Digital PDFs

EK-DFSLD-SD-002

May 1991

162 pages

Original

0.8MB

Document:	Fiber Distributed Data Interface System Level Description
Order Number:	EK-DFSLD-SD
Revision:	002
Pages:	162
Original Filename:

OCR Text

For: KNILES
Printed on: Thu, May 23, 1991 10:36:27
From book: Frontcover in 7X9SLD_5/12
Document: final_cover
Last saved on: Mon, May 20, 1991 08:52:11
Document: Title_final
Last saved on: Wed, May 22, 1991 13:05:44
Document: copyright_hw
Last saved on: Mon, May 20, 1991 08:59:Á15
Document: TOC
Last saved on: Wed, May 22, 1991 13:01:16
Document: Pref
Last saved on: Mon, May 20, 1991 09:03:14
Document: CHAP1
Last saved on: Wed, May 22, 1991 16:37:51
Document: CHAP2
Last saved on: Wed, May 22, 1991 11:47:57
Document: CHAP3
Last saved on: Wed, May 22, 1991 12:14:49
Document: CHAP4
Last saved on: Wed, May 22, 1991 12:13:43
Document: CHAP5
Last saved on: Wed, May 22, 1991 12:12:49
Document: CHAP6
Last saved on: Thu, May 16, 1991 13:29:42
(
...)

Fiber Distributed Data Interface

System Level Description

Order Number: EK–FDSLD–SD–002

Fiber Distributed Data Interface

System Level Description

May 1991
This manual introduces and describes Digital’s implementation
of Fiber Distributed Data Interface.

Order Number: EK–FDSLD–SD–002

EK–FDSLD–SD–002
MAY 1991

The information in this document is subject to change without notice and should not be construed as
a commitment by Digital Equipment Corporation. Digital Equipment Corporation assumes no responsibility for any errors that may appear in this document.
Copyright  1991 by Digital Equipment Corporation
All Rights Reserved.
Printed in U.S.A.

The following are trademarks of Digital Equipment Corporation:
DEBET
DEC
DEC FDDIcontroller
DECbridge
DECconcentrator
DECconnect
DECelms

DECnet
DECstation
DECsystem
DECUS
DECwindows
LAN Traffic Monitor
PDP
ThinWire

TURBOchannel
ULTRIX
UNIBUS
VAX
VAXcluster
VAXstation
VMS

This manual was produced by Telecommunications and Networks Publications.

Contents
Preface

Overview
1.1
1.2
1.2.1
1.2.2
1.2.3
1.3
1.3.1
1.3.2
1.3.3
1.4
1.4.1
1.4.2
1.5
1.5.1
1.5.2
1.5.3
1.5.4
1.5.5
1.6
1.6.1
1.6.2

FDDI Defined . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Background to FDDI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ASC X3T9.5 Task Group . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ANSI Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI and the OSI Model . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Single Attachment Stations (SAS) . . . . . . . . . . . . . . . . . . . . .
Dual Attachment Stations (DAS) . . . . . . . . . . . . . . . . . . . . . .
Concentrators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Application Environments . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Backbone . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Workgroup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Topologies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Standalone Concentrator . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tree of Concentrators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring of Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Homing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring of Trees, Topology of Choice . . . . . . . . . . . . . . . . . .
Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scalability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1–2
1–4
1–4
1–7
1–8
1–9
1–10
1–10
1–11
1–12
1–13
1–14
1–14
1–16
1–17
1–18
1–20
1–22
1–23
1–27
1–28

iii

1.6.3
1.7
1.7.1
1.7.2
1.7.3
1.7.4
1.8
1.9

1–28
1–28
1–29
1–29
1–29
1–30
1–30
1–31

Concepts
2.1
2.2
2.3
2.3.1
2.3.2
2.3.3
2.4
2.4.1
2.4.2
2.4.3
2.4.4
2.4.5
2.4.6
2.4.7
2.4.8
2.5
2.5.1
2.5.2
2.5.3
2.5.4
2.6
2.6.1
2.6.2
2.6.3

Maintainability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Digital Product Line . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
LAN Management Software . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
General Planning and Configuration . . . . . . . . . . . . . . . . . . . . .

Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physical Layer Medium Dependent (PMD) . . . . . . . . . . . . . . . .
Fiber Cable Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . .
Optical Bypass Relay . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Connector Keying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Physical Layer Protocol (PHY) . . . . . . . . . . . . . . . . . . . . . . . . .
Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Line States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4B/5B Data Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Modified NRZI Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . .
Clock Recovery and the Elasticity Buffer . . . . . . . . . . . . . . .
Smoothing Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Repeat Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Scrubbing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Media Access Control (MAC) . . . . . . . . . . . . . . . . . . . . . . . . . .
Ring Scheduling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Data Link Addressing Scheme . . . . . . . . . . . . . . . . . . . . . . . .
Universally Administered Addressing . . . . . . . . . . . . . . . . . .
Canonical Addresses and the Order of Transmission . . . . . . .
Protocol Data Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Tokens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Frame and Token Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2–1
2–2
2–3
2–4
2–6
2–7
2–7
2–8
2–11
2–12
2–13
2–13
2–14
2–14
2–14
2–15
2–16
2–16
2–17
2–18
2–19
2–20
2–21
2–21

2.7
2.7.1
2.7.2
2.7.3
2.7.4
2.7.5
2.8
2.9
2.9.1
2.9.2
2.9.3
2.9.4
2.10
2.10.1
2.10.2
2.11
2.11.1
2.11.2

2–23
2–23
2–25
2–27
2–28
2–29
2–29
2–30
2–30
2–30
2–31
2–33
2–36
2–37
2–38
2–40
2–40
2–41

Concentrators
3.1
3.2
3.2.1
3.2.2
3.2.3
3.3
3.4
3.5
3.5.1
3.5.2

Station Management (SMT) . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Connection Management (CMT) . . . . . . . . . . . . . . . . . . . . . .
Connection Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ring Management (RMT) . . . . . . . . . . . . . . . . . . . . . . . . . . .
SMT Frame Services . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
SMT Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Logical Link Control (LLC) . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ring Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Asynchronous Ring Transmission . . . . . . . . . . . . . . . . . . . . .
Timers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Target Token Rotation Time, Claiming, and Initialization . . .
Steady State Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Ring Maintenance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Beaconing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Trace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Digital Enhancements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Frame Content Independent Stripping Algorithm . . . . . . . . .
Ring Purger Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrator Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
MAC and MAC-less Concentrators . . . . . . . . . . . . . . . . . . . .
Station Bypass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Network Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrator Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrator Planning Considerations . . . . . . . . . . . . . . . . . . . .
Configuration Guidelines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrator Connection Rules . . . . . . . . . . . . . . . . . . . . . . .
Concentrators and Structured Wiring . . . . . . . . . . . . . . . . . . .

3–1
3–5
3–5
3–6
3–6
3–7
3–7
3–7
3–9
3–10

Bridges
4.1

Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

4–2

4.2
4.2.1
4.2.2
4.2.3
4.2.4
4.2.5
4.2.6
4.3
4.3.1
4.4
4.5
4.6
4.6.1

Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adapter Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Host Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adapter Manager . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Packet Memory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Accessing the Ring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Network Management Functions . . . . . . . . . . . . . . . . . . . . . .
Adapter Planning Considerations . . . . . . . . . . . . . . . . . . . . . . . .
Adapter Configuration Guidelines . . . . . . . . . . . . . . . . . . . . . . .

5–1
5–1
5–2
5–3
5–3
5–3
5–3
5–4
5–4
5–4
5–5
5–5

Network Management
6.1
6.2
6.2.1
6.2.2

4–4
4–4
4–5
4–5
4–6
4–6
4–7
4–7
4–8
4–9
4–10
4–10
4–10

Adapters
5.1
5.2
5.2.1
5.2.2
5.2.3
5.2.4
5.3
5.3.1
5.3.2
5.3.3
5.4
5.5

Bridge Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Translation Not Encapsulation . . . . . . . . . . . . . . . . . . . . . . . .
Frame Forwarding and Filtering . . . . . . . . . . . . . . . . . . . . . .
Source Address Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Spanning Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Network Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Internet Protocol Fragmentation . . . . . . . . . . . . . . . . . . . . . . .
Bridge Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Port . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Performance Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Bridge Planning Considerations . . . . . . . . . . . . . . . . . . . . . . . . .
Bridge Configuration Guidelines . . . . . . . . . . . . . . . . . . . . . . . .
General Connection Rules . . . . . . . . . . . . . . . . . . . . . . . . . . .

Network Management Benefits . . . . . . . . . . . . . . . . . . . . . . . . .
Network Management Functions . . . . . . . . . . . . . . . . . . . . . . . .
Network Configuration Management . . . . . . . . . . . . . . . . . . .
Fault Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

6–2
6–2
6–3
6–3

6.2.3
Performance Management . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2.4
Accounting Management . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.2.5
Security Management . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3
Digital’s Network Management Capabilities . . . . . . . . . . . . . . .
6.3.1
Device Naming . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.2
Device Registry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.3
Event Messages and Event Log . . . . . . . . . . . . . . . . . . . . . . .
6.3.4
Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.5
Configuration Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.6
Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.7
Bridge Port Enabling and Configuring . . . . . . . . . . . . . . . . . .
6.3.8
PHY Port Configuration and Control . . . . . . . . . . . . . . . . . . .
6.3.9
Address Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.10
Protocol Filtering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.11
Network Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6.3.12
FDDI Physical Ring Mapping . . . . . . . . . . . . . . . . . . . . . . . .
6.3.13
SMT Neighborhood Information Frame (NIF) Display . . . . .
6.3.14
SMT Station Information Frame (SIF) Display . . . . . . . . . . .
6.3.15
Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

OSI Reference Model
A.1
A.1.1
A.1.2
A.1.3
A.1.4
A.1.5
A.1.6

6–3
6–3
6–4
6–4
6–4
6–4
6–5
6–5
6–5
6–6
6–6
6–6
6–6
6–7
6–7
6–7
6–7
6–7
6–8

Physical Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Data Link Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Network Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Transport Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Session Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Presentation Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Application Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

A–3
A–3
A–3
A–3
A–4
A–4
A–4

Additional Configurations
B.1
B.2
B.3

Standalone Configurations Using Concentrators . . . . . . . . . . . .
Tree Configurations Using Concentrators and Bridges . . . . . . .
Dual Ring Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

B–1
B–3
B–6

vii

B.4
B.5

FDDI Dual Homing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B–8
Single-Mode Fiber Configurations . . . . . . . . . . . . . . . . . . . . . . . B–10

Related Documents

Glossary
Index
Figures
1–1
1–2
1–3
1–4
1–5
1–6
1–7
1–8
1–9
1–10
1–11
1–12
1–13
1–14
1–15
1–16
1–17
1–18
1–19
1–20
1–21

viii

Dual Ring of Trees Topology with Extended LANs . . . . . . . . .
FDDI Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI and Other LAN Standards . . . . . . . . . . . . . . . . . . . . . . . .
Relationship Between the OSI Model and FDDI . . . . . . . . . . . .
Single Attachment Station Model . . . . . . . . . . . . . . . . . . . . . . . .
Dual Attachment Station Model . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Concentrator Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Backbone Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Standalone Workgroup Configuration . . . . . . . . . . . . . . . . . . . .
FDDI Physical Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Standalone Concentrator Topology . . . . . . . . . . . . . . . . . . . . . .
Tree of Concentrators Topology . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring—Conceptual View . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring—Isolating a Link Failure . . . . . . . . . . . . . . . . . . . . .
Dual Ring—Isolating a Station Failure . . . . . . . . . . . . . . . . . . .
Dual Ring of Trees Topology . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Homing Topology Using Concentrators . . . . . . . . . . . . . .
Dual Homing Topology Using Concentrators and a DAS Bridge
Single Ring Failure of a non-FDDI Topology . . . . . . . . . . . . . .
Dual Ring Failure of an FDDI Topology . . . . . . . . . . . . . . . . . .
Concentrators in a Dual Ring of Trees Configuration . . . . . . . .

1–3
1–6
1–7
1–9
1–10
1–11
1–12
1–13
1–14
1–15
1–16
1–17
1–18
1–19
1–20
1–21
1–22
1–23
1–24
1–25
1–27

2–1
2–2
2–3
2–4
2–5
2–6
2–7
2–8
2–9
2–10
2–11
2–12
2–13
2–14
2–15
2–16
2–17
2–18
2–19
3–1
3–2
3–3
3–4
3–5
3–6
4–1
4–2
4–3
4–4
5–1
5–2
A–1
B–1
B–2

FDDI Network Connection Model . . . . . . . . . . . . . . . . . . . . . . .
Optical Bypass . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
MIC Receptacle Keying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Symbol Flow Through the PHY . . . . . . . . . . . . . . . . . . . . . . . . .
NRZI Encoding Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
48-Bit Address Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Universal Address Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Order of Transmission for Addresses . . . . . . . . . . . . . . . . . . . . .
Token Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Frame Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Station Management Architecture . . . . . . . . . . . . . . . . . . . . . . .
PHY- Port Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Connection Rules Matrix . . . . . . . . . . . . . . . . . . . . . . . . .
SMT Frame Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Logical Link Control and FDDI . . . . . . . . . . . . . . . . . . . . . . . . .
Claiming Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Capturing the Token . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Typical Ring Recovery . . . . . . . . . . . . . . . . . . . . . . . . . . .
Detection of Defective Station . . . . . . . . . . . . . . . . . . . . . . . . . .
Concentrator in a Standalone Topology . . . . . . . . . . . . . . . . . . .
Concentrators in a Dual Ring of Trees Topology . . . . . . . . . . . .
Concentrator Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Dual Ring of Concentrators in a Backbone Wiring System . . . .
Concentrators Cascaded in Equipment Rooms . . . . . . . . . . . . . .
Dual Ring of Trees in a Campus Environment . . . . . . . . . . . . . .
Extended LAN Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Bridge Functional Diagram . . . . . . . . . . . . . . . . . . . . . . .
SAS Bridge-to-Concentrator Connection . . . . . . . . . . . . . . . . . .
Bridge Distance Requirements for Multimode Connections . . .
Adapter Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Adapter Distance Requirements for Multimode Fiber . . . . . . . .
OSI Seven-Layer Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Standalone Workgroup Installation . . . . . . . . . . . . . . . . . . . . . .
Dedicated Network in a Dual Ring Installation . . . . . . . . . . . . .

2–3
2–6
2–7
2–9
2–13
2–16
2–17
2–18
2–20
2–21
2–24
2–25
2–26
2–29
2–29
2–32
2–34
2–37
2–39
3–2
3–3
3–4
3–8
3–11
3–12
4–2
4–3
4–8
4–11
5–2
5–6
A–2
B–2
B–3

B–3
B–4
B–5
B–6
B–7
B–8
B–9
B–10

FDDI Tree of Concentrators and Bridges, Campus
Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . B–4
Tree of Concentrators Installed in a Building . . . . . . . . . . . . . . . B–5
Dual Ring of Concentrators Installed in a Building . . . . . . . . . . B–6
Dual Ring of Concentrators and Bridges Installed in a Campus
B–7
Concentrators Installed for Dual Homing . . . . . . . . . . . . . . . . . . B–8
DAS Bridge Installed for Dual Homing . . . . . . . . . . . . . . . . . . . B–9
Connecting to the Dual Ring Through a Single-Mode Link . . . B–10
Dual Ring with Mixed Link Lengths . . . . . . . . . . . . . . . . . . . . . B–11

Tables
1–1
1–2
1–3
2–1
2–2
2–3
2–4
2–5

ANSI and ISO Standards for FDDI . . . . . . . . . . . . . . . . . . . . . .
FDDI General Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . .
Recommended Fiber Counts . . . . . . . . . . . . . . . . . . . . . . . . . . . .
Multimode Fiber Cable Specifications . . . . . . . . . . . . . . . . . . . .
Alternate Multimode Fiber Cable Specifications for FDDI . . . .
Single-Mode Fiber Cable Specifications . . . . . . . . . . . . . . . . . .
Symbol Encoding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
FDDI Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1–5
1–30
1–32
2–4
2–4
2–5
2–10
2–19

Preface

The Fiber Distributed Data Interface System Level Description describes the American National Standards Institute’s (ANSI) Fiber Distributed Data Interface (FDDI),
100-Mb/s, timed-token ring network standard. This manual also describes Digital
Equipment Corporation’s implementation of the ANSI standard. It discusses the various FDDI components designed and built by Digital and how these devices connect
to form an FDDI network.
This manual discusses the topologies recommended by Digital and the benefits to
users who implement them. In addition, this manual explains Digital’s approach to
network management and describes the facilities provided by network management
software.
Audience
This manual is intended for users who are familiar with basic network concepts and
who want an in-depth understanding of FDDI. By explaining FDDI and Digital’s
FDDI product set, this manual provides users with an understanding of the complexity of FDDI and the power of Digital’s products. After reading this document, users
will better understand the benefits that FDDI provides in the workplace.
Document Structure
This manual consists of six chapters and three appendixes as follows:
Chapter 1

Defines ANSI FDDI and provides background information
on how it evolved. This chapter also describes the FDDI
topologies recommended by Digital and lists Digital’s
FDDI product set. An overview of FDDI configuration
guidelines is included at the end of this chapter.

xii

Chapter 2

Discusses the basic design features of FDDI. This chapter
lists and defines the four standards that, taken as a whole,
make up ANSI FDDI. Topics covered include frame and
token structure, clocking, the physical connection of FDDI
stations, and basic ring operation.

Chapter 3

Contains a general description of concentrators and how
they are used in an FDDI environment. It emphasizes the
importance of the concentrator in Digital’s FDDI topology.
Configuration guidelines for concentrators are included at
the end of this chapter.

Chapter 4

Describes bridges and their role in the FDDI environment.
Bridge configuration guidelines are included at the end of
this chapter.

Chapter 5

Provides a general description of adapters and their use in
an FDDI environment. Adapter configuration guidelines are
included at the end of this chapter.

Chapter 6

Describes network management as it applies to FDDI and
the extended LAN environment. This chapter provides an
overview of network management capabilities. It also discusses concepts of Digital’s Enterprise Management Architecture (EMA).

Appendix A

Describes the International Organization for Standardization (ISO) seven-layer Open System Interconnect (OSI)
network architecture.

Appendix B

Contains examples of FDDI network configurations built
on Digital’s implementation of the commercial building
wiring scheme.

Appendix C

Contains a list of related documents that can be ordered
from Digital Equipment Corporation.

1
Overview

In an effort to standardize high-speed networking protocols, the American National
Standards Institute (ANSI) has defined a 100-Megabit-per-second (Mb/s) local area
network (LAN). Fiber optic cable is the first transmission medium defined in the
ANSI specification. Called Fiber Distributed Data Interface (FDDI), this ANSI
specification consists of a set of standards that define the components of FDDI:
Physical Layer Medium Dependent (PMD), Physical Layer Protocol (PHY), Media
Access Control (MAC), and Station Management (SMT). This manual describes
FDDI and addresses questions, including:

• Why did ANSI develop FDDI?
• What are the goals for FDDI?
• What problem is ANSI solving with FDDI?
This manual also discusses the benefits of adopting Digital’s implementation of
FDDI, based on the dual ring of trees network topology. These benefits include the
following:

• Reliability
• Scalability
• Maintainability
Digital’s implementation of FDDI complies with the ANSI standards as they are
written. This chapter provides a brief history of the FDDI specification. It also in-

1–1

cludes a description of the physical environments where FDDI is used and an introduction to the basic components of Digital’s implementation of FDDI.
Throughout this manual, the words station, device, and node are used extensively in
descriptions of equipment connected to the FDDI ring. Keep the following definitions for these words in mind:

1.1

Node

Any attachment device to the FDDI network. It can be a
single attachment station (SAS), a dual attachment station
(DAS), or a concentrator.

Station

As used in the FDDI environment, any node that possesses
a MAC entity.

Device

Generic hardware that connects to the FDDI network. Can
be a station, a node, or the equipment that connects to the
network through the FDDI components.

FDDI Defined
FDDI consists of a set of standards that, when taken together, define a 100-Mb/s,
timed-token passing, local area network that uses fiber optic medium to transmit data
between attached devices. Note that fiber optic medium is the first medium defined
by the standards. Work is underway to add other media. The standards define several
types of networking devices. These include concentrators, dual attachment stations,
and single attachment stations. These devices allow users to construct various network configurations based on the FDDI standard.
Digital Equipment Corporation recommends a basic network topology, called a dual
ring of trees, that provides a backup physical link for increased reliability. In this topology, an FDDI ring is constructed as trees of stations connected to the FDDI dual
ring. Figure 1–1 illustrates this dual ring of trees topology. This figure shows how the
various FDDI components connect to form an integrated, fully functional network.
In this topology, dual attachment stations and single attachment stations connect to
the FDDI ring through concentrators forming the branches of the trees.
Figure 1–1 emphasizes the importance of the concentrator as the key component in a
dual ring of trees topology. In Figure 1–1, the dual ring connects concentrators. Stations connect to the dual ring through a tree of concentrators.This topology offers a
campus-wide network free from disruption by everyday events.

1–2

Fiber Distributed Data Interface System Level Description

Network management software provides the tools to maintain and manage the network. Other sections of this manual provide details on the benefits of this topology
and how SAS and DAS bridges connect FDDI LANs to extended LANs.
Figure 1–1:

Dual Ring of Trees Topology with Extended LANs

Concentrator

Dual ring

Location of dual ring

Concentrator

• ••

Concentrator

Tree

Node

Concentrator

FDDI bridge
•••

Node
Bridge connects FDDI
to extended LAN

Concentrator

Node

FDDI bridge

Node

LKG-5102–91I

Overview

1–3

Digital, by making its implementation fully compliant with the ANSI standards,
achieves interoperability with other vendors whose products are also truly ANSI
compliant. By maintaining compliance with the standard, Digital also makes it easier
for users to design networks using off-the-shelf components, rather than exotic, proprietary hardware and software.

1.2 Background to FDDI
FDDI is an ANSI standard for a high-speed, general-purpose network for connecting
large computers, workstations, desktop computers, and other equipment. ANSI defines and oversees the standards creation process and also charters other organizations to write standards for domestic use. This section explains the various ANSI
specifications that make up FDDI. Section 1.2.3 maps the ANSI standards to the
ISO/OSI model.
1.2.1 ASC X3T9.5 Task Group

ASC X3T9.5 is the ANSI Task Group responsible for the design of the overall ANSI
FDDI specification. The Accredited Standards Committee (ASC), X3, develops
standards for Information Processing Systems. Technical Committee X3T9 is responsible for writing standards for I/O interfaces. The X3T9.5 Task Group is chartered to write the FDDI standard.
The ANSI standards, and related ISO versions, that compose the FDDI specification
are listed in Table 1–1 and described in the following sections. These sections are
overviews of the individual standards. For further information refer to the individual
ANSI or ISO document for each standard.

1–4

Fiber Distributed Data Interface System Level Description

Table 1–1:

ANSI and ISO Standards for FDDI

ANSI/ISO Standard

FDDI Standard

X3.166–1990/
ISO 9314-3:1990

Physical Layer Medium Dependent —This standard corresponds to the lower portion of the OSI Physical layer. PMD defines the transmit/receive power levels, optical transmitter and
receiver interface requirements, error rates, and cable and connector specifications.

X3.148–1988/
ISO 9314-1:1989

Physical Layer Protocol —This medium-independent standard corresponds to the upper portion of the Physical layer. PHY
defines symbols, line states, encoding/decoding techniques,
clocking requirements, and data framing requirements.

X3.139–1987/
ISO 9314-2:1989

Media Access Control —This standard corresponds to the
lower portion of the OSI Data Link layer. MAC defines data link
addressing, frame formatting and checking, medium access,
error detecting, and token handling. Its primary function is to deliver formatted data to nodes attached to the FDDI LAN.

X3.T9.5/84–49
(under development)

Station Management —This draft standard defines the system
management services for the FDDI protocols. SMT includes facilities for connection management, node configuration, recovery from error conditions, and the encoding of SMT frames.

Overview

1–5

Figure 1–2 shows the relationship among the various FDDI standards and how they
combine to form the complete FDDI standard. Figure 1–3 shows the relationship
among the FDDI standards and other common industry standards. Note that the
802.2 Logical Link Control and IEEE 802.1 standards are common across the various lower layer standards.
Figure 1–2:

FDDI Model

Logical Link Control
(IEEE 802.2 LLC)

Data
Link
layer

Physical
layer

Media Access Control
(MAC)
• Addressing
• Frame construction
• Token handling
Physical Layer Protocol
(PHY)
• Encoding/decoding
• Clocking
• Symbol set

Station Management
(SMT)

•
•
•
•

Ring monitoring
Ring management
Connection management
SMT frames

Physical Media Dependent
(PMD)
• Physical link parameters
• Connectors and cabling
LKG–4042–90I

1–6

Fiber Distributed Data Interface System Level Description

Figure 1–3:

FDDI and Other LAN Standards

IEEE 802.1

Data
Link
layer

Logical Link Control
(IEEE 802.2 LLC)
MAC

PHY

SMT

Physical
layer

IEEE
802.3
CSMA/CD

IEEE
802.4
Token
bus

IEEE
802.5
Token
ring

PMD
FDDI

LKG–4043–90I

1.2.2 ANSI Specifications

The ANSI standards that make up the FDDI specification include X3.166–1990,
X3.148–1988, X3.139–1987, and X3.T9.5/84–49. Together, these standards define
a fully functional high-speed fiber optic network. Chapter 2 describes the individual
standards in greater detail. Following are summaries of the FDDI standards.
X3.166–1990 defines the Physical Layer Medium Dependent (PMD) sublayer. This
sublayer corresponds to the lower portion of the Physical layer of the OSI seven-layer model. This standard provides the specifications for the optical transmitter and receiver used by FDDI, the power levels, allowable bit error rates, fiber optic
transmission wavelength, fiber optic cable, and connectors.
X3.148–1988 defines the Physical Layer Protocol (PHY). This medium-independent layer corresponds to the upper portion of the OSI Physical layer. This standard
defines the encoding scheme, clock synchronization, and serial-parallel conversion.

Overview

1–7

X3.139–1987 defines the Media Access Control (MAC) sublayer. The MAC corresponds to the lowest sublayer of the OSI Data Link layer (DLL). This standard defines medium access, addressing, and error detection protocols. Its primary function
is the delivery of frames to attached nodes on the FDDI ring. It also provides access to
the ring for higher level protocols, specifically, the Logical Link Control (LLC) protocol.
X3.T9.5/84–49 defines the system management applications for the FDDI protocol
layers. Called Station Management (SMT), it provides the specifications for the station-level control necessary to ensure proper node operation in an FDDI ring. These
include topology, station configuration, ring management, and connection management.
1.2.3 FDDI and the OSI Model

The Open Systems Interconnection (OSI) model is a reference for defining the architecture necessary to provide communications between open systems. Developed by
the International Organization for Standardization (ISO), it provides a framework
for the development of international standards for computer communications. The
OSI model is based on the layering technique where the various communications
functions are partitioned into seven discrete units. This section describes the relationship between the OSI layered model and the FDDI standard. For further information concerning the OSI model, refer to Appendix A.
The FDDI standards apply to the Physical and Data Link layers of the OSI model.
Figure 1–4 shows the relationship between the OSI model and the FDDI standards.
In the OSI model, the Physical layer defines data transmission over the physical link.
Issues such as bit signaling, mechanical and electrical specifications, cabling, and
connectors are covered here. In FDDI this is further divided into the PHY and PMD
sublayers, which separate the media and transmission details into two distinct parts.
The OSI Data Link layer defines parameters for medium access, frame construction,
data link addressing, error detection, and connection to higher layers. FDDI divides
this layer into two sublayers. The upper sublayer, Logical Link Control, is provided
by the IEEE 802.2 LLC standard. This is the same LLC used in IEEE 802.3, .4, and
.5. It provides a means for exchanging data between LLC users across an FDDI
LAN.
The lower sublayer is defined in the MAC and SMT standards in FDDI. This sublayer defines the data link packets specifically required by the FDDI token passing

1–8

Fiber Distributed Data Interface System Level Description

scheme. These include addressing and token handling. In Figure 1–4, Service Access Points (SAPs) are logical connections between layers in the OSI model. They
are shown here for informational purposes only.
Figure 1–4:

Relationship Between the OSI Model and FDDI

Application
Presentation

Service
Access
Points
(SAPs)

Session
Transport

Logical Link Control

Network

MAC

Data Link

PHY

Physical

PMD

OSI model

SMT

FDDI standards
LKG–4044–90I

1.3 FDDI Devices
The FDDI standard defines two types of devices: stations and concentrators. Stations
can be further divided into single attachment stations (SAS) and dual attachment stations (DAS). The concentrator provides connections for the attachment of multiple
nodes to the FDDI ring. The concentrator can be a single attachment concentrator
(SAC), dual attachment concentrator (DAC), or a standalone concentrator. All of
these device types act as connection points to the FDDI network. Digital implements
the SAS, DAS, and DAC in the dual ring of trees topology. The following sections

Overview

1–9

describe the attachment devices defined by FDDI. Chapter 2 provides a detailed discussion of the components defined by the FDDI standards.
1.3.1 Single Attachment Stations (SAS)

The single attachment station (SAS) permits access to the FDDI ring through the
concentrator. The SAS provides a single connection point to the FDDI ring. It does
not have the facilities for connecting to the dual ring; therefore, it cannot wrap the
ring in case of a station or fiber failure. The concentrator provides fault control. In the
dual ring of trees topology, the SAS proves to be a reliable, cost-effective method for
connecting to the FDDI network.
The components of the SAS are a single instance of the MAC entity, the PHY, the
PMD, and the SMT. Figure 1–5 shows a model of the SAS.
Figure 1–5:

Single Attachment Station Model

Station Management (SMT)

Receive
MAC

PHY

PMD

To concentrator
Transmit
LKG–4045–90I

1.3.2 Dual Attachment Stations (DAS)

The dual attachment station (DAS) permits a device to connect to the primary and
secondary rings of the FDDI dual ring, allowing a device to access the dual ring architecture of FDDI. The DAS consists of two PHYs, two PMDs, one or (optionally)
two MACs, one SMT, and an optical bypass relay (optional). Figure 1–6 shows a
model of the DAS.

1–10

Fiber Distributed Data Interface System Level Description

In the dual ring of trees topology employed by Digital, DAS nodes are supported, but
their use is not required. Concentrators, SAS and DAS bridges, and SAS nodes compose a typical dual ring of trees topology as implemented by Digital.
Figure 1–6:

Dual Attachment Station Model

Station Management (SMT)

MAC
(optional)

MAC

PHY 2

PMD 2
Bypass
relay
(optional)

PHY 1

To primary
FDDI ring
To secondary
FDDI ring

PMD 1

LKG–4046–90I

1.3.3 Concentrators

The concentrator is also specified as part of the ANSI standards. The concentrator
allows the attachment of multiple SAS, DAS, or concentrators to the primary (or optionally, secondary) ring of the FDDI network. The concentrator can also be used as
the root of a tree topology. As shown in Figure 1–7, the concentrator consists of SMT
circuitry, an optional MAC entity, and several PHY/PMD entities. A complete physical connection consists of a PHY/PMD in a concentrator connected by a medium to
the PHY/PMD in another node.
An FDDI concentrator retimes data, detects faults, performs fault isolation, and is
manageable by network management software. Unlike concentrators found in other
networks (IEEE 802.5, for example), the FDDI concentrator is an active device that
can actually control the physical topology of a network.

Overview

1–11

The concentrator can reconfigure the network by inserting or removing connected
nodes. The configuration switch shown in Figure 1–7 provides this capability to the
concentrator.
In Figure 1–7, the individual PHY/PMD entities have separate paths to the SMT, as
indicated by the dashed lines through subsequent blocks in the figure. Chapter 3 provides a closer look at the concentrator as it is used by Digital in the dual ring of trees
topology.
Figure 1–7:

FDDI Concentrator Model

MAC
(optional)
Station
management
(SMT)

Configuration switch

PHY 1

PHY 2

PHY 3

PHY 4

PHY n

PMD 1

PMD 2

PMD 3

PMD 4

PMD n

To
primary and
secondary
ring

To attached devices
LKG–4047–90I

1.4 Application Environments
Digital’s implementation of FDDI supports various LAN environments including
backbones and workgroups. These environments provide great flexibility in the design and implementation of an FDDI network. The following sections and supporting illustrations further describe these network environments.

1–12

Fiber Distributed Data Interface System Level Description

1.4.1 Backbone

Figure 1–8 shows an FDDI backbone configuration. This configuration ties together
multiple LANs in a campus or multibuilding environment into one integrated network. The backbone provides a high-bandwidth facility for the connection of FDDI
devices. Because attached devices can be subjected to frequent power down by the
user, connection to the FDDI ring is optimized by the use of concentrators. Backbones provide high reliability; the use of concentrators maintains ring integrity even
when one or all devices attached to it are shut down or unplugged. The backbone supports radial wiring schemes such as the one defined in EIA/TIA 568 (TR–41.8.1).
Figure 1–8:

Backbone Configuration

Building 1

Building 2

IDF
Concentrator

MDF
Concentrator

IDF
Concentrator

HDF

Concentrator

SAS

DAS
Workgroup
HDF = Horizontal distribution frame
IDF = Intermediate distribution frame
MDF = Main distribution frame

Bridge

HDF
Concentrator

SAS

DAS
Bridge

SAS = Single attachment station
DAS = Dual attachment station
LKG–4048–90I

Overview

1–13

1.4.2 Workgroup

The workgroup configuration connects workstations, personal computers, and minicomputers through one or more concentrators. This network is characterized by a
relatively small number of attached devices spread over a limited geographical area
and with FDDI’s ability to transfer large amounts of data quickly and reliably.
The typical workgroup configuration normally has distances between stations less
than the 2 kilometers (1.2 miles) specified in the FDDI standards. The dual ring of
trees architecture employed in the FDDI ring allows network to sustain the loss of
one or more attached nodes and still function. A workgroup can be part of a backbone configuration as shown in Figure 1–8. Figure 1–9 illustrates a standalone
workgroup configuration using a concentrator.
Figure 1–9:

Standalone Workgroup Configuration

Concentrator

SAS

SAS
LKG–4050–90I

1.5 FDDI Topologies
An FDDI ring network consists of a number of serially attached stations that are connected by a transmission medium to form a closed loop. An active station transmits
information serially as a stream of symbols to the next active station on the ring. As
each active station receives these symbols, it regenerates and repeats them to the next
active device on the ring (its downstream neighbor).

1–14

Fiber Distributed Data Interface System Level Description

The physical connection (see Figure 1–10) is basic to the FDDI ring design: Through
this connection attached devices pass information over the ring. An FDDI physical
connection is formed between the Physical layers (PHY/PMD) of two stations connected by the medium. FDDI requires that all connections in the ring be point-topoint, full-duplex, and bidirectional. Each attachment to the network has a transmit
and receive component. For example, a valid fiber optic connection contains two fibers providing a point-to-point, full-duplex path.
Figure 1–10: FDDI Physical Connection

Physical connection
Medium
Station A
PHY

Station B
PMD

PMD

To MAC

PHY
To
MAC

Cable
plant

LKG–4051–90I

As part of the FDDI standard, ANSI permits a number of topologies. Of these topologies, four are of particular importance to the users of Digital’s FDDI products. These
topologies are as follows:

•

Standalone concentrator with attached nodes

•

Tree of concentrators

•

Dual counter-rotating ring

•

Dual ring of trees

These topologies are discussed in detail in the following sections.

Overview

1–15

1.5.1 Standalone Concentrator

The standalone concentrator topology consists of a single concentrator and its attached stations (Figure 1–11). These stations can be either SAS or DAS devices. Because of their lower cost, SAS devices are the preferred station for this topology. This
topology does not require access to the FDDI dual ring. It offers all the advantages
associated with the concentrator, including reliability and ease of configuring, to
small, independent workgroups. This topology allows the use of star wiring. This
means existing structured fiber optic wiring can be used, affording significant cost
savings in prewired sites. A standalone concentrator topology grows into a tree of
concentrators as the network is expanded.
Figure 1–11: Standalone Concentrator Topology

SAS

Concentrator

SAS

DAS

LKG–4056–90I

1–16

Fiber Distributed Data Interface System Level Description

1.5.2 Tree of Concentrators

The tree of concentrators, shown in Figure 1–12, is used when wiring large groups
of user devices together. Concentrators are wired in a star topology with one concentrator serving as the root of the tree. Additional concentrators can connect to the second tier of concentrators as needed to support new users. This topology can be used
to wire a single building or to service many stations on one floor of a building. This
topology lends itself to structured wiring systems.
Figure 1–12: Tree of Concentrators Topology

Root
concentrator

Concentrator

SAS

SAS
Concentrator

SAS
LKG–4236–90I

Overview

1–17

1.5.3 Dual Ring

One of the basic concepts in the FDDI standard is the dual counter-rotating ring. The
dual ring consists of a primary ring and a secondary ring. The direction of data flow
in the secondary ring is opposite (counter to) that of the primary ring. The main functions of this secondary path are to assist in the ring initialization and reconfiguration
process and to provide backup to the first ring. Inherent in this design is the ability to
continue ring transmission if a device on the ring fails. The ring is restored by wrapping to the secondary ring to maintain the transmission path. This redundancy in the
ring design provides a degree of fault tolerance not found in other network standards.
Figure 1–13 illustrates the basic concepts of the dual ring.
Figure 1–13: Dual Ring—Conceptual View

Station 1

Primary ring

Station 2

Secondary ring

Total fiber length = 200 km/124 mi (100 km/62 mi x 2 rings)

Station 4

Station 3

All devices on the ring are dual attachment stations or dual attachment
concentrators.
LKG–4053–90I

1–18

Fiber Distributed Data Interface System Level Description

Dual attachment stations or dual attachment concentrators on the dual ring connect
to each other by means of the primary and secondary fiber cables. Under normal conditions, the secondary ring is idle with no data being transmitted. If a fiber failure
occurs, the stations on either side of the failure reconfigure by wrapping the primary
ring to the secondary ring, effectively isolating the fault. This maintains connectivity
between stations in the dual ring, allowing normal operation to continue.
Figure 1–14 shows the FDDI technique of isolating a link failure.
Figure 1–14: Dual Ring—Isolating a Link Failure

Station 1

Primary ring

Station 2

Secondary ring

Wrapped configuration

Break in the ring
Station 4

Station 3

All devices on the ring are dual attachment stations or dual attachment
concentrators.
LKG–4054–90I

If an attached station fails, again the stations on either side of the failing station reconfigure, effectively isolating the failing station from the ring. Figure 1–15 shows
the FDDI technique of isolating a station failure.
FDDI limits total fiber length to 200 km (124 mi). Since the dual ring topology effectively doubles media length in the event of a ring wrap, the actual length of each ring
is limited to 100 km (62 mi).

Overview

1–19

Figure 1–15: Dual Ring—Isolating a Station Failure

Station 1

Primary ring

Station 2

Secondary ring

Wrapped configuration

Defective station
Station 4

Station 3

All devices on the ring are dual attachment stations or dual attachment
concentrators.
LKG–4055–90I

1.5.4 Dual Ring of Trees

A fourth topology described in FDDI, the dual ring of trees (DRT), connects concentrators and DAS bridges located at a Main Distribution Frame (MDF) together in a
dual ring. This provides the reliability of the dual ring where it is needed most, in the
campus backbone, while allowing for the use of structured wiring in all other campus locations. Trees of concentrators then connect to the dual ring, providing connections at the Intermediate Distribution Frames (IDF) and Horizontal Distribution
Frames (HDF) of the structured wiring system. Stations then connect to concentrators located in the HDFs of the structured wiring system. The tree can branch as required by simply adding concentrators as long as the station number or ring distance
limits are not exceeded. Figure 1–16 shows the DRT topology.

1–20

Fiber Distributed Data Interface System Level Description

Figure 1–16: Dual Ring of Trees Topology

Concentrator

Node

Node
Concentrator

Node

Node
Concentrator

Node

Node
Node

Node
LKG–4052–90I

Overview

1–21

The concentrator plays a key role in fault isolation in this topology. The dual ring at
the backbone guarantees a secondary data path to all concentrators and dual attachment stations in the dual ring. Beyond this level, in the trees, individual stations can
be removed from the ring as needed with no interruption of service to the rest of the
network. The concentrators can bypass inactive or defective stations, as required.
1.5.5 Dual Homing

The dual homing topology uses backup links to enhance fault tolerance in a tree topology. One link is called the primary link, and the other link is called the backup
link. Figure 1–17 shows a dual homed topology using concentrators. If concentrator
2 or the primary link to concentrator 3 fails, the backup link to concentrator 3,
through concentrator 1, is activated. This ensures that the devices connected to concentrator 3 will have uninterrupted service. Activation of the individual links follows
the FDDI connection rules, described in Section 2.7.2.
Figure 1–17: Dual Homing Topology Using Concentrators

Dual
ring

DAS
Concentrator 1

Concentrator 2
Tree

SAS

Backup
link

Primary link

Concentrator 3

SAS
LKG-4812-91I

1–22

Fiber Distributed Data Interface System Level Description

Figure 1–18 shows a dual homed topology using concentrators and a DAS bridge. If
concentrator 2 or the primary link to the DAS bridge fails, the backup link to the DAS
bridge, through concentrator 1, is activated. This ensures that the devices connected
to the DAS bridge will have uninterrupted service. Activation of the individual links
follows the FDDI connection rules, described in Section 2.7.2.
Figure 1–18: Dual Homing Topology Using Concentrators and a DAS Bridge

Dual
ring

DAS
Concentrator 1

Concentrator 2
Tree

SAS

Backup
link

Primary link
DAS Bridge
External non-FDDI network

Node

Node
LKG-4934-91I

1.6 Dual Ring of Trees, Topology of Choice
Network failure problems that are inherent in a single ring topology can be corrected
by implementing a dual ring topology. However, a dual ring topology creates segments if multiple station or cable failures occur. A dual ring of trees topology using
concentrators corrects the problems that are inherent in both single and dual ring designs. This section explains Digital’s choice of the dual ring of trees as the recommended topology for FDDI.

Overview

1–23

A single ring topology consists of a series of point-to-point physical links in a closed
loop. If a failure on the ring occurs, the entire network fails. Figure 1–19 shows the
single-point failure problem in a non-FDDI single-ring topology. In large installations where the probability of ring or station failure is high, this type of disruption is
unacceptable.
Figure 1–19: Single Ring Failure of a non-FDDI Topology

Station

Single failure
results in failed
ring.

Station

Station
LKG–4057–90I

To eliminate the single-point-of-failure problem, the ANSI FDDI standards committee designed the dual ring topology. This topology solves the problem of a single failure disrupting the ring by providing a secondary transmission path. The secondary
transmission path assists in start-up, initialization, and reconfiguration of the primary path.

1–24

Fiber Distributed Data Interface System Level Description

Although the counter-rotating ring design eliminates the problems associated with
the single ring, a new issue arises when multiple failures occur. Figure 1–20 illustrates what happens to a ring of DAS devices when multiple ring failures or disconnects occur. The result is two or more (depending on number of failures) segmented
rings.
Figure 1–20: Dual Ring Failure of an FDDI Topology

Station

Two or more failures
result in segmented
rings with each ring
functioning independently of the other.
Station

Station

LKG–5095–90I

Although each ring is fully functional, the rings cannot access each other because of
the multiple failures. These breaks can be caused simply by a user disconnecting a
workstation. Again, in large installations where the risk of multiple failures increases
with the number of stations or large number of users added to the network, this is not
an acceptable situation.

Overview

1–25

The dual ring solution can also give rise to other concerns such as:

•

The dual ring is not well suited to the use of structured wiring schemes.

•

Dual attachment stations require twice the connectors and cables at the drop
site.

•

The site backbone cabling is accessible to each user on the dual ring, which
presents a ring integrity risk at every user location.

•

Because each station is part of the backbone wiring, the behavior of each user
is critical to the operation of the ring.

•

Manual intervention and manipulation of the dual ring cabling is required for
all additions, moves, and changes, resulting in periods of ring instability.

Digital recognizes the segmented ring issue inherent with DAS installations and addresses it with the use of FDDI concentrators configured in a dual ring of trees. The
DRT topology allows the creation of a large, sophisticated network that can sustain
the loss of all stations connected to the concentrators without losing ring integrity. In
this topology, concentrators electronically bypass ring disruptions. Connection
Management allows the removal of defective links with no disruption of service to
remaining nodes. Figure 1–21 illustrates this concept. It shows two concentrators attached to the dual ring with two bypassed stations.
Only network components (concentrators or bridges) should be connected to the
dual ring. These devices are typically in locked closets or cabinets and only accessible by network-knowledgeable individuals. End user stations, such as computers or
workstations, should only be connected as SAS.
Further inspection of the DRT reveals it to be a more robust and dynamic solution to
configuration problems than either the single or dual ring approaches to network design. With the dual ring of trees, the best of both topologies is available to network
users. The dual ring implemented at the root of the tree provides transmission path
redundancy to attached concentrators. The single ring cabling between concentrators and SAS stations or other concentrators on the tree branches provides significant
cost savings compared to the purchase of DAS devices and the installation of multiple dual fiber cables at all user sites.

1–26

Fiber Distributed Data Interface System Level Description

Figure 1–21: Concentrators in a Dual Ring of Trees Configuration

DAS
Concentrator

Concentrator

Dual
ring

Concentrator

Node Node*
Concentrator

Node • • •

Node*

* Multiple station failures in a dual ring of trees result in no loss of access
to the ring. Failing stations are simply removed from the ring.
LKG–4237–90I

Key points in the choice of the dual ring of trees topology are reliability, scalability,
and maintainability. These factors provide the power not found in the other topologies.
1.6.1 Reliability

A dual ring of trees provides a backup physical link at the dual ring. In addition, concentrators in the dual ring or in the tree can bypass an inactive or defective station and
allow the insertion or removal of stations as needed. This is done without the intervention of the network manager and without disrupting the network.

Overview

1–27

Because the network is unaffected by single- or multiple-point failures, it is always
available to attached stations. Barring a power failure or major natural disaster, there
is no ordinary incident that would cause the loss of ring access to other stations connected to the network. The dual ring of trees allows any number of stations to be removed from the ring without denying ring availability to remaining attached
stations.
1.6.2 Scalability

Increasing or decreasing ring size is a simple matter of adding or removing concentrators. In a dual ring topology, recabling is necessary every time a station is added or
removed. The inherent flexibility of the concentrator allows network growth to occur as needed; recabling is not necessary in a dual ring of trees topology. Starting
with a single concentrator and attached SAS devices, a user can expand the network
to fit the needs of the site as long as the station number and ring distance limit are not
exceeded. The dual ring of trees permits additions, moves, and changes without disrupting the network. This flexibility in wiring complies with the wiring scheme specified in the EIA/TIA 568 standard.
1.6.3 Maintainability

Proper maintenance is critical to the operation of any network. Digital provides intelligent concentrators and distributed networking tools that permit the user to monitor
and maintain the network down to individual attached devices. Proper installation in
dedicated equipment rooms allows the concentrators to be carefully controlled and
maintained. The physical topology recommended by Digital meets the specifications set forth in the EIA/TIA 568 standard.

1.7 Digital Product Line
The Digital implementation of ANSI FDDI allows for great flexibility in the design
and construction of a high-speed fiber network. Digital designs and builds a number
of products for the construction of an FDDI ring. They are fully compliant with the
ANSI FDDI standards. The following sections provide an overview of the FDDI devices produced by Digital.

1–28

Fiber Distributed Data Interface System Level Description

1.7.1

Concentrators

Concentrators are necessary in large networks where DAS configurations are difficult to configure and maintain, and are expensive and unreliable. Digital implements
the DECconcentrator 500 as the recommended device for connecting stations to the
FDDI ring. Digital supports the use of a DAS configured into a dual ring or as a member of a tree.
The DECconcentrator 500 complies with the ANSI X3T9.5 standards. It provides
for attachment of nodes (SAS, DAS, or concentrator) in a standalone environment.
The concentrator also supports the hierarchical building and campus physical wiring
topologies discussed in previous sections by allowing concentrators to be interconnected in a tree topology. Finally, it supports the dual ring interconnection of concentrators and DAS, referred to as the dual ring in the ANSI standards. Refer to Chapter
3 for details concerning concentrator functionality.
1.7.2 FDDI Bridges

To preserve previous investments in equipment, the FDDI LAN must be able to interface with existing network technologies. Bridges provide a means to do this. In
Digital’s implementation of FDDI, the DECbridge 500/600 series provides the interface necessary to connect the 100-Mb/s FDDI ring to the 10-Mb/s, 802.3/Ethernet
LAN. The DECbridge 500/600 series contains all necessary buffering and translation logic for interfacing to the 802.3/Ethernet network. In addition, the DECbridge
600 series allows multiple 802.3/Ethernet segments to be connected to the FDDI network.
The DECbridge 500/600 series contains SAS and DAS bridges. SAS and DAS
bridges interconnect in a tree topology and in a dual ring of trees topology through a
concentrator. Additionally, DAS bridges interconnect in a dual ring topology. Refer
to Chapter 4 for further details concerning the bridge.
1.7.3 Adapters

Adapters connect different devices to the FDDI ring through the concentrator.
Adapters can be SAS or DAS. An adapter connects to a device such as a workstation
or minicomputer with an I/O bus, for example, a TURBOchannel bus. The adapter
contains all necessary logic to act as an interface between the FDDI ring and device
system memory. Refer to Chapter 5 for more details on adapter functionality.

Overview

1–29

1.7.4 LAN Management Software

Proper network management is critical to the successful implementation of a LAN.
To this end, Digital has developed sophisticated LAN management software. This
software allows a user at a management station to control and observe bridges, FDDI
concentrators, and other nodes in the extended local area network. Network managers use the management software to monitor and control a network; field service
technicians use it to troubleshoot the LAN. Refer to Chapter 6 for details on network
management and the Digital management philosophy.

1.8 FDDI Specifications
FDDI allows the creation of a flexible, robust high-performance network. Table 1–2
lists the specifications for FDDI as set forth by Digital and the ANSI standards. For
detailed information concerning building or site-cabling procedures and requirements, refer to the DECconnect System Fiber Optic Planning and Configuration and
Fiber Optic Installation manuals.
Table 1–2:

FDDI General Specifications

Item

Specification

Transmission rate

125 megabaud (100 Mb/s at the data link)

Physical layer entities

1000 (maximum)

Fiber length

200 km (124 mi) (maximum)

Link loss budget

11.0 dB (maximum loss — multimode fiber optic cable)
22 dB (maximum loss — single-mode fiber optic cable)

Interstation link length*

2 km (1.2 mi) (maximum — multimode fiber optic cable)
40 km (24.8 mi) (maximum — single-mode fiber optic cable)

Transmission medium

Fiber optic cable

Network topology

Dual ring of trees

Media access method

Timed-token passing

*FDDI link length is bandwidth limited to 2 km (1.2 mi) for multimode fiber.

The number of stations that can be installed in an FDDI LAN depends on the number
of Physical layer entities that are present. A Physical layer entity consists of one instance of PHY/PMD and is the equivalent of one PHY port. A station can have one
Physical layer entity (SAS), two Physical layer entities (DAS), or multiple Physical
layer entities (concentrators).

1–30

Fiber Distributed Data Interface System Level Description

The ANSI FDDI standard specifies a maximum of 1000 physical layer entities and
1000 MAC entities in an FDDI LAN. The standard allows up to 500 stations depending on the topology. Digital supports the configurations specified in the ANSI FDDI
standard. However, Digital recommends that an FDDI LAN generally be limited to
100 stations and 20 concentrators due to performance and reliability characteristics
of large rings.

1.9 General Planning and Configuration
This section lists some general points to consider when planning an FDDI network. It
does not, however, address all network planning requirements. For further details regarding the planning and installation of Digital fiber optic cable, refer to the following two manuals.

•

DECconnect System Fiber Optic Planning and Configuration

•

DECconnect System Fiber Optic Installation

In the initial planning stages, it is important to determine proper fiber counts at FDDI
device locations. Consider the following points when determining fiber count:

•

Multiple services support

•

Level of redundancy

•

Planned growth and expansion

Table 1–3 lists the minimum recommendations for fiber optic cable between various
network components. The drop locations are defined according to the EIA/TIA 568
standard.

Overview

1–31

Table 1–3:

Recommended Fiber Counts
Minimum Recommended Fiber Count

Drop Location

Multimode

Single Mode

MDF to IDF

IDF to HDF

HDF to office wallbox

n/a

MDF = Main Distribution Frame
IDF = Intermediate Distribution Frame
HDF = Horizontal Distribution Frame
n/a = not applicable

1–32

Fiber Distributed Data Interface System Level Description

2
Concepts

This chapter explains the basic concepts of FDDI, giving a general overview of a
complex design. Topics include Protocol Data Units (PDUs), Target Token Rotation
Time (TTRT), and signaling techniques. It concludes with a discussion of ring operation. This chapter provides basic information about how FDDI operates and why certain functions were included in the FDDI design. It also discusses FDDI in terms of
the four standards that compose FDDI:

•

Physical Layer Medium Dependent (PMD)

•

Physical Layer Protocol (PHY)

•

Media Access Control (MAC)

•

Station Management (SMT)

While reading about the intricacies of FDDI, keep in mind that Digital has incorporated the power of FDDI in a standard chipset that is used throughout the Digital
family of FDDI products.

2.1 Overview
The ANSI FDDI committee reviewed many options during the development process. Key design concepts of FDDI include the following:

•

Timed-Token Protocol

•

Symbols for data transmission and line state signaling

2–1

•

4B/5B and NRZ/NRZI encoding

•

Synchronous, asynchronous, and restricted service

•

Token rotation time determined by negotiation among all connected stations

•

Immediate transmission of data after token acquisition

•

Distributed clocking to support a large number of stations on the ring

Because of the requirement for a 100-Mb/s transmission rate (125 megabaud), FDDI
presents a new set of challenges to designers. By making use of the techniques noted
above, designers have enabled the FDDI network to maintain a high transmission
rate over great distances among a large number of stations.

2.2 FDDI Structure
FDDI is built on four sublayers as defined in the ANSI standards. Sections 2.3
through 2.7 discuss these items in detail. Figure 2–1 illustrates the basic FDDI network connection model and shows how these sublayers interact. The four standards
that make up FDDI are the PMD, PHY, MAC, and SMT.
These four sublayers plus the IEEE 802.2 Logical Link Control (LLC) standard
combine to provide essential networking services to devices attached to the FDDI
network. These services include hardware connection, ring initialization, error detection, token handling, and overall station management.

2–2

Fiber Distributed Data Interface System Level Description

Figure 2–1:

FDDI Network Connection Model

Physical
connection
Station A

Station B
SMT

LLC

MAC

PHY

SMT

PMD

PHY

MAC

Bidirectional
medium

LLC

LKG-4807-91I

2.3 Physical Layer Medium Dependent (PMD)
The PMD standard defines the lower sublayer of the OSI Physical layer. It describes
the physical requirements for nodes that attach to the FDDI network and the hardware connection between these nodes. It also provides all services necessary to transmit an NRZI-encoded digital bit stream between stations. The PMD addresses the
following:

•

Transmitter and receiver requirements, including power budget

•

Transmitter and receiver waveform characteristics

•

Conversion of electrical data signals to and from optical signals

•

Jitter limits and permissible bit error rates.

•

Mechanical mating requirements for the Media Interface Connector (MIC),
including the keying methods and connector footprint

•

Fiber optic cable requirements

Concepts

2–3

The PMD standard specifies a nominal optical wavelength of 1300 nm for data transmission over 62.5/125 µm multimode, graded index, fiber optic cable. This permits
the use of lower-cost LEDs in place of the more expensive laser devices required for
single-mode fiber.
2.3.1 Fiber Cable Requirements

Fiber optic cable is available in different sizes and types. The FDDI standard specifies the use of 62.5/125 µm graded index, multimode, fiber optic cable. An appendix
to the PMD standard lists cable alternatives. Table 2–1 lists Digital’s recommended
fiber cable specifications in both the 1300 nm and 850 nm windows. Digital recommends the use of dual-window fiber to support FDDI 1300 nm requirements and existing or future non-FDDI applications requiring a 850 nm window. Table 2–2 lists
alternate multimode fiber cable specifications for FDDI in the 1300 nm window.
Table 2–1:

Multimode Fiber Cable Specifications

Recommended Cable

Parameters

Nominal core diameter

62.5 ± 3.0 µm

Cladding diameter

125.0 ± 2.0 µm

Numerical aperture

0.275 ± 0.015

Modal bandwidth minimum

500 MHz• km @ 1300 nm, 160 MHz• km @ 850 nm

Maximum attenuation rate

1.5 dB/km @ 1300 nm, 3.5 dB/km @ 850 nm

Power budget

11.0 dB @ 1300 nm (for FDDI), varies by product
@ 850 nm

Table 2–2:

Alternate Multimode Fiber Cable Specifications for FDDI

Alternate Cable

100/140

50/125

Nominal core diameter

100 µm* ± 4.0 µm

50 µm* ± 3.0 µm

Cladding diameter

140 µm* ± 6.0 µm

125 µm* ± 2.0 µm

Optical wavelength

1300 nm

Numerical aperture

0.290 ± 0.015

0.200 ± 0.015

Modal bandwidth minimum

500 MHz• km @ 1300 nm

Power budget for FDDI

9.0–13.0 dB

6.0–7.0 dB

Maximum distance

1.6 km (.96 mi)

2 km (1.2 mi)

*Power budget contingent on core and cladding tolerances.

2–4

Fiber Distributed Data Interface System Level Description

The FDDI single-mode PMD standard (currently under development) describes the
traits of single-mode fiber for long-distance links. Table 2–3 lists the single-mode
fiber optic cable specifications.
Table 2–3:

Single-Mode Fiber Cable Specifications

Recommended Cable

Parameters

Mode Field diameter

8.2 µm to 10.5 µm

Cladding diameter

125 µm ± 2 µm

Fiber cladding noncircularity

2% maximum

Core to cladding concentricity error

1 µm maximum

Nominal operating wavelength

1300 nm

Fiber cutoff wavelength

1270 nm maximum

Zero dispersion wavelength

1300 to 1322 nm

Zero dispersion slope

0.095 ps/(nm2km) maximum

Optical power attenuation

≤ 0.40 dB per km @ 1310 nm

Minimum power budget

22 dB @ 1300 nm

Minimum required loss

12 dB @ 1300 nm

Maximum distance

40 km (24.8 miles)

Concepts

2–5

2.3.2 Optical Bypass Relay

The optical bypass relay (option) shown in Figure 2–2 can be used to maintain connectivity of the FDDI ring in the absence of power or during fault conditions in a
node. The bypass relay allows the light to bypass the optical receiver in the faulty
node. In this way, the faulty node is bypassed and the operation of the FDDI ring is
maintained.
Figure 2–2:

Optical Bypass

Secondary ring
Optical
bypass

Primary ring

Dual attachment station

LKG–5092–91I

Optical bypass relays have a power penalty, however, which may cause the maximum allowable loss between stations to be exceeded. This limits the number of serially connected relays in the ring.
Other considerations when using optical bypass relays include:

2–6

•

Bypass relays introduce additional loss in the network, and they do not perform repeater functions of amplifying and restoring the bit stream.

•

By bypassing a station, the new distance between adjacent stations may exceed the maximum allowable value.

•

Bypass relays, as any mechanical devices, may introduce less than reliable
service to the network.

Fiber Distributed Data Interface System Level Description

2.3.3 Connector Keying

The Media Interface Connector (MIC) is the device used in FDDI to connect a node
to multimode fiber. The MIC is designed to align the fiber cable properly with the
transmit and receive optics in the node. The MIC consists of a plug and a receptacle.
The receptacle is keyed to prevent improper plug attachment. Proper keying prevents misconnecting port types, which avoids misconfiguring the network.
Four key types are defined for the MIC receptacle: MIC A and MIC B keying provide
for the attachment of DAS and DAC into the dual ring; MIC M and MIC S keying
provide for attachment of a SAS or a SAC (MIC S) to a concentrator (MIC M).
Figure 2–3 shows the MIC receptacle keying design.
The position of the keyway on the MIC plug determines the type of receptacle to
which it connects. A MIC S plug fits all receptacles. The keyway on the other MIC
plugs and receptacles must match for proper mating.
Figure 2–3:

MIC Receptacle Keying

Connector keying:
MIC A
MIC B
MIC M
MIC S

Primary in/secondary out–DAS A port
Primary out/secondary in–DAS B port
Concentrator M port
SAS S port
LKG–5072–90I

2.4 Physical Layer Protocol (PHY)
The PHY standard defines the media independent portion of the Physical layer in
FDDI. The PHY establishes clock synchronization with the incoming data stream.
The PHY standard defines the following components. The sections that follow discuss these components in greater detail.

Concepts

2–7

•

Symbols—The smallest primitive signaling element used by the MAC

•

Line state—Indicates connection status between neighboring stations as determined by receiving station and based on symbols received from adjacent
neighbor

•

Encode/decode process—4B/5B and NRZ/NRZI data encoding

•

Data recovery—Recovers data bits from the incoming signal

•

Elasticity buffer—Compensates for clock frequency differences between
neighboring stations

•

Smoothing—Compensates for preamble bit loss

•

Repeat filter—Prevents propagation of code violations and certain line states
from inbound to outbound links

•

Scrubbing—Prevents orphan protocol data units (PDUs) from continually
circulating on the ring

To transmit data signals over the fiber, the PMD must first convert the electrical signals from the PHY to an optical data stream. In the optical data stream, a binary 1 is
represented by a polarity transition high to low or low to high, a binary 0 is represented by the absence of a transition. A method must be provided for the receiving
station to synchronize its clock to the transmitting station clock.
To solve this problem, the PHY encodes data in such a way as to guarantee the presence of transitions in the data stream. To achieve efficiency in data transmission,
FDDI combines two types of data encoding: 4B/5B and NRZ/NRZI. This dual coding guarantees at least one transition for each transmitted symbol, ensuring recovery
of the clock by the receiving device. Section 2.4.1 describes the use of symbols in
FDDI; Sections 2.4.3 and 2.4.4 discuss the PHY encoding methods.
2.4.1 Symbols

Although data is transmitted over the fiber in a serial bit stream, the basic unit of information is the symbol. The symbol transmits information between stations on the
FDDI network. The PHY specification defines the symbol set. FDDI defines three
types of symbols: line state, data, and control. Each symbol consists of four bits. In
the PHY, these four bits are passed through a 4B/5B encoder that generates a code
group consisting of five bits.

2–8

Fiber Distributed Data Interface System Level Description

The five-bit code group establishes the signaling rate for FDDI at 125 Megabaud to
meet the 100-Mb/s data rate requirement. Figure 2–4 is a simplified diagram depicting the flow of symbols through the 4B/5B and NRZ/NRZI encoders onto the medium.
Figure 2–4:

4B
Symbols

Symbol Flow Through the PHY

1111

4B/5B
encoder

11101

NRZ/NRZI
encoder

1111 is the symbol for F.
11101 is the code group for F.
The order of bit transmission is leftmost first.
LKG–4239–90i

Table 2–4 lists and defines the symbols used in FDDI. As shown, only 16 of the 32
symbols represent input data; the remaining symbols define control and line state
conditions.

Concepts

2–9

2–10

Table 2–4:

Symbol Encoding

Symbol

Code Group

Assignment

Symbol Type

00000

Quiet

Line state symbol

11111

Idle

Line state symbol

00100

Halt

Line state symbol

11000

First of sequential Starting
Delimiter Pair

10001

Second of sequential Starting
Delimiter Pair

11110

Zero

Data symbol

01001

One

Data symbol

10100

Two

Data symbol

10101

Three

Data symbol

01010

Four

Data symbol

01011

Five

Data symbol

01110

Six

Data symbol

01111

Seven

Data symbol

10010

Eight

Data symbol

10011

Nine

Data symbol

10110

Data symbol

10111

Data symbol

11010

Data symbol

11011

Data symbol

11100

Data symbol

11101

Data symbol

01101

Ending Delimiter

00111

Control indicator/logical zero
(reset)

11001

Control indicator/logical one (set)

V or H

00001

V or H

Invalid code assignment

V or H

00010

V or H

Invalid code assignment

00011

Invalid code assignment

00101

Invalid code assignment

Fiber Distributed Data Interface System Level Description

Table 2–4 (Cont): Symbol Encoding
Symbol

Code Group

Assignment

Symbol Type

00110

Invalid code assignment

V or H

01000

V or H

Invalid code assignment

01100

Invalid code assignment

V or H

10000

V or H

Invalid code assignment

Note: Order of transmission is leftmost bit first (12345).

2.4.2 Line States

FDDI reserves the symbols Q, I, and H, for PHY-to-PHY communications. Called
line state symbols, they can be combined in various ways to determine and control
the status of the physical connection between attached stations.
Line states are determined by the receiving PHY. They indicate a physical link condition that exists longer than can be represented by a single symbol or pair of symbols.
SMT Connection Management (CMT) and Ring Management (RMT) services use
line state symbols to verify and maintain ring integrity. The various line states are as
follows:
Quiet Line State (QLS)

Indicates the absence of a physical connection. A station enters this state upon receipt of at least 16 consecutive Quiet (Q) symbols.

Master Line State (MLS) Part of the physical connection process, a station enters
this state upon receipt of 8 consecutive HQ or QH symbol pairs.
Halt Line State (HLS)

Part of the physical connection process, a station enters
this state upon receipt of at least 16 consecutive H
symbols.

Idle Line State (ILS)

Establishes and maintains clock synchronization on
the outbound physical link. A station enters this state
upon receipt of at least 4 consecutive Idle (I) symbols.

Concepts

2–11

Active Line State (ALS) Indicates that the inbound symbol stream is a MAC
frame sequence and that the neighboring PHY has enabled the associated physical connection. A station enters this state upon receipt of a Starting Delimiter (JK)
symbol pair.
Noise Line State (NLS)

Indicates noise on the inbound physical link and, if
NLS persists, that the associated connection is faulty.
A station enters NLS upon receipt of 16 noise events
without satisfying the criteria for entry into another
line state. Potential noise events include receipt of a Q,
H, J, K, or V symbol.

2.4.3 4B/5B Data Encoding

To convert the 4-bit MAC symbols into 5-bit code groups, the PHY standard specifies the use of 4B/5B encoding. The fifth bit added to the symbol group is for clocking purposes and to provide control symbols that are unique from data. Figure 2–4
shows the symbol flow through the 4B/5B encoding logic. Properties of 4B/5B encoding include the following:

2–12

•

Idle symbols have the maximum clock information for synchronization of
station clocks.

•

Maximum of one leading 0 and two trailing 0’s in a code group to prevent
more than three contiguous 0’s in a data stream during active line state.

•

Starting delimiter, JK, is always recognized regardless of symbol boundary.

•

Line state recognition is independent of symbol boundaries.

Fiber Distributed Data Interface System Level Description

2.4.4 Modified NRZI Encoding

Once a symbol is converted to a 5-bit code group, it is further encoded using a
Non-Return-to-Zero/Non-Return-to-Zero Invert on Ones (NRZ/NRZI) coding
scheme. NRZ/NRZI encoding allows data to be transmitted at 100% efficiency by
means of polarity changes. A change in amplitude or optical power level indicates a
one; no change indicates a zero. Figure 2–5 illustrates this concept.
Figure 2–5:

NRZI Encoding Process

8 ns
8 ns clock pulse
Pre-encoded data
NRZI encoded data
1

1
LKG–4062–90I

2.4.5 Clock Recovery and the Elasticity Buffer

At 100 Mb/s, the use of a centralized clocking scheme would put critical demands on
the internal phase locked loop circuitry of each station. To eliminate the need for expensive circuits, FDDI specifies the use of distributed clocking on the network. Each
station has an autonomous clock source for transmitting or repeating information on
the ring. Each receiver recovers clock and data from its upstream neighbor.
Within the PHY standard, clocking is defined as part of the data receive function. The
clocking circuitry within each station ensures that received data is properly decoded.
A phase locked loop circuit derives a 125 MHz clock based on the incoming data.
The resulting clock is used to sample and restore incoming data to logic levels.

Concepts

2–13

To compensate for any possible clock frequency differences between stations, each
station has an elasticity buffer. Data is clocked into the buffer at the clock rate recovered from the incoming bit stream. It is clocked out of the buffer at the station’s internal clock rate. The MAC entity that originates a frame inserts at least 16 Idle symbols
before the frame. The elasticity buffer in a subsequent station can change Idle symbol count as necessary to maintain clock synchronicity among attached stations.
The PHY standard specifies a frequency tolerance of .005% for a station, which
translates to a .01% frequency variation between the incoming clock and the station
internal clock. The standard further specifies a minimum station elasticity of ± 4.5
code bits. Multiplying the .01% frequency difference times the maximum frame size
of 4500 bytes (45,000 code bits) results in the 4.5 code-bit elasticity buffer size.
2.4.6 Smoothing Function

As a MAC data frame circles the ring, it is possible that repeated elasticity buffer operations could cause sufficient preamble shrinkage to cause a station to lose frames.
To prevent this, a smoother is built into each PHY. This function compensates for the
possibility of preamble shrinkage by absorbing symbols from longer preambles and
distributing them into shorter preambles. The smoother can insert or delete Idle symbols in the preamble as necessary to ensure the proper reception of frames.
2.4.7 Repeat Filter

The repeat filter prevents propagation of code violations and certain line states from
the inbound link to the outbound link while permitting the propagation of valid
frames. It also provides for the propagation of lost (damaged) frames so that they can
be counted by the next MAC entity on the ring. This filter allows a station to pass data
from its upstream neighbor to its downstream neighbor without involving the MAC
entity.
2.4.8 Scrubbing

When a ring is reconfigured, attached devices can enter or leave the ring. During this
process, orphan protocol data units (PDUs) are sometimes created. These orphan
PDUs may not be identified as belonging to any particular attached device because
they are generated by a device no longer part of the original configuration. For example, if a station suddenly fails, the station abruptly leaves the ring without completing
normal exit procedures. The adjacent station cannot identify the PDUs left behind.

2–14

Fiber Distributed Data Interface System Level Description

To remove these orphans, the adjacent station scrubs the ring when a station is inserted or removed. The adjacent station generates a series of Idle symbols while
stripping all frames and tokens that are received. This operation is performed long
enough to guarantee that all frames have been removed from the ring and that the ring
will reinitialize. This prevents orphan frames from continually circulating on the
ring and prevents old frames from being received long after they were originally
sent.

2.5 Media Access Control (MAC)
An FDDI token ring consists of stations connected in series by medium forming a
closed loop. Data is transmitted serially as a symbol stream from one attached station
to its downstream neighbor. Each station in turn regenerates and repeats each symbol, passing it to its downstream neighbor.
MAC is the lower sublayer of the Data Link layer defined by the OSI reference model. The functions of the MAC include the delivery of LLC PDUs, frame construction,
transmission, receiving, repeating, and stripping. The MAC is further responsible
for claiming and beaconing. Error detection mechanisms are also defined in the
MAC standard. Functions defined by the MAC standard include the following:

•

Fair and equal access to the ring through use of the timed token

•

Communication between attached devices using frames and tokens

•

Ring initialization (claiming)

•

Ring fault isolation (beaconing)

To gain the right to transmit on the ring, stations must first acquire the token, which is
a unique symbol sequence that circulates around the ring following a data transmission. Once a station acquires the token, it removes the token from the ring and begins
transmitting data. At the end of transmission, the station issues a new token onto the
ring, providing other stations with the opportunity to transmit.

Concepts

2–15

2.5.1 Ring Scheduling

One of the fundamental concepts of the FDDI network is the use of a timed-token
ring architecture and a Timed-Token Protocol (TTP). To guarantee synchronous service time, the MAC standard specifies that a Target Token Rotation Time (TTRT)
must be observed by all attached stations. TTRT is determined by mutual agreement
among all active stations. The TTP guarantees that the token appears at a station by
two times the TTRT. The MAC standard defines the timed token as the means for
acquiring access to the ring in order to transmit data from one attached device to the
next.
The TTP is a major difference from such commonly used LAN standards as IEEE
802.5. In some other LANs, a station captures the token and flips a bit within the token to indicate the station has captured it. In FDDI, a station captures the token, transmits for its allotted time period, and reissues the token onto the ring once it has
completed transmission or has exhausted its available transmission time.
2.5.2 Data Link Addressing Scheme

Digital’s FDDI devices employ the 48-bit addressing scheme defined by the FDDI
standards. In this scheme, 48 bits specify the destination and source addresses of active nodes. Two of the bits also determine whether the address is an individual or
group address, and whether address administration is local or by means of the IEEE
address administrator. The 48-bit address must be unique for all nodes. No two
nodes in the extended LAN can have identical individual addresses. Figure 2–6
shows the format and defines bits used in the address.
Figure 2–6:

48-Bit Address Format

48-bit address
I/G

U/L

Remaining 46 address bits

I/G = Individual or group bit
U/L = Universal or local bit
LKG–4065–90I

2–16

Fiber Distributed Data Interface System Level Description

In the destination address field, a 0 in the I/G bit indicates that the 48-bit address is an
individual address; a 1 indicates that it is a group address. A 1 in the U/L bit indicates
a locally administered address; a 0 in this bit position indicates a universally administered address. In the source address field when source routing is used, the I/G bit is set
to 1. Otherwise, the I/G bit is set to 0.
Digital recommends using universally administered addressing to eliminate the risk
of duplicate addressing on the extended network and to eliminate the need for the
network manager to assign addresses. Because every address is unique, this technique also eliminates the need to resolve addressing conflicts when LANs are
merged.
2.5.3 Universally Administered Addressing

Universally administered addresses ensure that any address assigned on the extended LAN is unique. This is done by the use of a hierarchical address administration. In the 48-bit addressing of FDDI, the 24 Organizationally Unique Identifier
(OUI) bits are assigned to the owner of the OUI by the IEEE Address Administrator.
The remaining 24 bits are assigned by the owner of the OUI. Controlling address assignments in this manner avoids address duplication on the LAN. Figure 2–7 shows
the universal address format.
Figure 2–7:

Universal Address Format

48-bit address
I/G

U/L

OUI
24 bits long

24 bits assigned by OUI
owner

OUI = Organizationally Unique Identifier
LKG–4240–90i

Although OUI assignments are specified as 24 bits, the owner of the OUI can set the
I/G bit to 0 or 1 as needed to define an individual or group address. The U/L bit is set
to 0 to indicate a universally administered address.

Concepts

2–17

2.5.4 Canonical Addresses and the Order of Transmission

The IEEE canonical form for a 48-bit address represents the address as a 6 octet hexadecimal string, with hyphens separating the octets, as shown in the canonical address
line of Figure 2–8 part A. The first octet contains the first 8 bits of the address, with
the I/G bit as the least significant bit, the U/L bit as the next significant bit, and so on.
(In Figure 2–8 part A, the first octet of the binary representation line is the leftmost
octet.)
The IEEE strictly defines the order of transmission for addresses in the DA and SA
fields. Addresses in the DA and SA fields must be transmitted I/G bit first, no matter
what transmission order is used by the LAN. Transmission order is left to right, as
shown in the transmission order line of Figure 2–8 part A.
Because FDDI transmits symbols MSB first, the correct symbols must be transmitted to guarantee that the DA and SA fields meet the IEEE transmission order requirements. The FDDI symbols line in Figure 2–8 part A shows the FDDI symbols
that must be transmitted to produce the proper bit order. This order of transmission
also applies to addresses carried in SMT frames.
Figure 2–8:

Order of Transmission for Addresses

Address in SA and DA fields:

Canonical address
Binary representation
Transmission order
FDDI symbols

DE–BC–68–A5–34–27
11011110–10111100–01101000–10100101– 00110100– 00100111
01111011 00111101 00010110 10100101 00101100 11100100
7B
3D
16
A5
2C
E4

Address in the INFO field:

Canonical address
DE–BC–68–A5–34–27
802.3/802.4 transmission 01111011 00111101 00010110 10100101 00101100 11100100
FDDI/802.5 transmission 11011110 10111100 01101000 10100101 00110100 00100111
FDDI symbols
DE
BC
68
A5
34
27

LKG–4241–90I

2–18

Fiber Distributed Data Interface System Level Description

The rules for order of transmission of addresses apply only to the DA and SA fields.
Addresses carried as information in data fields should be transmitted in accordance
with the transmission order rules for the specific LAN. For FDDI and 802.5 LANs,
this is MSB first. For 802.3 and 802.4 LANs, it is LSB first.
The example in Figure 2–8 part B shows the transmission order, left to right, for the
different LAN types. In the line labeled 802.3/802.4 transmission, the I/G bit of the
address is the leftmost bit of the leftmost octet.
In the line labeled FDDI/802.5 transmission, the I/G bit is the rightmost bit of the
leftmost octet. The FDDI symbols that represent this transmission order are shown in
the FDDI symbols line. Addresses carried as data in the info field for interchange
between higher layer protocols (for example, ISO 9542 or TCP/IP ARP) should be
transmitted in their canonical (6-octet string) form.

2.6 Protocol Data Units
Protocol Data Units (PDUs) are the units of transfer between peer entities in the layered model. FDDI MAC defines two PDU formats: tokens and frames. Frames are
further divided into MAC frames, Station Management (SMT) frames, Logical Link
Control (LLC) frames, Void frames, and Implementer frames. Table 2–5 lists frame
types and Frame Control values that define them. Sections 2.6.1, 2.6.2, and 2.7.5 describe each type of PDU in greater detail.
Table 2–5:

FDDI Frames

Frame Control
(FC) Bit Field

Frame Type

Description

0x00 0000

Void

Void frame

1000 0000

Token

Nonrestricted token

1100 0000

Token

Restricted token

1L00 0001 to 1111

MAC

MAC frame

1L00 0010

MAC

MAC beacon frame

1L00 0011

MAC

MAC claim frame

0L00 0001 to 1111

SMT

SMT frame

0L00 1111

SMT

Next station addressing
frame

CLFF FFFF

Concepts

2–19

Table 2–5 (Cont.):

FDDI Frames

Frame Control
(FC) Bit Field

Frame Type

Description

CL01 r000 to r111

LLC

LLC frame

0L01 rPPP

LLC

Asynchronous priority transmission, priority determined
by PPP in bit field

1L01 rrrr

LLC

Synchronous transmission

CL10 r000 to r111

Implementer

Implementer frame

CL11 rrrr

Reserved

Reserved for future use by
standard

C = class
L = length
P = priority

r = reserved
x = don’t care

2.6.1 Tokens

The token is the means by which the right to transmit is passed from one attached
node to another. In FDDI, stations acquire and absorb the token when they want to
transmit data. Immediately after data transmission, the station puts a new token onto
the ring. Figure 2–9 illustrates the MAC token structure. Section 2.6.3 defines token
fields.
Figure 2–9:

Token Format

FDDI token — 22 symbols long
PA

PA = Preamble — 16 symbols
SD = Starting delimiter — 2 symbols
FC = Frame control — 2 symbols
ED = Ending delimiter — 2 symbols
LKG–4063–90I

2–20

Fiber Distributed Data Interface System Level Description

2.6.2 Frames

Although a token transfers the right to transmit, frames are the means for carrying
MAC control data or user information among stations. Frames are divided into several fields, each containing information that defines a specific frame parameter.
Figure 2–10 shows the FDDI frame format and component fields. Each frame field
consists of a number of FDDI symbols. Refer to Section 2.4 for information concerning symbols and data encoding; Section 2.6.3 provides frame field definitions.
Figure 2–10: FDDI Frame Format

FDDI frame — 4500 bytes maximum length
PA

PA =
SD =
FC =
DA =
SA =

SD FC DA SA

Preamble
Starting delimiter
Frame control
Destination address
Source address

INFO

FCS ED FS

INFO =
FCS =
ED =
FS =

Information
Frame check sequence
Ending delimiter
Frame status

LKG–4066–90I

2.6.3 Frame and Token Fields

Frame and token fields are defined as follows (refer to Table 2–5 for bit and field positions within the frame or token):
Preamble (PA)

This field is used for Receive clock synchronization
in frames and tokens. The preamble is transmitted
with at least 16 Idle symbols. Frames can be received
that contain more or fewer Idles, depending on station behavior.

Starting Delimiter (SD)

A unique 2-symbol pattern that indicates the start of
the frame or token. In FDDI this pattern is always the
symbols J and K. No PDU is considered valid unless
it begins with this sequence.

Concepts

2–21

Frame Control (FC)

A 2-symbol field that defines frame or token type.
The bits in this field are defined as follows from left
to right:
C = Class bit, indicating class of service: synchronous or asynchronous. In Digital’s implementation,
it is set to 1 indicating asynchronous frames.
L = Address Length bit, indicating either 48- or
16-bit addressing. In Digital’s implementation, it is
set to 1 indicating 48-bit addressing.
FF = Format bits. Used in conjunction with the other
FC bits, they define the frame type being transmitted:
Void, Nonrestricted Token, Restricted Token, SMT,
MAC, LLC, Implementer, and Reserved.
ZZZZ = Control bits used with the CLFF bits to define the frame.

Destination Address (DA) This 48-bit field identifies the station to receive the
frame. The DA can be a single station address, a
broadcast address, or a multicast address.

2–22

Source Address (SA)

This 48-bit field identifies the originator of the
frame.

INFO

This field contains the data to be transmitted over
the ring. It can be either MAC control information or
user data. Field length is variable but is limited by the
maximum FDDI frame size of 4500 bytes.

Frame Check Sequence

This 32-bit field contains a Cyclic Redundancy
Check (CRC) value based on the contents of the FC,
DA, SA, INFO, and FCS fields. The receive station
uses this value to verify the accuracy of the transmitted frame.

Ending Delimiter (ED)

This field indicates the end of a frame. For tokens, the
ED is always two T (Terminate) symbols. For
frames, the ED is always one T symbol.

Fiber Distributed Data Interface System Level Description

Frame Status (FS)

This field consists of the Error Detected (E), Address
Recognized (A), and Frame Copied (C) indicators.
These indicators can contain an Reset (R) or Set (S)
symbol.

2.7 Station Management (SMT)
Station Management (SMT) provides the necessary services at the station or local
level to monitor and control an FDDI node. It allows stations to work cooperatively
over the ring, and ensures proper station operation. SMT provides the following services to an FDDI node:

•

Connection management (including insertion and removal)

•

Configuration management

•

Encoding of SMT frames

•

Fault isolation and recovery

Although FDDI stations can have multiple instances of the MAC, PHY, and PMD,
all FDDI station types have only one SMT entity. SMT contains three major protocols: Connection Management, Ring Management, and SMT Frame Services. The
three SMT protocols are discussed in the following sections. Figure 2–11 shows how
SMT is incorporated into the FDDI architecture.
2.7.1 Connection Management (CMT)

Connection Management (CMT) is that portion of SMT that performs Physical layer
insertion and removal, and controls the configuration of PHYs and MACs within a
node. CMT controls the connection of nodes to the FDDI network, the physical connection of PHYs in separate nodes (including concentrators), and other functions
such as scrubbing and trace. Figure 2–11 shows CMT in relation to other elements of
SMT. The CMT components are as follows:

•

Concepts

Physical Connection Management (PCM)—Provides for management of the
physical connection between adjacent PHYs. This includes connection establishment, link confidence testing, and enforcement of connection rules.

2–23

•

Configuration Management (CFM)—Provides for the configuration of PHY
and MAC entities within a node. There is only one CFM entity in a station;
there is only one CFM entity for each PHY in a concentrator.

•

Entity Coordination Management (ECM)—Provides for control of bypass
relays and signaling to control PCM.

Figure 2–11: Station Management Architecture

FDDI station
To LLC

Ring
Management

MAC
Configuration
switch

Local
management
interface

Connection
Management

PHY

PMD
SMT frame
services
Station
Management
Transmit data path
Receive data path
LKG–4067–90I

2–24

Fiber Distributed Data Interface System Level Description

2.7.2 Connection Rules

In the FDDI topology, there are four physical port (PHY-port) types: A, B, M, and S.
Figure 2–12, a simplified diagram of a concentrator, shows the various PHY-ports
used in FDDI. These PHY ports are defined as follows:

•

PHY A—Connects to the incoming primary ring and the outgoing secondary
ring of the FDDI dual ring. This port is part of a dual attachment station
(DAS) or a dual attachment concentrator (DAC).

•

PHY B—Connects to the outgoing primary ring and the incoming secondary
ring of the FDDI dual ring. This port is part of a DAS or a DAC. This port is
also used to connect the DAS to the concentrator.

•

PHY M—Connects a concentrator to a SAS, DAS, or another concentrator.
This port is implemented only in the concentrator.

•

PHY S—Connects a SAS to a concentrator PHY-M port. PHY S also connects to a DAS or another SAS, but these configurations are not always recommended.

Figure 2–12: PHY- Port Types

Secondary ring
PHY A

PHY B

Primary ring

Concentrator

PHY M

• ••

PHY M

•••

PHY M

PHY B

PHY S

DAS or DAC

SAS
LKG–4809–91I

Concepts

2–25

In the ANSI FDDI standards, connection rules are specified to ensure against the
construction of illegal topologies. Connections between stations have the status of in
use or not in use. When a cable is connected a decision is made to use or not use that
cable in the token path. The ring connection rules for Digital devices are listed in
Figure 2–13. In the figure, This Node is the reference point for all connection decisions.
Figure 2–13: FDDI Connection Rules Matrix

This Node

Other
Node

PHY port

Rule 1

Yes

Rule 2

Yes

Rule 3

Yes
Note 1

Yes

Yes = Accepts this connection
No = Does not accept this connection
LKG-4810-91I

Rule 1:

Rule 2:

PHY B of This Node accepts connection to PHY A of Other Node if:

•

PHY A of This Node is not connected, or

•

PHY A of This Node is connected to PHY B or PHY S of another node.

PHY A of This Node accepts connection to PHY B of Other Node if:

•

PHY B of This Node is not connected, or

•

PHY B of This Node is connected to PHY A or PHY S of another node.
See also Note 2.

2–26

Fiber Distributed Data Interface System Level Description

Rule 3:

PHY A of This Node accepts the connection to PHY M of Other Node if:

•

PHY B of This Node is not connected, or

•

PHY B of This Node is connected to PHY A or PHY S of another node.
See also Note 2.

Note 1:

Backup tree connection (dual homing)
The connection of PHY B of This Node to PHY M of Other Node always
takes precedence over the connection of PHY A of This Node to PHY M
of another node.
When a connection between PHY A of This Node and a PHY M or PHY B
of another node exists, it is disabled when PHY B of This Node connects to
PHY M of Other Node.
If the connection breaks between PHY B of This Node and PHY M of Other
Node, the connection is reestablished between PHY A of This Node and
PHY M of another node.

Note 2:

Tree connections are preferred over dual ring connections
Tree connections are preferred over dual ring connections when only one of
two connections may be accepted.
The connection of PHY A or B of This Node to PHY M of Other Node always takes precedence over the connection of the remaining PHY (A or B)
of This Node to PHY A or PHY B of another node.
The connection is disabled between the remaining PHY (A or B) of This
Node and PHY A or PHY B of another node when PHY A or PHY B of This
Node connects to PHY M of Other Node.

2.7.3 Ring Management (RMT)

The Ring Management (RMT) portion of SMT receives status information from the
MAC and CMT configuration logic. RMT then reports this status to the SMT and
higher level processes. Services provided by RMT include stuck beacon detection
and resolution by means of Tracing, MAC availability for transmission, and duplicate address detection and resolution.

Concepts

2–27

In the case of a stuck beacon (a station is locked into continuous beacon frame transmission), a stuck-beacon timer in RMT measures the length of beacon transmission.
If beaconing exceeds the timer limit, RMT begins stuck-beacon recovery procedures
to restore the station to normal operation, if possible.
This process begins with the transmission of a directed beacon, which informs the
ring of the stuck condition. If the ring has not recovered by the end of this transmission, then RMT initiates the Trace function. Trace provides a recovery mechanism,
using Physical layer signaling, for stuck beacon conditions on the FDDI ring when a
stuck beacon cannot be localized to a single link. Refer to Section 2.10.2 for a detailed discussion of the Trace function.
Duplicate addresses on the FDDI ring can prevent the ring from becoming operational. If RMT logic detects that its MAC has a duplicate address, RMT takes action to
allow the ring to resume normal operation.
2.7.4 SMT Frame Services

SMT frame services provide control and observation mechanisms for the FDDI network. These services are implemented by means of different SMT frame classes.
Within each frame class, different frame types are also defined. Frame class identifies the function the frame is performing. Type indicates whether the frame is an announcement, request, or response to a request. The Header field of an SMT frame
contains class and type definitions for the SMT frame. These classes include, but are
not limited to, the following:

2–28

•

Neighborhood Information Frames (NIF) are used by a station to announce
its address and a description of the station. Stations also use these frames to
request this information of a neighbor station.

•

Station Information Frames (SIF) are used to request, and provide in response to another request, a station’s configuration and operating information. In addition, SIF frames contain information about the attachment status
of each port. This information can be used to create a physical ring map that
show the token path and the topology.

•

Request Denied Frames (RDF) are generated in response to an unsupported
frame request.

Fiber Distributed Data Interface System Level Description

2.7.5 SMT Frames

The SMT frame consists of three fields as shown in Figure 2–14. The MAC header
portion identifies the frame as an SMT frame and contains the source and destination
addresses. The SMT header identifies the SMT class and type of the frame. The SMT
INFO field contains the SMT parameters to be transmitted.
Figure 2–14: SMT Frame Format

MAC header

SMT header

SMT INFO

13 octets

20 octets

0 to n octets
LKG–4242–90i

2.8 Logical Link Control (LLC)
Although not part of the standard, FDDI requires LLC for proper ring operation. It
resides in the second, or Data Link, layer of the OSI model. As in most Data Link
standards, LLC defines the transmission of a frame of data between two stations with
no intermediate switching nodes. Figure 2–15 shows the relationship between LLC
and FDDI. FDDI assumes implementation of the IEEE 802.2 LLC standard.
Figure 2–15: Logical Link Control and FDDI

Station Management (SMT)

Logical
Link
Control

MAC

Configuration switch

Rx
PHY

PMD
Tx
LKG–4068–90I

Concepts

2–29

Data is transmitted between LLCs by means of LLC PDUs, which are frames with
Frame Control fields designating the INFO field as containing LLC information. As
part of the LLC PDU, a service access point address is included indicating that the
receiver is a particular LLC client. LLC frames can cross bridges into the extended
LAN; MAC and SMT frames cannot do this.

2.9 Ring Operation
This section describes basic FDDI ring operation. It provides an overview of ring
functionality using the concepts described in the preceding sections. It also discusses
timing, steady state operation, ring initialization, and the claiming process. Section
2.10 describes the beacon process and the trace function. Because of the complexity
of FDDI, it is beyond the scope of this document to describe every detail of ring operation.
2.9.1 Asynchronous Ring Transmission

In Digital’s FDDI implementation, all attached stations use asynchronous data transmission to pass data around the ring. In asynchronous data transmission, all attached
stations are dynamically allocated a transmission time based on the TTRT (discussed
previously). Stations acquire the token and transmit until the Token Holding Time
(THT) expires.
Asynchronous transmission, when used in the context of the FDDI environment,
pertains to ring traffic and services that are not extremely delay sensitive. Do not confuse this usage with the asynchronous transmission of data on the physical media,
which refers to the intercharacter timing of a transmission as used on low speed lines
for devices such as ASCII terminals. Digital’s FDDI devices receive synchronous
and asynchronous transmissions, but they transmit asynchronous only.
2.9.2 Timers

Each station on the FDDI ring uses three timers to regulate operation of the FDDI
ring: the Token Rotation Timer (TRT), the Token Holding Timer (THT), and the Valid Transmission Timer (TVX). These timers are administered locally by the individual stations.

2–30

Fiber Distributed Data Interface System Level Description

The TRT, which times the period between the receipt of tokens, is critical to the
successful operation of the FDDI network. It is initialized to different values depending on the status of the ring. This timer controls ring scheduling during normal operation and it is also used for timing error recovery in the node. The TRT is set during
normal operation to the value of the Target Token Rotation Time (TTRT) negotiated
by stations attached to the FDDI ring. Section 2.9.3 discusses this in more detail.
The THT controls the length of time during which a station can start transmission of
asynchronous frames. Attached stations can begin asynchronous transmissions if
THT has not expired. THT is initialized with the amount of time remaining until
TTRT expires when the token is captured. The station can then initiate asynchronous
transmission until THT expires.
The TVX times the period between valid transmissions on the ring and allows the
station to recover from transient ring errors. The TVX is used in the detection of excessive ring noise, token loss, and other faults. It is reset on receipt of a valid frame or
token.
2.9.3 Target Token Rotation Time, Claiming, and Initialization

This section describes normal station power up and the physical connection process,
which is controlled by CMT using PHY signaling. At power up, stations recognize
their neighbors on the fiber by transmission of a defined line state sequence (QLS–
HLS–ILS). Once stations acknowledge each other, they exchange information on
port type and connection rules. They negotiate link confidence test (LCT) duration,
run LCT, and exchange results. If LCT passes, they complete the physical connection by transmitting another defined line state sequence (HLS–MLS–ILS).
Once the physical connection is complete, FDDI requires the establishment of a
TTRT to ensure that stations receive guaranteed service time. The TTRT default value in the Digital environment is 8 ms; this value is not normally changed. If necessary, TTRT can be set to a different value based on the number of attached stations,
the length of the ring, and the time required by each station to transmit data over the
ring. TTRT is negotiated during ring initialization by means of a claim token bidding
process (described in the following paragraphs). Figure 2–16 shows how the claiming process operates.

Concepts

2–31

Figure 2–16: Claiming Process

FDDI
station 2

FDDI
station 1

Claiming process
winner

FDDI
station 3

FDDI
station 4

1. All active stations start issuing claim frames.
2. Station 2 receives a shorter claim and yields.
3. Station 3 receives a longer claim and continues
sending claim frames.
4. Stations 1 and 4 receive a shorter claim and yield.
5. Station 3 receives its own claim frame signaling
that it has won the bidding process.
6. Station 3 issues the token.

LKG–4064–90I

The claiming process determines which station will initialize the ring. The process
begins when one or more MACs enter the claim state. In this state, the MAC continually transmits claim frames. The claim frame contains the originator’s address and
the station’s bid (called T_Bid) for TTRT.
Receiving stations take incoming claim frames and compare the received TTRT with
their bid for TTRT. If the received T_Bid represents a shorter (in time) TTRT than the
station’s own bid value, then the station repeats the claim frame and stops transmitting its own claim. If the received bid value represents a longer (in time) TTRT, then
the station strips the claim frame and continues transmitting frames with its own
T_Bid onto the ring. The claim winner is determined when a station receives its own
claim frame. In the event that two or more stations make identical bids, then the station with the longest and highest address wins the bidding.

2–32

Fiber Distributed Data Interface System Level Description

Eventually, the ring is filled with claim frames from the winning station, which initializes the ring by issuing a nonrestricted token. This token passes around the ring
without being captured by any station. Each station in turn, upon receipt of the initial
token, resets its TTRT to the negotiated value (T_Neg) and sets its Ring Operational
bit. Synchronous transmissions are allowed on the second token rotation. On the
third token rotation, nodes can begin asynchronous data transmission. At this point,
the ring is in steady state operation.
Because FDDI uses nonprioritized, asynchronous transmission with a nonrestricted
token, bandwidth is time-sliced among all active stations on the ring. This mode supports fair access to the ring at a frame-level granularity.
2.9.4 Steady State Operation

In steady state operation, the ring has been initialized and all attached stations are
fully functional. A station that is ready to transmit waits for a token. The station captures the token from the ring, which allows it to transmit. The station continues initiating transmissions until either it has no more data to transmit or the THT expires.
Capturing the token in an FDDI ring differs from the process used in an 802.5 ring in
which a station ready to transmit captures the token and flips a bit within the token. A
station ready to transmit data in FDDI must wait until it detects a token. The station
then captures the token, stops the repeat process, and begins transmitting frames. After the station captures the token, there is no token on the ring, effectively prohibiting
other stations from transmitting data onto the ring.
As each station receives a frame, it checks for its own address and for frame errors.
Stations use the Frame Status field, which consists of a number of symbols called
Control Indicators, to signal frame condition. If the frame is destined for the station
that has received it, the station sets the A Control Indicator in the Frame Status field.
It then copies the frame and sets the C indicator, if the copy is successful. Repeating
stations simply retransmit the data frames onto the ring to their downstream neighbor
while checking for errors. Figure 2–17 shows the process of capturing the token and
transmitting a frame on an FDDI ring.

Concepts

2–33

Figure 2–17: Capturing the Token

T
A

C
F
1

1. A has a frame of data to
transmit and captures the
token.

2. A begins transmitting the frame
with Destination Address = C
and Source Address = A .

F
1
C

F
1

F
2
B

B
3. At completion of transmission,
A reissues the token. C copies
the frame addressed to it.

F
1

4. B captures the token and transmits
F2 with DA = D and SA = B.

F
2

T
C

T
B

C
F
2
B

5. B reissues the token. D copies F2.
A, recognizing its SA, strips F1
from the ring.

6. B sees its SA and strips F2.

LKG–4070–90I

2–34

Fiber Distributed Data Interface System Level Description

The receiving station then inserts a symbol in one or more of the frame status indicators depending on the status of the incoming frame.

•

An S symbol in the E indicator signals that an error has been found in the
frame.

•

An S symbol in the A indicator signals that the receiving station recognized
its own address.

•

An S symbol in the C indicator signals that the receiving station has copied
the frame.

Depending upon actions taken by layers above the Data Link layer, the MAC can
make use of, or ignore, the Frame Status Control indicators. If used, the combination
of the A indicator set and the C indicator reset can be interpreted as notification that
the destination station did not copy the frame. Since bridges do not set the A indicator
when forwarding frames, and since the addressed station may exist elsewhere in the
extended LAN, it is not correct to interpret the A indicator reset as notification that a
station does not exist. In most applications, the A and C indicators are ignored.
All stations, except the transmitter, receive and then repeat each frame from the
transmitting station. This results in a delay being introduced on the ring by each repeating station. The total delay introduced by the transmitting station and all repeating stations, plus the cable propagation delay, is the ring latency. As soon as a station
finishes transmitting, it places a new token on the ring. If the ring latency is greater
than the length of a station transmission, it is possible for a token to be issued before a
frame is stripped from the ring.
Frames placed onto the ring by the transmitting station, after circulating around the
ring, are stripped by the transmitting station. Stripping causes partial frames, called
fragments or remnants, to be left on the ring after a station transmits. Logic circuitry
in the FDDI stations ensures that frame remnants do not degrade ring operation.
Frame fragments are removed from the ring when they encounter a transmitting station.

Concepts

2–35

2.10 Ring Maintenance
The responsibility for monitoring ring integrity is distributed among all attached stations. Each station monitors the ring for conditions that require ring initialization.
These conditions include ring inactivity longer than TVX, invalid frames on the ring,
or a physical or logical break in the ring. To detect a lack of ring activity, stations use
the internal timer TVX to track the length of time between reception of valid frames
or tokens. If this time exceeds the value in TVX, an error condition is indicated and
stations enter the claim process. If the claim process fails, the station enters the beacon process.
Stations leave the steady state whenever a ring or station malfunction occurs.
Figure 2–18 shows the events that occur when there is a break in the fiber of an FDDI
ring. In Figure 2–18, the token path on the primary ring is from station A to station B,
and so forth. If a break in the primary ring occurs between station A and station B,
PCM at PHY A of station B sees the loss of optical signal on the input fiber. Station B
then shuts down the connection, which turns off the optical transmitter on the secondary ring (that transmits to PHY B of station A). Station B then wraps.
PHY B of station A detects loss of signal on the receiver, so it shuts down the connection and wraps. Stations A and B then attempt to restart the physical connection, but
the PCM protocol detects the break and does not complete the initialization sequence. The ring remains operational because stations A and B are wrapped until the
fault is corrected or until another fault partitions the ring.

2–36

Fiber Distributed Data Interface System Level Description

Figure 2–18: FDDI Typical Ring Recovery

Station D

Primary ring

A
B

Secondary ring

Station C
B
A

Wrapped stations

A
B
Station A

Break in the
fiber

B
A
Station B

1. A break in the primary ring causes port A of station B to detect signal
loss and wrap.
2. Port B of station A also detects signal loss and wraps.
3. Stations A and B attempt to restart the physical connection. SMT PCM
detects the break and does not complete initialization.
Note: All devices on the ring are dual attachment stations or dual attachment
concentrators.
LKG–4071–90I

2.10.1 Beaconing

If a ring malfunction causes a claim to fail, or if SMT requests it, an FDDI station can
start the beacon process. In this state, a station begins continuous transmission of a
beacon frame (refer to Table 2–5 for frame structure). The MAC uses this frame to
announce that the ring is broken and to locate and isolate the fault, if possible.
If a station receives a beacon from its upstream neighbor, it stops beaconing and repeats the upstream beacon. If no beacon is received, the station transmits a continuous beacon stream. If a station receives its own beacon, it assumes the logical ring is
restored, ceases beacon transmission, and begins the claim process.

Concepts

2–37

2.10.2 Trace

The trace function uses PHY signaling to recover from a stuck beacon condition. In
the event of a stuck beacon condition, the fault is localized to the beaconing MAC
and its nearest upstream neighbor MAC. The trace function causes all stations and
concentrators in the suspect section of the ring to leave the ring and run Pathtest in an
attempt to localize the fault.
Stations run Pathtest to check the components of the failed path. This test checks the
transmit and receive components of the path, including the optics and switching logic. If Pathtest finds a fault, the station remains off the ring.
Figure 2–19 illustrates the trace function. The transmitter in station A is defective
and begins issuing a continuous stream of Idle symbols onto the ring. The TVX in
station B times out. Station B attempts to recover by claiming—and failing that—by
beaconing. Finally, station B determines that it is stuck beaconing and begins a trace.
The trace signal (MLS) from station B propagates from B to A over the secondary
ring. Stations A and B leave the ring and begin running Pathtest. Running Pathtest
should allow station A to detect its transmitter fault. If it does not, the fault is detected
when stations A and B attempt to restart the connection. PCM will not complete the
initialization and the connection remains down (as in the example of Figure 2–18).
Although this example deals with stations on the dual ring, trace operates the same
way in a tree topology. The trace always occurs between two MACs on the ring, with
the trace being propagated upstream over the transmit fiber of the ring from the beaconing MAC back to the defective station.

2–38

Fiber Distributed Data Interface System Level Description

Figure 2–19: Detection of Defective Station

Station D

Station C

Primary ring

Secondary ring

1L000 0010

Beacon frame

Master line state
(MLS)

00100 00000 00100 00000 00100 00000
. . . . I I I I I I I I I I I I I I
Station A

Station B

1. Defective station A transmits continuous idle symbols.
2. Downstream station B fails the claim process and begins issuing beacon
frames.
3. Station B detects a stuck beacon and initiates the trace function by sending
MLS to its upstream station.
4. Stations A and B leave the ring and begin Pathtest.
5. If Pathtest detects a fault in station A, station B rejoins the ring and A
remains down.
6. If Pathtest fails to detect the fault, stations attempt to rejoin the ring. At this
time, PCM prevents defective station A from reconnecting to the ring.
LKG-4243-91I

Concepts

2–39

2.11 Digital Enhancements
Digital has developed two algorithms for FDDI: the Frame Content Independent
Stripping algorithm and the Ring Purger algorithm.
2.11.1 Frame Content Independent Stripping Algorithm

On FDDI a MAC entity (also called a MAC) is required to remove (strip) its own
transmitted frames after exactly one traversal of the frames around the ring. The
ANSI FDDI MAC Standard uses a Source Address Match algorithm for frame stripping, where a received frame or fragment is stripped if the Source Address field
matches the MAC’s address.
Digital developed the Frame Content Independent Stripping (FCIS) algorithm for
use by a station to remove or strip frames that it has transmitted onto the ring. The
algorithm is also called the Bridge Stripping algorithm.
The algorithm allows frames to be stripped independently of the content of the
frame. This is extremely important for implementations, like bridges, that have to
transmit frames with source addresses of other stations on different LANs. The algorithm allows implementations like bridges to strip frames without having to do any
address comparisons in real time. The FCIS algorithm interoperates with and is compatible with implementations of the ANSI FDDI MAC standard.
The FCIS algorithm is based on stripping the same number of frames that the station
has transmitted since the last token capture. It does so by keeping a local count of
frames transmitted but not yet stripped. It strips frames until the count is zero.
Each time a frame is transmitted the algorithm increments the count. Each time an
error-free frame is received and stripped, the algorithm decrements the count. Also,
if the count is greater than zero, the algorithm strips frames even with a Frame Check
Sequence error without decrementing the count.
In addition to the count mechanism, the algorithm requires the transmission of a well
formed Void frame (Void with FCS), which is transmitted at the end of transmitting
the frames and prior to transmitting the token. When the well formed Void frame is
received, and if it is error free, the local count is set to zero unconditionally. On receipt of a token, claim, or beacon frame the count is set to zero. When the count is
zero, the MAC strips frames based on the Source Address Match algorithm.

2–40

Fiber Distributed Data Interface System Level Description

The ANSI FDDI MAC standard defines a well formed Void frame as a special frame
that causes the Valid Transmission Timer (TVX) in all stations to be reset. The standard further stipulates that Void frames are not to be copied by any MAC. All well
formed Void frames used by FCIS contain the transmitter’s address as the source address so that they can be properly stripped by the transmitter. The well formed Void
frame is 17 octets long, which is the minimum size frame allowed by the standard.
Digital presented the FCIS algorithm to the ANSI FDDI Committee in November
1988 and in June 1989, where it was highly regarded. The presentation showed the
committee that the algorithm has all the right properties and is robust. Digital also
published and presented papers on this algorithm at SIGCOMM 90 and at the Local
Computer Network Conference in September and October 1990.
2.11.2 Ring Purger Algorithm

On all types of token rings, a frame can circulate indefinitely if it is not removed. A
frame that is not removed (stripped) after its first traversal of the ring is called a No
Owner frame (NOF). While the occurrence of an NOF may be rare, the consequences of allowing the circulation of NOFs are severe. Station resources are consumed by receiving these No Owner frames, possibly several times. For example, a
single NOF on FDDI may cause a station to receive duplicate copies of a frame at the
rate of up to 20,000 frames per second and may consume almost 80 Mb/s bandwidth
on the ring and the receiving station.
The ANSI FDDI standard does not specify a guaranteed method of removing NOFs.
It should be noted that other token rings, such as the IEEE 802.5 token ring, have
mechanisms for removing NOFs. In the IEEE 802.5 token ring, one of the tasks performed by the monitor station is to remove NOFs.
Digital has invented a simple, yet robust, algorithm (called the Ring Purger algorithm) that corrects problems posed by NOFs by removing them reliably from the
ring. The mechanism used by a Ring Purger is extremely simple, requiring the transmission of Void frames, which are not received by the stations on the ring. The ring
bandwidth consumed is small and is an insignificant proportion of the total ring
bandwidth.

Concepts

2–41

The Ring Purger algorithm consists of two parts: an election algorithm to designate,
in a distributed manner, a station as the purger; and the purging algorithm itself. Any
one of the Digital stations in the ring can be elected to act as the ring purger. A unique
station capable of being the ring purger is not needed. All Digital stations participate
in this set of algorithms.
The primary purpose of the Ring Purger election algorithm is to ensure that there is
one, and only one, ring purger operating in the ring most of the time.
The Ring Purger algorithm first elects one Digital station as the ring purger. The designated ring purger station then removes (purges) NOFs and fragments from the ring
transparently. Although the ring purger is not part of ANSI FDDI standard, it interoperates with and is compatible with implementations compliant with the ANSI
FDDI standard.
The Ring Purger algorithm recovers from ring initializations and from failure of a
station acting as the ring purger. Each time the ring is initialized, a Digital station becomes the ring purger if it is the FDDI MAC Claim token winner or if it was the ring
purger prior to the ring initialization.
A ring purger announces its presence periodically by transmitting a Purger Hello
frame. If there are multiple stations capable of acting as ring purgers, the station with
the highest address is elected the ring purger. If there is no ring purger after ring initialization or any time during ring operation, a new ring purger is elected using a distributed election protocol.
The ring purger election protocol frames (for example, Purger Hello frames) use the
SMT Extended Service frames. The election algorithm is designed to elect one, and
only one, station to be the ring purger. The overhead of the purger election algorithm
is negligible because the ring purger only periodically (approximately every 10 seconds) sends a short SMT Extended Service frame to the other participating stations
on the ring.

2–42

Fiber Distributed Data Interface System Level Description

The removal of NOFs from the ring is achieved by the purging algorithm, which operates as follows. The ring purger waits for a token. Once it captures a token, if it has
frames to transmit, it transmits the frames. Once the transmission is completed, the
ring purger transmits two well formed Void frames followed by the token. These
Void frames are extremely small (17 bytes long) and consume little, if any, of the
ring’s usable bandwidth. The Void frames demarcate the end of transmission of the
ring purger’s frames. The Void frames are transmitted with the destination and
source address equal to the My Long Address (MLA) of the ring purger.
When the ring is operating normally, the ring purger receives only the frames it sent
followed by the Void frames. The ring purger strips the ring of all frames and fragments until it receives one of its Void frames. It then strips the two Void frames it
transmitted and ceases stripping frames from the ring. If there are NOFs circulating
on the ring, the ring purger also strips these NOFs while stripping all the frames it
receives before receiving its Void frames. The ring purger does not begin another
purging operation until it receives the token. The ring purger ceases to operate when
it sees a Beacon or Claim frame.
An active ring purger ensures that no frame on the ring circulates as an NOF for more
than one traversal of the ring. These frames will therefore be received at most twice
by the destination stations. The presence of the ring purger strictly limits the degradation in performance of the stations on the ring and the bridged-LAN due to NOFs.
The impact of the ring purger on the performance of the ring and the stations on the
ring is negligible. Stations compliant to the ANSI FDDI MAC standard experience
no overhead because the Void frames have no effect. The ANSI FDDI standard specifies that stations are not to copy Void frames. This prevents station resources from
being consumed by the Void frames.
The ring bandwidth used by the ring purger is also negligible, because the two Void
frames are transmitted only when a token is received. Also, the transmission of two
Void frames takes only 4.48 microseconds. The Ring Purger algorithm is designed so
that the bandwidth consumed becomes smaller as the load on the ring increases.
When the stations on the ring use more of the available bandwidth, the overhead consumed by the ring purger decreases.

Concepts

2–43

The ring purger transmits two Void frames each time it sees the token. On an idle ring,
the ring purger transmits Void frames at a high rate. However, under this idle condition, the Void frames are only using bandwidth on the ring that is being wasted as a
result of no stations transmitting on the ring. When the ring is idle, the consumption
of ring bandwidth by the Void frames is of no consequence to the stations, which also
do not receive these frames. As soon as the load on the ring is increased, the rate of
Void frames decreases, until it finally reduces down to two Void frames every TTRT
period.
Under heavy load, the ring purger transmits the Void frames less frequently. For example, under a very heavy load on the ring, it can take up to a TTRT (which can range
from 4 to 173 milliseconds) for the token to make one traversal around the ring. Thus,
the load posed by the ring purger’s Void frames backs off when there is a load on the
ring. In the worst case, on a heavily loaded ring, the ring purger’s effect on the usable
bandwidth is less than 0.22% (approximately).
The ring purger algorithm cleans NOFs from the ring transparently, and therefore,
helps to improve the robustness and reliability of the FDDI token ring operation.

2–44

Fiber Distributed Data Interface System Level Description

3
Concentrators

The concentrator plays a central role in FDDI topologies. It provides a connection
point for end stations or other concentrators in a physical tree. It supports building
and campus wiring topologies through the creation of hierarchical trees. It also supports the dual ring design specified in the FDDI standards. This chapter provides a
general explanation of the concentrator and its role in FDDI applications. Section 3.5
provides configuration guidelines for using the concentrator in an FDDI network.
Digital’s concentrator, the DECconcentrator 500, provides the flexibility to accommodate the diverse building wiring schemes found at customer sites. It is the key
component in the Digital dual ring of trees topology. It also supports small workgroup topologies when used in a standalone configuration. The concentrator enhances reliability and maintainability in both large and small networks.

3.1 Description
The concentrator is an optoelectronic device that connects stations or other concentrators to the FDDI network. It maps the logical token ring into a hierarchical physical tree. The concentrator is an active device that decodes, retimes, and modifies, if
necessary, the data stream as it passes through the concentrator. By cascading concentrators, network builders can create dual ring of trees topologies. In a standalone
topology, it is the hub in a workgroup topology.
The main advantage of the concentrator is that it removes control of ring topology
from individual stations and places it in the concentrator itself. This feature increases
the reliability of the network for all users. Network managers can use the concentrator to alter the network topology by selectively removing or inserting attached nodes.

3–1

The FDDI standard defines two concentrator types: a single attachment concentrator
(SAC) and a dual attachment concentrator (DAC).
The DAC has all the necessary components for attaching to the dual ring or into a tree
(acting as a SAC). The SAC cannot connect directly to the dual ring because it lacks
the necessary hardware and logic. Figure 3–1 and Figure 3–2 show two concentrator
topologies, the standalone and the dual ring of trees topology.
Figure 3–1:

Concentrator in a Standalone Topology

Concentrator

SAS

SAS
LKG–4050–90I

3–2

Fiber Distributed Data Interface System Level Description

Figure 3–2:

Concentrators in a Dual Ring of Trees Topology

Concentrator

SAS
Concentrator

SAS

Concentrator

SAS

SAS
LKG–4244–90I

Figure 3–3 illustrates the basic architecture of a concentrator. The optional MAC entity provides MAC services to the concentrator. If the MAC entity is implemented, it
is placed just in front of the outbound port. This helps to ensure that SMT station
mapping can be performed accurately. The configuration switch provides internal
configuration capabilities to the concentrator, allowing it to disconnect any or all attached nodes from the ring. With the configuration switch, the concentrator can route
the data path as required according to the number of attached devices. This switch
also disconnects nonfunctional stations attached to the concentrator.

Concentrators

3–3

As shown in Figure 3–3, the concentrator contains a number of PHY and PMD entities. Each PHY/PMD combination makes up an FDDI PHY port. These PHY ports
are full duplex with a transmit and receive function.
Figure 3–3:

Concentrator Architecture

Station Managment (SMT)

MAC

(Optional)

Configuration
switch

PHY

•
•
•
•

PHY

PMD

Port

PMD

•
•
•
•

LKG–4073–90I

Concentrator PHY ports can be type A, B, M, or S. The DECconcentrator 500 has
PHY ports type A, B, and M only. Type A and B ports connect the concentrator to the
primary and secondary rings of the dual ring. Type B ports can also connect one concentrator to the M port of another concentrator. Type M ports can connect stations or
other concentrators to the FDDI network. Refer to Figure 2–13 for device connection rules.

3–4

Fiber Distributed Data Interface System Level Description

Concentrators with type S ports are part of a SAC. They connect the SAC to the
FDDI network through the M port of another concentrator.

3.2 Concentrator Functions
The concentrator performs three reconfiguration functions: station insertion, station
bypassing, and scrubbing. Station insertion and station bypass are performed by the
configuration switch, based on inputs from SMT. Ports can be switched into and out
of the ring either by control of SMT Connection Management or with a management
command. Refer to Section 2.4.8 for a description of the scrubbing function.
Depending on the requirements of a network or the desired application, the concentrator can be equipped with a MAC or it can be MAC-less. Section 3.2.1 describes the
two available concentrator types and their differences. The remaining sections describe MAC functions in further detail.
3.2.1 MAC and MAC-less Concentrators

The ANSI standards specify that concentrators can be constructed with or without
the MAC. The addition of a MAC allows the concentrator to be managed by a remote
management station using the same FDDI (in-band remote management).
Concentrators normally operate without need of manual intervention and provide all
basic services independent of MAC capability. A concentrator can be managed
through a local management interface. A network manager can observe and control
the concentrator directly, or remotely through a communications channel. Remote
management may be provided through the same FDDI (referred to as in-band remote
management) or through a separate channel (referred to as out-of-band remote management).
Management adds the capabilities of remote monitoring and network configuration.
In large networks, this functionality is usually required to ensure proper maintenance
and control of the network.

Concentrators

3–5

Concentrators automatically configure the ring and connect stations to the ring
through the services of SMT Configuration Management (CMT) and Ring Management (RMT). MAC-less concentrators are most useful in cases either where out-ofband remote management is needed for reliability or where local management is
sufficient, such as in standalone topologies serving a limited geographical area. In
these cases, in-band management is an unnecessary expense.
3.2.2 Station Bypass

Station bypass allows the concentrator to logically and physically disconnect the attached station from the ring. The concentrator does this according to a specific set of
conditions as set forth by SMT Physical Connection Management (PCM). The occurrence of any one of these conditions results in the station being removed from the
ring. The concentrator will disconnect the station if one of the following events occurs:

•

The attached station is powered off or the media is unplugged.

•

The concentrator PCM detects a defective connection.

•

The concentrator PCM is explicitly requested to reconfigure by a network
management entity, removing an attached station from the network.

•

The concentrator PCM detects a high error rate on the connection.

PCM, a service of the SMT entity, is part of the Connection Management (CMT) portion of SMT. It initializes the connection of neighboring PHYs. PCM also provides
support for connection maintenance and detection of marginal connection conditions.
3.2.3 Network Management

Independent of network management, the concentrator has the capability to remove
and insert any connection. For example, the concentrator can disconnect itself, a
SAS, or a DAS based on the results of the Link Error Monitor (LEM), which is a
function of CMT. The LEM monitors the connection’s bit error rate (BER). If a station’s LEM determines that the BER is too high, the concentrator removes the connection from the ring.

3–6

Fiber Distributed Data Interface System Level Description

From a network perspective, network management software allows the manager to
configure a concentrator to suit the needs of the network. Under control of the network manager, the concentrator can remove or insert stations as necessary. The network manager can also access the management information maintained by the
concentrator. The resultant information can be used for error checking.

3.3 Concentrator Operation
Frames pass through the concentrator by way of the PHY A, B, and M ports. If the
concentrator is attached directly to the dual ring, frames enter by way of the PHY A
port (if the primary ring is active), flow through the PHY M ports to attached nodes,
and exit by way of the PHY B port. If the concentrator is part of a tree and not connected to the backbone, frames from the higher level concentrator enter through the
PHY B port, flow through the PHY M ports to attached nodes, and back through the
PHY B port to the higher level concentrator.
If the concentrator is dual homed and the backup link is active, frames from the higher level concentrator enter through the PHY A port, flow through the PHY M ports to
attached nodes, and exit by way of the PHY A port.

3.4 Concentrator Planning Considerations
When planning for the installation of concentrators, such as the DECconcentrator
500, in an FDDI network, consult the following Digital manuals:

•

DECconnect System Fiber Optic Planning and Configuration

•

DECconnect System Fiber Optic Installation

3.5 Configuration Guidelines
This section provides an overview of rules and requirements for the configuring of a
concentrator into an FDDI network. In most cases, Digital products autoconfigure.
Physically installing a DECconcentrator is described in the installation manuals for
the product.

Concentrators

3–7

The concentrator can be an integral part of the following network topologies:

•

Dual ring of trees in a building or campus wiring system.

•

Dual ring of concentrators in a backbone wiring systems (Figure 3–4).

•

Standalone workgroup configurations such as LANs connecting high-performance workstations and servers. These workgroups can also be connected
to a backbone system.

Figure 3–4:

Dual Ring of Concentrators in a Backbone Wiring System

VAX

Workstation

Minicomputer

VAX

Concentrator

Workstation

100 Mb/s FDDI ring

Workstation

Concentrator

Workstation

Mainframe

File
server

Concentrator

VAX

Workstation
LKG–3812–89I

3–8

Fiber Distributed Data Interface System Level Description

3.5.1 Concentrator Connection Rules

Use the following rules when connecting a concentrator into an FDDI network.

•

Maximum distance between concentrators must not exceed 2 km (1.2 mi) for
multimode connection, 1 km (0.62 mi) for low-power multimode connections, and 40 km (24 mi) for single-mode connections.This includes cables
connected between the patch panel and the device.

•

Total length of the network must not exceed 200 km (124.2 mi). This distance must include the length of the secondary ring, if a dual ring is implemented.

•

Distance between concentrators and attached stations must not exceed 2 km
(1.2 mi) for multimode connections, 1 km (0.62 mi) for low-power multimode connections, and 40 km (24 mi) for single-mode connections.

•

Maximum number of stations in an FDDI LAN is 500 (1000 Physical layer
entities). A station can have one Physical layer entity (SAS), two Physical
layer entities (DAS), or multiple Physical layer entities (concentrators). Digital supports the configurations specified in the ANSI FDDI standard. However, Digital recommends that an FDDI LAN generally be limited to 100
stations and 20 concentrators due to performance and reliability characteristics of large rings (see also Section 1.8).

•

Maximum number of concentrators in an FDDI network (dual ring of trees
and dual ring) is limited by the maximum number of Physical layer entities
that can connect to the ring (see also Section 1.8).

•

Maximum optical attenuation allowable between a concentrator port and its
neighbor is 11 dB for multimode connections, 7 dB for low-power multimode
connections, and 22 dB for single-mode connections. Minimum allowable
optical attenuation for single-mode connections is 12 dB.

Refer to Section 2.7.2 for FDDI device connection rules.

Concentrators

3–9

3.5.2 Concentrators and Structured Wiring

Commercial building wiring standards, such as EIA/TIA 568, address all communication needs including voice, data, and video for single or multiple building applications. EIA/TIA 568 defines a wiring system that supports a multivendor, multiproduct environment. At the center of the wiring system is a hierarchy of connection
equipment that supports the network.
The connection hierarchy specified in Digital’s implementation of EIA/TIA 568 is
the Main Distribution Frame (MDF), Intermediate Distribution Frame (IDF), Horizontal Distribution Frame (HDF), and the Wallbox. The concentrator can be located
anywhere in the structured wiring scheme.
Digital’s implementation of FDDI fully supports the EIA/TIA 568 Building Wiring
Standard as it applies to fiber optic cabling. The FDDI wiring subsystem structures
consist of the following:

•

Interbuilding backbone (multibuilding system) that goes from a Main Distribution Frame (MDF) in one building to the Intermediate Distribution Frame
(IDF) in one or more buildings.

•

Intrabuilding backbone (single building system) that goes from the IDF to
one or more HDFs.

•

Horizontal wiring system that goes from an HDF to an office Wallbox.

In the Interbuilding Campus Backbone, concentrators are connected in a dual ring at
the Main Distribution Frame in one building. From the MDF, fiber cable runs to concentrators at Intermediate Distribution Frames in the other buildings. Within each
building, concentrators at the IDFs connect to concentrators at one or more HDFs.
From the HDF, fiber cable can be run directly to the end station by means of the Wallbox.
Figure 3–5 is a conceptual view of this wiring scheme using the recommendations
set forth in EIA/TIA 568 for structured wiring installations. It is possible to install a
custom ring that exceeds the recommendations of this standard as long as FDDI distance and PMD-specific loss limits are not exceeded.

3–10

Fiber Distributed Data Interface System Level Description

Figure 3–5:

Concentrators Cascaded in Equipment Rooms

Main
distribution
frame (MDF)

(One per site)

Campus backbone

Intermediate
distribution
frame (IDF)

(One per building)
2000 m*
(6560 ft)
Building backbone
500 m*
(1640 ft)
Horizontal
distribution
frame (HDF)

Horizontal
wiring
Horizontal
wiring

To office wallbox
*Recommendations set forth in EIA/TIA 568 for structured wiring schemes.
LKG–3796–90I

Figure 3–6 shows the concentrator in dual ring of tree topology. Figure 3–6 is a conceptual view of a campus backbone wiring scheme. This figure also shows the PHY
port connections for each concentrator.

Concentrators

3–11

Figure 3–6:

Dual Ring of Trees in a Campus Environment
MDF

DAC 2
B

DAC 1

Secondary
A

Primary

IDF
Concentrator 1

IDF
Concentrator 3
B

A
M

B
M

SAS

DAS

SAS

DAS

HDF
Concentrator 2

HDF
Concentrator 4
B

B
M

SAS

DAS

SAS

DAS

SAS

DACs1 and 2 are connected through PHY ports A and B to the dual ring.
Concentrators attach to the ring through PHY ports M and B.
DASs connect to the ring through PHY port B. DAS PHY port A is not
used in this topology.
Arrows indicate direction of token flow.
LKG–5097–90I

3–12

Fiber Distributed Data Interface System Level Description

4
Bridges

Bridges are the connecting devices between local area networks in an extended LAN
architecture. Bridges allow the creation of extended LANs by acting as data link relays between similar or dissimilar LAN types (see Figure 4–1). In the FDDI environment, the bridge acts as a link between FDDI LANs or an FDDI LAN and other LAN
types. This chapter explains bridge operation and the DECbridge 500/600 series
FDDI-to-802.3/Ethernet SAS and DAS bridges. Also included in this chapter is a
functional bridge description along with planning and configuration considerations
for FDDI bridge. For a more detailed discussion of bridges, refer to the Bridge and
Extended LAN Reference manual.

4–1

Figure 4–1:

Extended LAN Examples

Concentrator
Concentrator

DAS bridge
Concentrator

Other networks
SAS
bridge
Other network

• •••••
LKG–4075–90I

4.1 Description
Bridges are protocol-independent, specialized LAN stations that provide the connection point between two or more local area networks. They are store-and-forward
devices operating at the Data Link layer. Bridges receive and store frames from attached LANs and selectively filter the frames or forward them to their destination
address. This reduces total traffic on the extended LAN by filtering unnecessary traffic from the network.

4–2

Fiber Distributed Data Interface System Level Description

In the FDDI environment, SAS bridges connect to the dual ring through a PHY M
port connection of a concentrator. DAS bridges connect directly to the dual ring or to
concentrators. One or more ports to the other network(s) complete the data path
through the bridge. Figure 4–2 is a functional diagram of a typical FDDI bridge. A
SAS bridge has a PHY S port. A DAS bridge has a PHY A and a PHY B port. Both
SAS and DAS bridges have all the additional elements of a SAS or DAS including
the PHY, PMD, MAC, and SMT.
Figure 4–2:

FDDI Bridge Functional Diagram

FDDI
Queue
manager

FDDI
packet
memory

FDDI
ring

FDDI PHY
port
(SAS or DAS)

Table
lookup/
address
table

Translation
processor

Other LAN(s)
filtering/
forwarding
control
processor
(multiport
models)

Other
network
packet
memory

Other network
port(s)

Other
network(s)
LKG–4076–90I

Bridges

4–3

4.2 Bridge Functions
Basic bridge functions include translation, frame forwarding and filtering, source
address tracking, participation in the spanning tree algorithm, network management,
and Internet Protocol (IP) fragmentation. FDDI bridge structure includes an FDDI
SAS or DAS port(s), one or more ports to other network(s), address table and table
lookup logic, packet memories, translation processor, and the forwarding/filtering
control processor.
4.2.1 Translation Not Encapsulation

A major reason for choosing FDDI is that it is designed to be an industry standard for
high-speed networks. To take advantage of this, users should avoid FDDI components that lock them into a specific vendor implementation. Otherwise, the user becomes tied to a closed, proprietary system with equipment choices limited to a single
source.
Encapsulation is a technique used by some vendors to send data over the extended
network. In encapsulation, packets received by the bridge from the non-FDDI network become the data in an FDDI frame. Using a proprietary protocol, the bridge
surrounds, or encapsulates, the packet with an FDDI header and trailer and retransmits it onto the FDDI ring. Only a bridge, or other receiving device, supplied by the
same vendor or one who subscribes to the same proprietary protocol can make use of
this encapsulated data. Although this method is fairly simple to implement, it results
in the user being forced into a closed, proprietary, single-vendor network architecture.
To avoid this pitfall, and in keeping with the open system philosophy, Digital does
not use encapsulation. Digital employs the translation process in its bridge architecture. In the translation process, the bridge modifies the fields of a forwarded packet,
making it compliant with the packet format of the network to which it is being forwarded. Although difficult to implement because it must be done in real time, translation allows the use of other vendor components in the network as long as they
adhere to the various network standards on both sides of the bridge. See Section 4.3
for further details on translation.

4–4

Fiber Distributed Data Interface System Level Description

4.2.2 Frame Forwarding and Filtering

The basic purpose of a bridge is to forward frames between its attached LANs, so that
the stations on each LAN can communicate with all other stations as if they were on
one large network. Bridges do not forward all frames that pass across a LAN. Bridges
forward only those frames destined for a station on another port. If a frame is destined
for a station known to be on the same port as the sending station, the bridge discards
the frame.
Under network management control, bridges can be configured to filter certain
frames from crossing the bridge. For example, individual addresses can be set so that
the bridge will always filter, or always forward, frames destined to these addresses.
The bridge does this regardless of whether or not it recognizes the addresses as belonging on the extended network.
Additionally, the bridge can be instructed to filter frames based on their source address and/or protocol type. These filters can be set on groups of ports, or they can be
set globally.
4.2.3 Source Address Tracking

Digital and IEEE 802.1 require that bridges be self-configuring. For a bridge to determine whether to forward or discard a given frame, it must maintain a list of all
known station addresses in the extended network, and the port on which the station is
located. Each bridge maintains a source address table that it automatically updates.
This table, called the Forwarding Database, is built from the source addresses in received frames.
When a bridge receives a frame, it records the source address of the frame into its
Forwarding Database (address table), along with an indication of the port from
which the frame was received. Because every frame contains the source address of
the transmitting node, the bridge is constantly updating its forwarding database,
learning the location of every station that sends a frame over the network. In an extended network, every bridge learns where stations are located, relative to its own
position in the network, as frames pass through the bridge.

Bridges

4–5

4.2.4 Spanning Tree

The logical topology of an extended LAN must be loop-free to prevent the circulation of packets indefinitely on the extended network.To eliminate this problem, the
extended network forms a logical configuration, called a spanning tree, which eliminates logical loops in the extended network. This spanning tree is based on the spanning tree algorithm defined by Digital and implemented with minor changes by
IEEE 802.1d. Digital bridges default to the IEEE 802.1d application and use the
LAN Bridge 100 spanning tree implementation if they detect bridges that do not support IEEE 802.1d.
The spanning tree algorithm has two primary functions: the creation of one and only
one path between any two bridges, and the proper connection of all LANs into a
single extended LAN. The algorithm works by having all bridges in a given LAN
exchange Hello messages (bridge PDUs). The bridges use the information in these
messages to elect which bridges can forward frames and which cannot.
Bridges that form redundant paths in an extended LAN go into backup mode. In
backup mode, bridges continue to monitor network traffic but do not forward frames.
If the active bridge fails, the backup bridge notices the failure and begins to forward
frames.
4.2.5 Network Management

Network management can monitor conditions on the extended network and adjust
the logical topology to optimize traffic patterns and to fix things if they go wrong. To
these ends, bridges maintain counters to keep track of the number of frames forwarded and discarded, and other such events.
The bridge forwarding database can be remotely observed and modified. This allows
network management to determine the location of user stations, or to manage various
regions of the extended network. By adjusting parameters in the spanning tree computational process, a network manager can optimize the algorithm performance for
specific topologies.
The spanning tree algorithm database also contains useful information about the extended network topology. This information is made available to the network manager through the use of higher-level protocol messages.

4–6

Fiber Distributed Data Interface System Level Description

Network management is also used to set special filters; for example, protocol filters.
Network management entries for these filters are stored in a nonvolatile memory so
that they can be restored easily on power up.
4.2.6 Internet Protocol Fragmentation

FDDI permits frames up to 4500 bytes in length. The 802.3/Ethernet LAN cannot
handle frame sizes this large. To ensure that frame size restrictions are not violated
when frames cross the bridge, the bridge performs a process called fragmentation on
large FDDI IP frames. Fragmentation occurs when a frame is larger than the frame
size allowed by the outbound port.
A frame is fragmented only if its LLC header indicates that it is an IP frame. The
bridge receives the frame and decides how many packets must be generated. It then
constructs the new packet(s) with a protocol type indicating IP frame and a new
checksum in the IP header. The bridge also manipulates the header bits to indicate
that the frame is a fragmented frame. The outbound port transmits the resulting packets with new data link CRCs for each packet. At the destination station, the frames are
reassembled in the proper sequence.

4.3 Bridge Operation
The FDDI bridge provides a transparent link between networks. Users on the FDDI
ring can freely exchange data with users on other networks. Data is transmitted over
the FDDI ring in frames up to 4500 bytes in length. As FDDI data is received in the
bridge, it is placed in packet memory buffers arranged in a queue structure. Each
buffer is assigned a descriptor containing status information about buffer contents.
The queue manager (QM) determines the ultimate destination of frames received
from the FDDI ring. The QM does this by looking at the descriptor, source and destination address, Frame Control field, type and quantity of previously received
frames, and the bridge forwarding database. Based on this information, the QM forwards the frames or discards them. Packets destined for another network are placed
into the FDDI forward queue for retransmission from the other network port.

Bridges

4–7

At this point, the translation processor takes the packets and descriptors and physically moves them into the packet memory of the other network interface. As they are
moved, the packets are converted to the format required by the other network. These
packets are then passed to the network interface controller and eventually clocked
onto the network.
Data received from the other network is handled in basically the same fashion. The
filtering/forwarding processor determines where the received frames are to be sent.
Packets are moved into the translation processor and converted to FDDI format.
Data to be transmitted to the FDDI ring is placed in the FDDI transmit queue for retransmission to the FDDI network.
4.3.1 FDDI Port

The DECbridge 500/600 series incorporates a PHY S port or a PHY A and PHY B
port in its design, making the bridge a SAS or DAS. FDDI entities in the bridge include a PHY, PMD, MAC, and SMT. Figure 4–1 shows SAS and DAS bridge connections to FDDI. Figure 4–3 shows a SAS bridge connection to a concentrator.
Figure 4–3:

SAS Bridge-to-Concentrator Connection

Concentrator
PHY
M

PHY
S
SAS bridge
LKG–4247–90I

4–8

Fiber Distributed Data Interface System Level Description

Data from the FDDI network passes through the FDDI interface where it is converted
from optical signals, error checking is performed, and the resultant data is passed to
FDDI packet memory.

4.4 Performance Considerations
A major reason for migrating to the FDDI network is its high bandwidth and resultant
large data-transmitting capacity. To take advantage of this power, a bridge must be
capable of supporting high throughput. It should not be a bottleneck on the ring. To
be effective, a bridge must make frame forwarding decisions in real time. Key bridge
performance characteristics are forwarding rate, filtering rate, and latency. Digital
provides the performance characteristics in all three areas to support the FDDI bandwidth.
The forwarding rate is based on the number of data packets that can be transferred to
and from the FDDI network. It should be large enough to allow a high volume of traffic between the FDDI ring and other networks. Forwarding rate is determined by the
speed of the bridge in parsing a packet, verifying the destination address by checking
its forwarding data base, and determining whether or not to forward the packet.
The filtering rate is the speed at which the bridge processes received packets to determine whether they are to be forwarded across the bridge. A high filtering rate keeps
the bridge from clogging with data packets that are not being forwarded. Efficient
filtering reduces the probability of overflowing internal storage areas and missing a
packet to be forwarded. Efficient filtering reduces the probability of a lost packet.
Latency is the delay incurred when data crosses the bridge. Certain applications and
protocols are delay sensitive. High latency results in nodes not being able to access
devices on remote networks. High latency can cause the delay time between end
nodes in an application to exceed the application’s response time requirements, causing poor performance or connection failures. Digital designs its bridges to minimize
latency time, thereby enhancing bridge performance.

Bridges

4–9

4.5 Bridge Planning Considerations
When planning for the installation of an FDDI Bridge such as the DECbridge
500/600 series, consult the following Digital manuals:

•

DECconnect System Fiber Optic Planning and Configuration

•

DECconnect System Fiber Optic Installation

4.6 Bridge Configuration Guidelines
This section lists general configuration guidelines for the DECbridge 500/600 series.
These bridges allow the use of a SAS or DAS connection to the FDDI network, and
provide one or three ports that interface to 802.3/Ethernet networks.
The DECbridge 500/600 series connects to the FDDI ring through a concentrator
(SAS bridges) or directly to the ring (DAS bridges). Although the FDDI standards
allow a PHY S port to connect to any other FDDI port type, Digital recommends that
a SAS bridge’s PHY S port connect to a PHY M port of the concentrator. Refer to
Section 2.7.2 for FDDI device connection rules.
4.6.1 General Connection Rules

Use the following guidelines when connecting one of the DECbridge 500/600 series
into an FDDI network:

4–10

•

SAS bridges should connect to a PHY M port on an FDDI concentrator to
work in an FDDI ring. DAS bridges can connect directly to the dual ring or
can be dual homed through a PHY M port on two separate concentrators.

•

Maximum distance between the bridge and the concentrator or adjacent station must not exceed 2 km (1.2 mi) for multimode connections (refer to
Figure 4–4) and 40 km (24 mi) for single-mode connections. This includes
cables between the patch panel and the device.

•

Fiber Distributed Data Interface System Level Description

Digital supports the configurations specified in the ANSI FDDI standard.
However, Digital recommends that an FDDI LAN generally be limited to 100
stations and 20 concentrators due to performance and reliability characteristics of large rings (see also Section 1.8).

•

Maximum number of bridges (SAS and DAS) in an FDDI network is limited
by the maximum number of Physical layer entities that can connect to the
ring (see also Section 1.8).

•

Maximum number of bridges between two stations on the extended LAN is
seven.

•

Maximum allowable optical attenuation between the bridge and an adjacent
FDDI station is 11 dB for multimode connections and 22 dB for single-mode
connections. Minimum allowable optical attenuation for single-mode connections is 12 dB.

Figure 4–4:

Bridge Distance Requirements for Multimode Connections

FDDI 100 Mb/s network
Concentrator
2 km (1.2 mi) maximum
distance to next station

2 km (1.2 mi) maximum
distance to next station

DAS
bridge

2 km (1.2 mi) maximum
distance from concentrator
Digital
FDDI bridge

Other network

• 50 m (165 ft) maximum distance if 802.3
transceiver is connected to the
AUI interface
• 185 m (610 ft) maximum segment length
if connected to the ThinWire port
LKG–4077–90I

Bridges

4–11

The following rules apply if a DECbridge 500/600 series unit is being installed using
the AUI connector on the 802.3/Ethernet side of the bridge:

•

If the bridge connects to an IEEE 802.3 transceiver such as the H4005, the
transceiver cable must be an IEEE 802.3 compliant transceiver cable
(BNE3H/K/L/M or BNE4C/D).

•

IEEE 802.3 transceiver cables and Ethernet transceiver cables cannot be interconnected.

•

If the bridge connects to an 802.3/Ethernet network, maximum length of the
802.3/Ethernet transceiver cable cannot exceed 50 m (164 ft). This maximum
length is reduced by the internal cabling equivalency of a device (such as a
DELNI) that is connected between the bridge and the transceiver, or by the
use of an office transceiver cable. The internal cabling equivalency of such a
device must be subtracted from the 50-m (164-ft) maximum.
NOTE
Cabling equivalency is a measure of the internal timing
delay of a device expressed in meters of transceiver
cable, for example:
If a device has a 5-m (16.4-ft) cabling equivalency, then
its maximum allowable transceiver cable length is 50 m
(164 ft) minus 5 m (16.4 ft) or 45 m (148 ft).
Office transceiver cable (BNE4x–xx), because of its
smaller diameter, has a signal loss that is four times that
of the (BNE3x–xx) transceiver cable. If office transceiver cable is used, the maximum transceiver cable distance
must be divided by 4. The maximum office transceiver
cable length allowed is 12.5 m (41 ft).

If a Digital FDDI-to-802.3/Ethernet bridge is being installed using the ThinWire
connector, maximum length of the ThinWire cable segment is 185 m (606.95 ft).
For device-specific information related to cabling equivalency, refer to the DECconnect System Planning and Configuration Guide.

4–12

Fiber Distributed Data Interface System Level Description

5
Adapters

FDDI adapters connect end-user nodes to the FDDI ring. Adapters allow attached
devices to act as single attachment stations (SAS) or dual attachment stations (DAS).
This chapter provides a functional overview of adapters. It also discusses planning
and configuration considerations for these devices.
Several different adapters are available in the Digital product set. This chapter provides a general overview of one type of adapter. Although construction and implementation may vary from one adapter to another, basic functionality and purpose, the
connection of a device to the FDDI network, remains the same.

5.1 Description
Digital’s family of FDDI adapters are single attachment stations that connect to the
FDDI ring by means of the concentrator. Part of the adapter is an interface to a specific workstation or system bus; for example, TURBOchannel, or other standard bus
interface (see Figure 5–1). The SAS components of the adapter consist of a PMD,
PHY, MAC, and SMT entity. Connection to the FDDI ring is through a PHY S port.
In order to properly interface to the ring, the adapter should attach to a PHY M port on
a concentrator. Section 5.5 lists basic configuration guidelines for the adapter.

5.2 Adapter Functionality
Adapters connect to the end node’s internal bus structure. Digital’s FDDI adapters
use internal microprocessor-controlled logic to permit data transfer between the end
node and the FDDI network without intervention of the host processor.

5–1

Digital FDDI adapters can be logically divided into four separate components as follows: (See Figure 5–1.)

•

Host interface logic

•

Adapter manager

•

Packet memory

•

FDDI SAS port

Figure 5–1:

Adapter Model

Ring entry
mover

Host
interface

To internal system bus

Packet memory and
packet memory controller

Adapter
manager

FDDI
SAS port

To concentrator
LKG–4078–90I

5.2.1 Host Interface

The host interface connects the adapter to the host’s internal bus. It contains all the
necessary logic to pass frames to and from host memory. As part of its structure, buffers in the host interface provide the capability for high-speed data movement to and
from the host.

5–2

Fiber Distributed Data Interface System Level Description

5.2.2 Adapter Manager

The adapter manager oversees the transfer of data between the node and the FDDI
ring. Functions performed by the adapter manager include the following:

•

Adapter initialization

•

Error handling

•

Device driver command implementation

•

Local management of the FDDI MAC, PHY, and PMD components within
the adapter

•

Collection and maintenance of node statistics

5.2.3 Packet Memory

Packet memory within the adapter operates in much the same manner as the memory
in a bridge. Data from the FDDI ring is placed in queues within this memory space.
Virtual addressing is used to access the data as it progresses through the queues. As in
the bridge, the data itself never actually moves until its final destination is determined. A descriptor word containing control information for each received packet is
actually passed around the entities. This word is changed as required when data ownership changes. Depending on the type of packet or frame received, the data is passed
onto the node, forwarded to the adapter manager, or discarded.
Similarly, data from the node moves through packet memory. Once it is determined
that the data is meant for transmission onto the ring, it is passed to the FDDI port.
5.2.4 FDDI Port

The adapter incorporates a PHY S port in its design, making the adapter a SAS node.
FDDI components in the adapter include a PHY, PMD, MAC, and SMT entity. Data
from the node that is destined for the FDDI ring passes to the FDDI port. Here the
data is formatted into FDDI frames and transmitted over the ring.

5.3 Operation
Actual adapter operation is a complex process. The following sections provide a
summary of the initialization and ring-access processes that an adapter goes through
when it is first powered on.

Adapters

5–3

5.3.1 Initialization

The adapter must be initialized when it is powered on or recovering from a fatal error
condition. In the initialized state, the adapter has allocated all buffers for internal
memory, loaded all registers with appropriate start addresses, and is ready to process
commands from the user device. During this process the following occurs:

•

Self-test diagnostics are run to ensure adapter integrity. If the tests pass, the
remaining events occur. If not, the adapter is not initialized.

•

Internal logic devices are initialized.

•

The 48-bit adapter LAN address is loaded into data registers.

5.3.2 Accessing the Ring

Once the adapter is initialized, it must access the FDDI ring. This is accomplished as
follows:

•

After successful initialization, the adapter transitions to the run state and
starts FDDI logic testing.

•

After successful completion, the FDDI logic requests connection to the ring
by generating a series of claim frames.

•

After synchronizing with the ring, physical and logical connections are established and the adapter is able to communicate with other attached devices.

5.3.3 Network Management Functions

To permit network management, Digital adapters incorporate a number of functions
in their basic design. Proper use of these functions provide the network or system
manager with the ability to monitor and control adapters individually, or as a distinct
group, from a remote location. These adapter functions include the following:

5–4

•

Request ID

•

System ID

•

Request counters

•

Remote boot

Fiber Distributed Data Interface System Level Description

5.4 Adapter Planning Considerations
When planning for the installation of an FDDI adapter, consult the following Digital
manuals:

•

DECconnect System Fiber Optic Planning and Configuration

•

DECconnect System Fiber Optic Installation

5.5 Adapter Configuration Guidelines
This section lists general configuration guidelines for Digital’s adapters. Digital
adapters are single attachment stations that connect to the FDDI ring by means of the
concentrator.
Although the FDDI standards allow a PHY S port to connect to any other FDDI port
type, Digital recommends that a PHY S port connect to a PHY M port of a concentrator. Refer to Section 2.7.2 for FDDI device connection rules.
Use the following guidelines when connecting an adapter into an FDDI network.

•

SAS adapters should connect to a concentrator PHY M port to work in an
FDDI ring. In a standalone configuration, the adapter PHY S port can connect to a concentrator PHY A or PHY B port.

•

Maximum allowable distance between the concentrator and an adapter is 2
km (1.2 mi) for multimode fiber connections, and 100 m (330 ft) for ThinWire and shielded twisted-pair. This includes cables between the patch panel
and the device. Figure 5–2 shows the distance requirements for an FDDI
adapter with multimode fiber. It also shows the connection to a concentrator.

•

Maximum number of stations in an FDDI LAN is 500 (1000 Physical layer
entities). A station can have one Physical layer entity (SAS), two Physical
layer entities (DAS), or multiple Physical layer entities (concentrators).
Digital supports the configurations specified in the ANSI FDDI standard.
However, Digital recommends that an FDDI LAN generally be limited to 100
stations and 20 concentrators due to performance and reliability characteristics of large rings (see also Section 1.8).

Adapters

5–5

•

Maximum number of adapters in an FDDI network is limited by the maximum number of physical layer entities that can connect to the ring (see also
Section 1.8).

•

Maximum number of patch-panel connections between the concentrator and
the adapter is two connections.

•

Maximum allowable loss between an adapter and the concentrator is 11 dB
for multimode connections.

Figure 5–2:

Adapter Distance Requirements for Multimode Fiber

PHY A

PHY B
Concentrator
PHY M

2 km (1.2 mi) maximum distance

PHY S
Port logic
Bus interface
Device internal bus
LKG–4079–89I

5–6

Fiber Distributed Data Interface System Level Description

6
Network Management

Proper management is important to the successful maintenance of a network. Without sophisticated and flexible management software, even the most powerful network can become an unwieldy, hard-to-maintain system.
Network management, as envisioned by Digital, is a means for network and systemlevel personnel to effectively design, implement, and maintain networks to meet specific organizational goals and technical needs. Digital provides a framework for the
management of heterogeneous, multivendor systems linked by a number of network
environments. This ability to manage devices over an extended network provides users with tremendous flexibility in terms of network design and cost.
In the Digital scheme, network management is one component of Enterprise Management. This term applies to Digital’s ability to manage not just a given network,
but a broad spectrum of systems, including voice and data, corporate-wide and personal computing, and distributed computing environments.
Digital believes that the concept of Enterprise Management Architecture is where
the future of system management lies. In designing EMA, Digital has followed the
model set forth in the OSI system management architecture. This chapter describes
part of the Enterprise Management Architecture through a discussion of management over the extended network.

6–1

6.1 Network Management Benefits
Digital’s network management tools provide the following benefits:

•

Flexibility

•

Extensibility

•

Integration

•

Growth Potential

A further benefit is protection of the network investment. Because existing components, as well as any additions, are supported, there is no need to discard existing
equipment when expanding network facilities.

6.2 Network Management Functions
Network management functions are those parts of the software system that network
and system-level personnel use to monitor, control, and maintain the network. As set
forth in the ISO/OSI management framework, these functions are as follows:

•

Network configuration management to set or change operating parameters of
the network

•

Fault management to detect, diagnose, and correct network faults and error
conditions

•

Performance management to monitor and evaluate network performance

•

Accounting management to monitor and evaluate network resources

•

Security management to provide a secure computing environment

If management functions such as fault notification and fault logging are implemented independently of the network components being managed, the result is an increase in the scope of the management system. Management functions must be
capable of being customized according to user requirements. They must be capable
of interaction, so that the output of one function can be used as input to another. This
provides for the implementation of more sophisticated management applications to
support the network systems throughout their existence, from planning and implementation to operation and change.

6–2

Fiber Distributed Data Interface System Level Description

The role of the ISO/OSI categories is to serve as the basis for developing usable management applications. For example, to be truly effective, network operation requires
the integration of performance, fault, and configuration management information.
The following sections describe the various ISO/OSI components in further detail.
6.2.1 Network Configuration Management

Network configuration management provides a means for configuring the network
and its components through setting or changing the operating parameters of the local
or extended network. This includes turning the parameters on or off, altering their
operational status, and collecting and distributing current status information of attached devices.
Ring mapping, a part of FDDI configuration management, enables users to identify
all stations on the FDDI network. This feature allows mapping of the FDDI network
configuration, providing network managers with valuable ring status information.
6.2.2 Fault Management

Fault management provides a means for detecting, diagnosing, and correcting network faults and error conditions. Fault management capabilities include the receipt
of unsolicited error messages from network devices, periodic polling for error messages, and setting error thresholds for individual devices. Through proper setting and
monitoring of operational parameters (for example, error alarms and threshold limits), network managers can actually prevent certain fault conditions.
6.2.3 Performance Management

Performance management allows the network manager to monitor and evaluate the
performance of local and extended networks by collecting and processing relevant
data. The information is gathered from device counters and error reports, then processed and analyzed for network planning and tuning.
6.2.4 Accounting Management

Accounting management allows managers to effectively monitor and evaluate network resources and their use. This function identifies usage costs associated with the
network, and provides the necessary tools for identifying and charging users for network time. Accounting management is an application function.

Network Management

6–3

6.2.5 Security Management

As networks become larger and more complex, and as more companies place highly
sensitive data on the network, the ability to create a secure environment increases in
importance. Security management defines those facilities required to manage network services: authentication, control of access to resources, and the protection of
confidential information on both the local and extended network.

6.3 Digital’s Network Management Capabilities
Network management is done through the use of a layered software product that runs
on a designated host station. This software communicates with corresponding software on network devices, allowing the network manager to remotely configure,
manage, and monitor devices on the extended LAN. These devices include the following:

•

FDDI and non-FDDI bridges

•

FDDI concentrators

The following sections describe some of the capabilities of Digital’s network management software.
6.3.1 Device Naming

With the network management software, all manageable bridges and concentrators
can be assigned unique ASCII names, as opposed to more cumbersome hardware addresses. These names can then be used in commands to the devices.
6.3.2 Device Registry

The management software creates a network registry containing information about
each manageable bridge and concentrator in the extended LAN. This registry contains device type, address, name, and a brief description for each attached device.
Registry information can be added or modified as required. Depending on the management software, this information can also be automatically collected as new
bridges or concentrators are added to the network.

6–4

Fiber Distributed Data Interface System Level Description

6.3.3 Event Messages and Event Log

The network management station can display error and event messages. The network manager collects these messages by polling and by error or event logging.
Event codes can also be recorded in an event log. The network management software
provides a simple mechanism for report generating based on the contents of the event
log.
6.3.4 Security

In any open network, security is always a primary concern. Digital’s network management software provides the ability to assign passwords to all bridges and concentrators in the extended LAN. Set from the network management station, these
passwords prevent the unauthorized control and configuration of network devices.
Devices store the passwords in their nonvolatile memory.
6.3.5 Configuration Parameters

Network management software allows the user to remotely set bridge and concentrator configuration parameters from the network management station. Configurable
parameters include the following:

•

Spanning tree parameters—Controls the extended LAN logical layout and
performance by manipulating bridge and line spanning tree parameters.

•

Firmware upgrade down-line loading—Enables a bridge or concentrator to
accept down-line loading of firmware upgrades, eliminating the need to perform upgrades at the bridge or concentrator site.

•

IP fragmentation—Enables an FDDI bridge to break large Internet Protocol
(IP) frames received from the FDDI LAN into smaller frames for retransmission across the bridge. For example, fragmentation occurs in Digital FDDIto-Ethernet bridges because maximum frame size in the FDDI LAN is 4500
bytes, but 802.3/Ethernet can only support a 1500-byte frame size. The DECbridge 500/600 series units come with the default mode set to fragment IP
traffic.

•

IP fragmentation error checking—Enables an FDDI bridge to perform header
checksum verification on large IP frames received from the FDDI LAN to
ensure data integrity of frames being sent across the bridge. The DECbridge
500/600 series units automatically perform IP fragmentation error checking.

Network Management

6–5

6.3.6 Initialization

The user can initialize bridges and concentrators from a network management station. Doing this forces the device to reset its internal counters to zero and run its selftest diagnostics. At initialization, bridges perform the spanning tree computation
process that determines active and standby bridges. Concentrators perform the initialization process described in Section 2.9.3.
6.3.7 Bridge Port Enabling and Configuring

From the network management station, bridge ports in the extended LAN can be enabled and disabled. As bridges are enabled and disabled, the spanning tree computation process determines which are the active and which are the standby bridges in the
network.
In the FDDI environment, network management software can set the maximum token rotation time (T_MAX), the requested token rotation time (T_REQ), and Valid
Transmission Time (TVX). Setting T_REQ and T_MAX for all attached devices on
the FDDI LAN allows the user to control the value of Target Token Rotation Time
(TTRT) on the FDDI LAN.
6.3.8 PHY Port Configuration and Control

The network manager can enable and disable the PHY ports of concentrators. This
allows the configuring of networks to fit the needs of the network users. Managers
also have the ability to set Link Error Monitoring (LEM) threshold limits for FDDI
devices. LEM performs continuous testing of links during ring operation to ensure
the bit error rate is being met.
6.3.9 Address Filtering

With Digital’s network management software, network managers can add or remove
addresses at the forwarding database of a bridge (FDDI and non-FDDI). This capability is used to block certain stations from sending or receiving frames across a
bridge, localizing traffic on a specific subnetwork or to stations within a subnetwork.

6–6

Fiber Distributed Data Interface System Level Description

6.3.10 Protocol Filtering

In addition to address filtering, network management software allows the filtering of
protocols at a bridge (FDDI or non-FDDI). This capability instructs the bridge to forward or filter (discard) frames based on the protocol information within the frame.
This feature is useful to regionalize traffic or to isolate broadcast storms or other protocol-related problems. Selective protocol filtering allows only selected protocols to
cross the bridge. Again, selected protocols are specified using the network management station and supported software.
6.3.11 Network Monitoring

Device status, characteristics, and counter contents are all accessible through the network management software. Bridges, concentrators, and the connection to each port
on these devices can be addressed. In the case of bridges, the spanning tree parameter
settings and the outcome of the spanning tree computational process are accessible
using the software.
6.3.12 FDDI Physical Ring Mapping

The network management software supports creation of a physical ring map of attached devices in the FDDI LAN. This map allows a user to determine the order of
token rotation among devices on the ring.
6.3.13 SMT Neighborhood Information Frame (NIF) Display

Besides displaying the NIF contents, network management software can use the NIF
frames to show the order in which each station appears in the token path. Other uses
of the NIF contents include downstream neighbor information gathering and duplicate address detection.
6.3.14 SMT Station Information Frame (SIF) Display

The network management software can display SIF frames from individual FDDI
stations. These frames contain a station’s configuration and operating information.
Network management software can create a physical ring map using the SIF configuration information. The operating information can be used as an aid in ring fault
detection.

Network Management

6–7

6.3.15 Troubleshooting

Digital’s network management software provides the network user with a wealth of
useful information. Proper use and interpretation of this information results in a
powerful troubleshooting tool. By properly analyzing the network performance and
error information, a user can diagnose and correct a variety of network problems
with minimal effect on overall network performance.
Combining the FDDI dual ring of trees with the power of Digital’s network management software results in a dynamic network resistant to failures caused by everyday
events. Proper use of the network management tools allows the user to isolate and
remove problem network devices with little or no effect on network performance or
throughput.

6–8

Fiber Distributed Data Interface System Level Description

A
OSI Reference Model

Chapter 1 describes FDDI and its relationship to the OSI reference model. This appendix provides an overview for readers unfamiliar with the OSI reference model.
The International Organization for Standardization (ISO) has established what has
become known as the Open Systems Interconnection (OSI) Reference Model. The
OSI model implements the design structure known as layering. In this structure, the
communications functions are divided into a vertical set of layers, with each layer
performing a related subset of the functions required to communicate with another
system. Each layer relies on the services of the next lower layer and the ability of the
protocols of the lower layer to conceal those functions from it and layers further up
the ladder. Layering provides modular design and allows independence of design
teams. As shown in Figure A–1, the OSI model consists of seven layers, with an interface between layers .
The layered approach to networking adheres to the following general guidelines:

•

Each level should perform a well-defined function.

•

The layers should comply with recognized standards.

•

Layer interfaces should be designed to minimize data flow between layers.

•

Layers should be sized to minimize complexity without making the layer unmanageable.

A–1

Figure A–1: OSI Seven-Layer Model

Application
layer

Application protocol

Application
layer

Presentation
layer

Presentation protocol

Presentation
layer

Session
layer

Session protocol

Session
layer

Transport
layer

Transport protocol

Transport
layer

Network
layer

Network protocol

Network
layer

Data Link
layer

Data link protocol

Data Link
layer

Physical
layer

Physical connection

Physical
layer

Physical transport medium

LKG–4080–90I

A–2

Fiber Distributed Data Interface System Level Description

A.1

Physical Layer

At the Physical layer, the physical interface between devices and the rules for bit
transmission are established. The Physical layer defines the mechanical, electrical,
functional, and procedural elements of the network. Examples of Physical layer
standards are RS–232–C, RS–449, RS–422, and RS–423.
A.1.1 Data Link Layer

The Data Link layer provides the raw bit-stream service. It is responsible for providing reliable data transmission from one node to another. It provides the means to activate, maintain, and deactivate the link. This layer also provides error detection and
control functions to the higher layers of the OSI model.
If communication is established between two devices not on the same fiber, however,
the result is a number of functionally independent data links. In this instance, the
higher layer must assume responsibility for overall error detection and correction.
Examples include HDLC and LLC.
A.1.2 Network Layer

The Network layer provides for the transparent transfer of data between transport
entities. It relieves the Transport layer of the need to be aware of the underlying data
transmission and switching technologies used to connect systems. Examples include
Internet Protocol and the OSI Router (ISO 9542).
A.1.3 Transport Layer

The Transport layer provides a reliable means for the exchange of data between processes in different systems. This layer ensures that data units (frames) are delivered
error free, in sequence, with no losses or duplications. The Transport layer can provide optimization and requested service capabilities. For example, the session layer
could specify error rates, delay and priority, and security parameters for a given session. It is the Transport layer that would provide them. TCP is an example of this layer.

OSI Reference Model

A–3

A.1.4 Session Layer

The Session layer provides the means for controlling the dialogue between presentation entities. The Session layer permits two presentation entities to establish and use
a connection, called a session. In addition, it provides services such as dialog type
and recovery. The recovery function provides a mechanism for retrieving data transmitted during a session, in the event of a session failure.
A.1.5 Presentation Layer

The Presentation layer offers applications and terminal handler programs a set of
data transformation services. This layer provides such services as data translation,
formatting, and syntax selection. Examples of protocols in this layer are text compression and encryption.
A.1.6 Application Layer

The Application layer provides a means for application processes to access the OSI
environment. It contains management functions and other mechanisms to support
distributed applications. Examples of protocols at this level are virtual file protocol
and job transfer and manipulation protocol.

A–4

Fiber Distributed Data Interface System Level Description

B
Additional Configurations

The EIA/TIA 568 commercial building wiring standard provides an infrastructure
for FDDI networks. This appendix contains examples of FDDI network configurations built on Digital’s implementation of the commercial building wiring scheme.

B.1 Standalone Configurations Using Concentrators
Figure B–1 shows a concentrator installed in a standalone workgroup. The workstations can be either SAS or DAS devices. SAS devices are preferred for this configuration. This configuration is useful for compute-intensive applications shared by a
limited number of users in a limited geographical area, or where security is paramount.
Figure B–2 shows a dedicated network on a dual ring. The attached workstations,
file server, mini-computer, and main frame are SAS devices. This configuration can
serve the same purposes as the standalone configuration with the added benefit of
extending the geographical area.
These configurations allow the use of star wiring, which means existing structured
cabling can be used.

B–1

Figure B–1: Standalone Workgroup Installation

(SAS)

File
server

Workstation

Concentrator

Workstation

File
server

(SAS)

SAS = single attachment station
LKG–3797–89I

B–2

Fiber Distributed Data Interface System Level Description

Figure B–2: Dedicated Network in a Dual Ring Installation

Workstation
Workstation
Concentrator
Workstation
File
server

Workstation

100 Mb/s
FDDI ring
Concentrator

Minicomputer
Mainframe

LKG–4456–89I

B.2 Tree Configurations Using Concentrators and Bridges
Tree configurations are used when wiring large groups of user devices together. Concentrators and bridges are wired in a star topology with one or more concentrators
serving as the root of the tree. Figure B–3 shows a tree of concentrators installed in a
campus configuration wired with Digital’s implementation of EIA/TIA 568.
Figure B–4 shows a tree of concentrators installed in a building with multiple HDFs.
These configurations also show the use of single and multiport bridges. Refer to Section 3.5.2 and Figure 3–5 for the definition of distribution frames in structured wiring.

Additional Configurations

B–3

Figure B–3: FDDI Tree of Concentrators and Bridges, Campus Configuration

Building 1

Building 2

(HDF)

B
Bridge

Concentrator

(IDF)

Concentrators
(MDF)

Building 3
(HDF)

(IDF)

Building 4
(HDF)

(IDF)
Bridge

Concentrator
B
B
B

IDF = Intermediate distribution frame
HDF = Horizontal distribution frame
MDF = Main distribution frame
B
= Bridge
LKG–3809–90I

B–4

Fiber Distributed Data Interface System Level Description

Figure B–4: Tree of Concentrators Installed in a Building

802.3/Ethernet 802.3/Ethernet
10 Mb/s LAN
10 Mb/s LAN

802.3/Ethernet
10 Mb/s LAN

Bridge

IDF
Concentrator

Workstation

Workstation
HDF
Minicomputer

HDF

Concentrator

Workstation

Concentrator
Workstation

Mainframe

HDF

File
server

Concentrator

Bridge

802.3/Ethernet
10 Mb/s LAN

HDF = Horizontal distribution frame
IDF = Intermediate distribution frame
LKG–3813–89I

Additional Configurations

B–5

B.3 Dual Ring Configurations
A dual ring of concentrators allows multiple devices to be attached to the dual ring
through the concentrators. Figure B–5 shows a dual ring of concentrators installed in
a building. The concentrators can be located in the MDF of a building, IDF of a building, or distributed throughout the building in HDFs.
Figure B–5: Dual Ring of Concentrators Installed in a Building

VAX

Workstation

Concentrator
Workstation
Minicomputer

100 Mb/s FDDI ring

Workstation

Concentrator

Workstation

Mainframe

File
server

Concentrator

VAX

Workstation
LKG–3812–89I

B–6

Fiber Distributed Data Interface System Level Description

Figure B–6 shows a dual ring of concentrators and a DAS multiport bridge installed
in a campus. Concentrators or bridges in the dual ring are located in the IDF of each
building.
Figure B–6: Dual Ring of Concentrators and Bridges Installed in a Campus

Building 1

Building 2
(HDF)

(HDF)
B

B
Concentrator

Concentrator
Server

(IDF)

Building 3

Server

(IDF)

Building 4

(MDF)
(IDF)

(HDF)

(IDF)

Server
Concentrator
B

Bridge

B
B

HDF = Horizontal distribution frame
MDF = Main distribution frame

IDF = Intermediate distribution frame
B = Bridge
LKG–3814–90I

Additional Configurations

B–7

B.4 FDDI Dual Homing
Figure B–7 shows three concentrators installed in a dual homing configuration. In
this configuration, if concentrator 2 or the primary link to concentrator 3 fails, the
backup link to concentrator 3, through concentrator 1, is activated. This ensures that
the devices connected to concentrator 3 will have uninterrupted service. Activation
of the individual links follows the FDDI connection rules, described in Section 2.7.2.
Figure B–7: Concentrators Installed for Dual Homing

PHY A

PHY B

PHY A

Concentrator 2

Concentrator 1
PHY M

Backup
link

PHY B

PHY M

PHY A

PHY B

Concentrator 3
PHY M

FDDI
dual ring

PHY M

Primary
link

PHY M
LKG-4813-91I

B–8

Fiber Distributed Data Interface System Level Description

Figure B–8 shows a DAS bridge installed in a dual homing configuration. In this
configuration, if concentrator 2 or the primary link to the bridge fails, the backup link
to the bridge, through concentrator 1, is activated. This ensures that the devices on
the LANs connected to the bridge will have uninterrupted service. Activation of the
individual links follows the FDDI connection rules, described in Section 2.7.2.
Figure B–8: DAS Bridge Installed for Dual Homing

PHY A

PHY B

PHY A

Concentrator 2

Concentrator 1
PHY M

Backup
link

PHY B

PHY M

PHY A

PHY B

DAS Bridge

FDDI
dual ring

PHY M

Primary
link

LKG-4935-91I

Additional Configurations

B–9

B.5 Single-Mode Fiber Configurations
Figure B–9 shows a SAS bridge connected to a dual ring through a long distance
single-mode fiber cable. This is the recommended method to connect remote sites
through FDDI.
Figure B–9: Connecting to the Dual Ring Through a Single-Mode Link

PHY A

Concentrator

PHY B

Link length from
2 km (1.2 mi) to
40 km (24 mi)

SAS bridge

LKG–5101–91I

B–10

Fiber Distributed Data Interface System Level Description

Figure B–10 shows a dual ring which requires some links to be longer than 2 km (1.2
mi). In this configuration, the maximum total length of all links must not exceed 100
km (60 mi). DAS connections count as twice the link length.
Figure B–10: Dual Ring with Mixed Link Lengths

< 2 km

Bridge

Bridge
B

> 2 km
single-mode
fiber

< 2 km
B

Concentrator

Concentrator
A

< 2 km

< 2 km
B

A
Bridge

Concentrator
A

> 2 km
single-mode
fiber
LKG–5000–91I

Additional Configurations

B–11

C
Related Documents

The following documents provide additional information concerning FDDI hardware and software products offered by Digital.

•

DECconcentrator 500 Installation (Order No. EK-DEFCN-IN)

•

DECconcentrator 500 Problem Solving (Order No. EK-DEFCN-PS)

•

DECbridge 500/600 Series Installation and Upgrade
(Order No. EK-DEFEB-IN)

•

DECbridge 500/600 Series Problem Solving (Order No. EK-DEFEB-PS)

•

DECelms Installation (Order No. AA-PAK1A-TE)

•

DECelms Use (Order No. AA-PAK2A-TE)

•

DECelms Reference (Order No. AA-PBWBA-TE)

For information concerning planning and configuration of fiber networks, consult
the following documentation.

•

DECconnect System Fiber Optic Planning and Configuration
(Order No. EK-DECSY-FP)

•

DECconnect System Fiber Optic Installation (Order No. EK-DECSY-FI)

C–1

The following documents provide information concerning other Digital network offerings.

•

Networks and Communications Product Documentation
(Order No. EK-NACPD-RE)

•

Telecommunications and Networks Buyer’s Guide

•

Bridge and Extended LAN Reference (Order No. EK-DEBAM-HR)

For additional information concerning Digital products that can connect to an FDDI
network and for information on miscellaneous networking topics, refer to the following documents.

C–2

•

VAX and VAX Workstation document set for hardware platform information

•

VAX and DECwindows document set for software platform information

•

Guide to DECnet-VAX Networking for DECnet node information

•

Introduction to Network Performance for network performance information

•

Network Troubleshooting for network problem-solving information

Fiber Distributed Data Interface System Level Description

Glossary

attenuation

The difference between transmitted and received power. Expressed in units of decibels (dB).
bridge

A connecting device between local area networks in an extended LAN architecture.
A bridge acts as a link between similar and dissimilar LANs permitting the creation
of extended LANs.
bypass

The ability of a station to be electronically or optically isolated from the network
while maintaining the integrity of the ring.
code bit

The smallest signaling element used by the Physical layer for transmission on the
fiber cable.
code group

The specific sequence of five code bits representing a PHY symbol.
concentrator

An FDDI node that provides additional attachment points for stations that are not attached directly to the dual ring, or for other concentrators in a tree structure. A concentrator has two or more Physical Layer entities and may have zero or more Data
Link layer entities. The concentrator is the focal point of Digital’s dual ring of trees
topology.
Connection Management (CMT)

That portion of the Station Management function that controls network insertion, removal, and connection of the PHY and MAC entities within a station.

Glossary–1

Dual Attachment Concentrator (DAC)

A concentrator that offers two connections to the FDDI network capable of accommodating the FDDI dual (counter-rotating) ring, and additional ports for connection
of other CONs or stations.
Dual Attachment Station (DAS)

An FDDI station that offers two connections to the FDDI network capable of accommodating the FDDI dual (counter-rotating) ring.
entity

A manageable active element within an Open Systems Interconnection (OSI) layer
or sublayer.
fiber

A dielectric waveguide that guides light.
Fiber Distributed Data Interface (FDDI)

A set of ANSI/ISO standards that, when taken together, define a 100 Mb/s, timed-token protocol, local area network that uses fiber optic cable as the transmission medium. The standards define Physical Layer Medium Dependent, Physical Layer,
Media Access Control, and Station Management entities.
fiber optics

A technology whereby signals are transmitted over an optical waveguide medium
through the use of light-emitting transmitters and light-detecting receivers.
fiber optic cable

A jacketed fiber or fibers.
frame

A MAC Protocol Data Unit (PDU) transmitted between MAC entities on a ring. A
frame consists of a variable number of bytes up to a maximum of 4500, including 2
bytes of preamble.
Horizontal Distribution Frame (HDF)

Located on the floor of a building. Consists of the active, passive, and support components that provide the connection between the building backbone cabling and the
horizontal wiring.
Intermediate Distribution Frame (IDF)

Located in an equipment room. Consists of the active, passive, and support components that provide the connection between interbuilding cabling and the intrabuilding cabling for a building.

Glossary–2

Main Distribution Frame (MDF)

Located in an equipment room. Consists of the active, passive, and support components that provide the connection of the interbuilding backbone cables between
IDFs.
Media Access Control (MAC)

The Data Link layer sublayer responsible for scheduling and routing data transmissions on a shared medium local area network (for example, FDDI).
Media Interface Connector (MIC)

An optical fiber connector pair which links the fiber media to the FDDI node or
another cable. The MIC consists of two halves, the plug and the receptacle. The MIC
plug terminates an optical fiber cable. The MIC receptacle is associated with the
FDDI node.
MIC plug

That portion of the Media Interface Connector which terminates an optical fiber
cable.
MIC receptacle

The fixed portion of the Media Interface Connector which is part of an FDDI node.
node

A generic term applying to any FDDI network attachment device (single attachment
station, dual attachment station, or concentrator).
Non-Return-to-Zero (NRZ)

A data representation for binary data in which a polarity level high, or low, represents
a logical 1 or 0 respectively.
Non-return-to-Zero-Invert on Ones (NRZI)

A data transmission technique in which a polarity transition from low to high, or high
to low, represents a logical 1. The absence of a polarity transition represents a 0.
Numerical Aperture (NA)

The sine of the radiation or acceptance half angle of an optical fiber multiplied by the
refractive index of the material in contact with the exit or entrance face of a fiber.
optical receiver

An optoelectronic circuit that converts an incoming optical signal to an electrical signal.
optical reference plane

The plane that defines the optical boundary between the MIC plug and the MIC receptacle.

Glossary–3

optical transmitter

An optoelectronic circuit that converts an electrical signal to an optical signal.
physical connection

The full-duplex physical layer association between adjacent PHYs in an FDDI ring.
Physical Layer Medium Dependent (PMD)

The Physical layer sublayer that defines the physical requirements for nodes that attach to the FDDI ring. Items defined by the PMD include transmit and receive power
levels, connector requirements, and fiber optic cable requirements.
Physical Layer Protocol (PHY)

The Physical layer sublayer that defines the media-independent portion of the Physical layer in FDDI. Items defined by the PHY include symbols, line states, data encode/decode process, and the data-recovery process.
physical link

The path, via PMD and attached cabling, from the transmit logic of one PHY to the
receive logic of an adjacent PHY in an FDDI ring. Part of the physical connection.
primitive

A basic element of the services provided by one entity to another.
Protocol Data Unit (PDU)

Information delivered as a unit between peer entities that may contain control information, address information, and data.
receive

The act of a station acquiring a frame, token, or control sequence from the ring.
repeat

The act of a station in receiving a code-bit stream (frame or token) from an upstream
station and placing it onto the physical link connected to its downstream neighbor.
The repeating station may examine, copy to a buffer, or modify frame status bits in
the code-bit stream as appropriate.
repeater

An FDDI node that minimally comprises the functionality of two PMDs and provides only a repeat function. A repeater does not have any MACs or concentrator
functionality.
ring

Two or more stations wherein information is passed sequentially between active stations, each station in turn examining or copying the information, finally returning it
to the originating station.

Glossary–4

services

The services provided by one entity to a higher entity or to Station Management
(SMT).
Single Attachment Concentrator (SAC)

A concentrator that offers one attachment to the FDDI network, and extra ports for
the attachment of stations or other CONs.
Single Attachment Station (SAS)

An FDDI station that uses one PHY-port of type S to connect to the FDDI network.
station

An addressable node on an FDDI ring capable of transmitting, receiving, and repeating data. A station has exactly one instance of SMT and at least one instance of MAC,
PHY, and PMD.
Station Management (SMT)

The entity within a station on the FDDI ring that monitors station activity and exercises overall appropriate control of station activity.
symbol

The smallest signaling element used by the MAC sublayer. The symbol set consists
of sixteen data symbols and eight control symbols. Each symbol corresponds to a
specific sequence of code bits (code group) to be transmitted by the Physical layer.
transmit

The act of a station that consists of generating a frame, token, or control sequence,
and placing it on the ring for receipt by the next (downstream) station.
wallbox

Connects workstation wiring to horizontal wiring. Also called faceplate, information outlet, or telecommunications outlet.
window

As applied to optics, an optical wavelength region of relatively high transmittance,
surrounded by regions of low transmittance. For example, the first window is at 850
nm, the second at 1300 nm. Also called Spectral Window or Transmission Window.

Glossary–5

Index
Numbers
4B/5B encoding, 2–12

A
Active Line State, 2–12
Adapters, 1–29, 5–1
configuration guidelines, 5–5
description, 5–1
distance requirements, 5–6
FDDI port, 5–3
functionality, 5–1
host interface, 5–2
model, 5–2
operation, 5–3
packet memory, 5–3
planning, 5–5
Address format
48–bit, 2–16
universal, 2–17
Alternate fiber specifications, 2–4
American National Standards Institute
role in FDDI standards, 1–1
X3, 1–4
X3.139 MAC standard, 1–8
X3.148 PHY standard, 1–7

X3.166 PMD standard, 1–7
X3.T9.5 Station management, 1–8
X3T9, 1–4
X3T9.5, 1–4
Application, environments
backbone, 1–12
workgroup, 1–12
Application layer, layer, A–4
Asynchronous transmission, 2–30

B
Backbone, 1–13
Beacon process
description, 2–37
directed beacon, 2–28
function, 2–37
stuck beacon, 2–28
Bridges, 4–1
configuration
example, B–3, B–7, B–9
guidelines, 4–10
definition, 4–1
description, 4–2
distance requirements, 4–11
dual homing, 1–23, B–9
encapsulation, 4–4

Index–1

FDDI-to-Ethernet
DAS devices, 4–1
SAS devices, 4–1
filtering, 4–5
forwarding
database, 4–5
frames, 4–5
fragmentation, 4–7
function, 1–29, 4–4
multiport, 1–29, B–3, B–5, B–9
network management, 4–6
operation, 4–7
performance, 4–9
planning considerations, 4–10
queue manager, 4–7
source address tracking, 4–5
spanning tree, 4–6
stripping algorithm, 2–40
translation
process, 4–4
processor, 4–8

C
Canonical addressing
description, 2–18
fields, 2–18
transmission rules, 2–19
Claim process, 2–32
Clock recovery
description, 2–13
frequency tolerance, 2–14
Code
bits, 2–8
group, 2–8
Concentrator
architecture, 3–3, 3–4

Index–2

configuration guidelines, 3–7
connection rules, 3–9
definition, 3–1
dual
homing, 1–22, B–8
ring of trees, 3–3
dual attachment concentrator, 1–9, 3–2
FDDI device, 1–2
function, 1–9, 1–11, 3–1, 3–5
in-band management, 3–5
network management, 3–6
operation, 3–7
out-of-band management, 3–5
Physical Connection Management,
3–6
planning, 3–7
single attachment concentrator, 1–9,
3–2
standalone topology, 3–2
station bypass, 3–6
structured wiring, 3–10
topologies, 1–29
types
dual attachment concentrator, 3–2
single attachment concentrator, 3–2
with a MAC, 3–5
without a MAC, 3–5
Configuration
campus backbone, B–4
dedicated network, B–3
dual homing, 1–22, 1–23, 2–27, B–8,
B–9
logical star, B–4
multiport bridge, B–7
standalone network, B–1
workgroup LAN, B–5
Configuration Management, 2–24
Connection Management

Configuration Management, 2–24
Entity Coordination Management,
2–24
function, 2–23
Physical Connection Management,
2–23
Connection rules
bridge, 4–10
concentrator, 3–9
PHY port types, 2–25
Connector keying, 2–7
Control indicators, 2–33
Counter-rotating ring, 1–18

connection rules, 1–23
model, 1–11
optical bypass relay, 1–10
Dual homing
configuration
guidelines, 2–27
with concentrators, 1–22, B–8
with DAS bridge, 1–23, B–9
description, 1–22
Dual ring of trees, 1–1, 1–2, 1–20
topology of choice, 1–23
Dual-window fiber, 2–4

E
D
DAC. See Concentrator
DAS. See Dual attachment station
Data Link layer, 1–8, A–3
Data recovery, 2–8
DECbridge 500/600 series, 1–29, 4–1,
4–10
DECconcentrator 500, 1–29, 3–1, 3–7
Destination address, 2–22
Device, definition, 1–2
Directed beacon, 2–28
Documents
Digital FDDI products, C–1
networks, C–1, C–2
Dual attachment station, 1–2, 1–9
bridges
configuration rules, 4–10
dual homing, 1–23, B–9
concentrator
components, 3–2
configuration rules, 3–9
dual homing, 1–22, B–8
FDDI connection, 1–9

Elasticity buffer
buffer size, 2–14
description, 2–13
function, 2–8
use of idle symbols, 2–14
Encapsulation, 4–4
Encode/decode process, 2–8
Ending delimiter, 2–22
Enterprise Management, 6–1
Entity Coordination Management, 2–24

F
FCIS. See Frame Content Independent
Stripping
FDDI. See FDDI port types; Fiber Distributed Data Interface
FDDI port types
PHY A, 2–25
PHY B, 2–25
PHY M, 2–25
PHY S, 2–25
Fiber Distributed Data Interface

Index–3

definition, 1–2
planning and configuration, 1–31
recommended fiber counts, 1–32
specifications, 1–1, 1–30, 2–4, 2–5
Fiber specifications, 2–4, 2–5
Fragmentation, 4–7, 6–5
Frame
check sequence, 2–22
control field
address length bit, 2–22
class bit, 2–22
control bits, 2–22
format bits, 2–22
description, 2–21
ending delimiter, 2–22
fields, 2–21
format, 2–21
fragments
creation of, 2–35
stripping, 2–40
preamble, 2–21
status field, 2–23
control indicators, 2–33
stripping
algorithm, 2–40
fragments, 2–40
process, 2–40
transmitting station, 2–35
void frames, 2–40
types, 2–19
void, 2–40
Frame Content Independent Stripping,
2–40

H
Halt Line State, 2–11

Index–4

Horizontal distribution frame, 1–20,
3–10

I
Idle Line State, 2–11
Idle symbols
origin of, 2–14
use in
clock recovery, 2–14
scrubbing, 2–15
trace function, 2–38
IEEE 802.2, 1–8
IEEE 802.3, 1–8
IEEE 802.4, 1–8
IEEE 802.5, 1–8, 1–11
IEEE Address Administrator, 2–17
In-band management, 3–5
INFO field, 2–22
Interbuilding backbone, 3–10
Intermediate distribution frame, 1–20,
3–10
International Organization for Standardization
FDDI standards, 1–4
Open Systems Interconnection model,
1–8, A–1
Internet Protocol fragmentation, 4–7,
6–5
Intrabuilding backbone, 3–10

L
LAN management, 1–30
Line states, 2–8, 2–11
Active Line State, 2–12
Halt Line State, 2–11
Idle Line State, 2–11

Master Line State, 2–11
Noise Line State, 2–12
Quiet Line State, 2–11
Link Confidence Test, 2–31
Logical Link Control, 1–6, 1–8, 2–29,
2–30

M
MAC. See Media Access Control
Main distribution frame, 1–20, 3–10
Maintainability, 1–28
Master Line State, 2–11, 2–38
Media Access Control
data link addressing, 2–16
FDDI standard, 1–1
functions, 2–15
ring scheduling, 2–16
universal addressing, 2–17
Media Interface Connector
description, 2–7
keying, 2–7
MIC A, 2–7
MIC B, 2–7
MIC M, 2–7
MIC S, 2–7
Multimode fiber, specifications, 2–4

N
Neighborhood Information Frame
function, 2–28, 6–7
uses, 6–7
Network connection model, 1–15, 2–2
Network layer, A–3
Network Management, 6–1
accounting management, 6–3

address filtering, 6–6
benefits, 6–2
bridge port
configuration, 6–6
enabling, 6–6
capabilities, 6–4
configuration
management, 6–3
parameters, 6–5
configuration management, 6–3
device
naming, 6–4
registry, 6–4
Enterprise Management, 6–1
event
log, 6–5
message, 6–5
fault management, 6–3
firmware upgrade, 6–5
initialization, 6–6
IP fragmentation, 6–5
management functions, 6–2
neighborhood information frame display, 6–7
network monitoring, 6–7
performance management, 6–3
PHY port
configuration, 6–6
control, 6–6
protocol filtering, 6–7
ring mapping, 6–3, 6–7
security
management, 6–4
password, 6–5
spanning tree parameters, 6–5
status information frame display, 6–7
troubleshooting, 6–8

Index–5

NIF. See Neighborhood Information
Frame
No Owner frame
definition, 2–41
removal from ring, 2–41
Node, definition, 1–2
NOF. See No Owner frame
Noise Line State, 2–12
NRZ/NRZI encoding, 2–13

O
Open Systems Interconnection
function, 1–8
layers, A–3, A–4
model, A–2
Optical bypass relay, 1–10
description, 2–6
limitation of, 2–6
Organizationally Unique Identifier, 2–17
OSI. See Open Systems Interconnection
Out-of-band management, 3–5

Media Interface Connector, 2–7
optical bypass relay, 2–6
Physical Layer Protocol, 1–1, 2–7
data recovery, 2–8
elasticity buffer, 2–8
encode/decode process, 2–8
line states, 2–8
repeat filter, 2–8
smoothing, 2–8
symbols, 2–8
PMD. See Physical Layer Medium Dependent
Preamble, 2–21
Presentation layer, A–4
Protocol data units
definition, 2–19
formats, 2–19
logical link control units, 2–15, 2–30
removal of orphans, 2–14
role of scrubbing, 2–8
valid format, 2–21

Q
P
PHY. See Physical Layer Protocol
PHY ports. See FDDI port types
Physical Connection Management,
2–23, 2–36, 2–38
Physical layer, 1–8, A–3
Physical layer entities
description, 1–30
number, 1–31, 3–9, 4–10, 5–5
Physical Layer Medium Dependent,
1–1, 2–3
alternate fiber specifications, 2–4
connector keying, 2–7
fiber specifications, 2–4

Index–6

Quiet Line State, 2–11

R
Reliability, 1–27
Repeat filter, 2–8, 2–14
Ring
latency, 2–35
maintenance
beacon process, 2–37
function, 2–36
recovery, 2–37
trace, 2–38
operation, 2–30

purger
algorithm, 2–41
election process, 2–42
function, 2–41
No Owner frame removal, 2–41
overhead, 2–42
scrubbing, 2–14
Ring Management
description, 2–27
stuck beacon, 2–28

S
SAC. See Concentrator
SAS. See Single attachment station
Scalability, 1–28
Scrubbing, 2–8, 2–14
Service Access Points, 1–9
Session layer, A–4
SIF. See Station Information Frame
Single attachment station, 1–2
concentrator, 1–9
FDDI devices, 1–9
model, 1–10
Single-mode fiber
configurations, B–10, B–11
specifications, 2–5
Smoothing, 2–8, 2–14
SMT. See Station Management
Source
address, 2–22
routing, 2–17
Standalone topology, 1–16
Starting delimiter, 2–21
Station
definition, 1–2
power up, 2–31

Station bypass, 3–6
Station Information Frame
function, 2–28, 6–7
uses, 6–7
Station Management
architecture, 2–24
Connection Management, 2–23
connection rules, 2–25
FDDI standard, 1–1
frame
fields, 2–29
format, 2–29
services, 2–28
function, 2–23
PHY-port types, 2–25
protocols, 2–23
Ring Management, 2–27
Steady-state operation
capturing the token, 2–33
control indicators, 2–33
description, 2–33
Stripping. See Frame
Stuck beacon, 2–28
Symbol
code bits, 2–8
code group, 2–8
definition, 2–8
encoding, 2–10, 2–11
function, 2–8

T
T_Bid, 2–32
T_Neg, 2–33
Target Token Rotation Time
default value, 2–31
function, 2–31
MAC requirement, 2–16

Index–7

negotiation process, 2–31, 2–32
operational value, 2–31
THT. See Token Holding Timer
Timed Token Protocol, 2–16
Timers, 2–30
Target Token Rotation Timer, 2–31
Token Holding Timer, 2–30
Token Rotation Timer, 2–30
Valid Transmission Timer, 2–30, 2–36
Token
acquisition, 2–16, 2–20, 2–34
definition, 2–15
fields, 2–21
format, 2–20
function, 2–20
issuing, 2–16, 2–20
Token Holding Timer, 2–30
Token Rotation Timer, 2–30
Topologies, 1–14
Trace, 2–38
Translation, 4–4

Index–8

Transport layer, A–3
Tree of concentrators, 1–17
TRT. See Token Rotation Timer
TTRT. See Target Token Rotation Time
TVX. See Valid Transmission Timer

U
Universal address
format, 2–17
unique identifier, 2–17

V
Valid Transmission Timer, 2–30, 2–36
Void frames, 2–40

W
Wallbox, 3–10
Workgroup, 1–12, 1–14

HOW TO ORDER ADDITIONAL DOCUMENTATION
DIRECT TELEPHONE ORDERS
In Continental USA
and Puerto Rico
call 800–258–1710

In Canada
call 800–267–6146

In New Hampshire
Alaska or Hawaii
call 603–884–6660

DIRECT MAIL ORDERS (U.S. and Puerto Rico*)
DIGITAL EQUIPMENT CORPORATION
P.O. Box CS2008
Nashua, New Hampshire 03061

DIRECT MAIL ORDERS (Canada)
DIGITAL EQUIPMENT OF CANADA LTD.
940 Belfast Road
Ottawa, Ontario, Canada K1G 4C2
Attn: A&SG Business Manager

INTERNATIONAL
DIGITAL
EQUIPMENT CORPORATION
A&SG Business Manager
c/o Digital’s local subsidiary
or approved distributor

Internal orders should be placed through Publishing and Circulation Services (P&CS),
Digital Equipment Corporation, 10 Forbes Road, Northboro, Massachusetts 01532–2597
*Any prepaid order from Puerto Rico must be placed
with the Local Digital Subsidiary:
809–754–7575

Fiber Distributed Data Interface
System Level Description
EK–FDSLD–SD–002
READER’S COMMENTS
What do you think of this manual? Your comments and suggestions will help us to improve the quality
and usefulness of our publications.
Please rate this manual:
Poor
1
1
1
1
1

Accuracy
Readability
Examples
Organization
Completeness

2
2
2
2
2

3
3
3
3
3

4
4
4
4
4

Excellent
5
5
5
5
5

Did you find errors in this manual? If so, please specify the error(s) and page number(s).

General comments:

Suggestions for improvement:

Name

Date

Title

Department

Company
City

Street
State/Country

Zip Code

DO NOT CUT – FOLD HERE AND TAPE
NO POSTAGE
NECESSARY
IF MAILED
IN THE
UNITED STATES

BUSINESS REPLY LABEL
FIRST CLASS PERMIT NO. 33 MAYNARD MASS.
POSTAGE WILL BE PAID BY ADDRESSEE

Telecommunications and
Networks Publications
550 King Street
Littleton, MA 01460–1289
DO NOT CUT – FOLD HERE

Printed in U.S.A.