Home Search Collections Journals About Contact us My IOPscience

A comparison of DQDB and FDDI for the interconnection of LANs

This article has been downloaded from IOPscience. Please scroll down to see the full text article.

1994 Distrib. Syst. Engng. 1 127

(http://iopscience.iop.org/0967-1846/1/3/001)

View the table of contents for this issue, or go to the journal homepage for more

Download details:
IP Address: 202.164.38.38
The article was downloaded on 10/09/2010 at 11:45

Please note that terms and conditions apply.

Distrib. Syst. Engng i(1994)127-134

A comparison of DQDB and FDDl for the

interconnection of LANs

Riaz Ahmad and Fred Halsall

Communications Research Group, Department of Electrical and Electronic
Engineering, University College of Swansea, Swansea, SA2 8PP,UK
Received 5 November 1993

Abstract. This paper identifies the design issues that must be considered when
interconnecting IEEE 802 LANs through either DQDB or FDDI backbone
subnetworks using MAC bridges. A series of simulation results are presented that
compare the performance of both subnetwork types for various types of LAN traffic,
backbone subnetwork sizes (physical coverage and number of LANs) and other
performance-criticalparameters. The paper concludes with some guidelines to be
followedto achieve an optimal performance to meet specific application
requirements.

1. Introduction subnetwork [ 5 ] . The general scheme is shown in figure 1.

Bridges are being proposed also for interconnecting high-
The interconnection of LANs over a complete site, without speed LANs and backbones [6].This paper is concerned
any performance loss, is currently a major issue being with a detailed performance comparison of DQDB and
addressed by network designers. The emerging IEEE FDDI when each is being used as a backbone subnetwork.
DQDB [l] and ANSI FDDI [2] standards are the two In the next section, the essential features of both
candidate networks being considered to meet such subnetwork types and bridges are presented and this is
requirements. Among the alternative interconnection followed by a description of the issues relating’ to the
methods, MAC bridges are popular since they are interconnection of IEEE 802 LANs that must be
transparent to end stations and can provide a high addressed. This is followed by a set of performance results
throughput 131. Moreover, they are now being which have been obtained from a series of computer
standardized [4]. One of the prefkrred ways to interconnect simulations. The paper concludes with some guidelines to
LANs that are distributed over a site is by the connection be followed to achieve an optimum performance to meet
of each LAN, via a bridge, to a high-speed backbone specific application requirements.

Figure 1. A typical network comprising a high-speed backbone interconnecting

several heterogeneous LANs.

The British Computer Society,The institution of Electrical Engineers and IOP Publishing Ltd
0967-1846/94/030127+08$19.50@1994
R Ahmad and F Halsall

2. FDDI token rotation time (TRT). On receipt of the token, if a

station has asynchronous (data) traffic to send, it computes
The architecture of a fibre distributed data interface (FDDI) the difference in the time between the TTRT and the actual
subnetwork consists of two independent optical fibre rings. token rotation time (TRT); that is (TTRT - 'RT). The
These are referred to as the primary ring and the secondary difference is known as the token holding time (THT). If
ring, each carrying data in opposite directions at a rate of THT is positive, the station can transmit for this interval
100Mbps. The proposed standard specifies the maximum prior to releasing the token. As can be deduced from this,
fibre path length to be 200km with up to 500 physical the TTRT establishes a guaranteed maximum response time
connections. A typical FDDI network configuration is for the ring since, in the worst case, the time between the
shown in figure 2. As can be seen, there are two types of arrival of two successive tokens will never exceed twice the
station: a dual attachment station, which is connected to TTRT value.
both rings, and a single attachment station, which is Various performance studies of FDDI parameters have
attached only to the primary Mg. In practice, most stations been reported in [9, IO], where the performance of FDDI
are attached to the ring via a wiring concentrator. Further has been shown to be dependent on the TTRT, the size of
information on the FDDI standard can be found in [7,8]. the network (length of cable), the total number of stations
FDDI uses a timed token rotation protocol to control and the frame size. The TTRT, however, has been shown to
access to the medium. Each station measures the time that be the key parameter that can be used to optimize the
has elapsed since a token was last received. As part of the performance of network. As will be seen, unlike DQDB,
ring initialization prmess, all stations negotiate a target FDDI provides fair access for all users for any network size
token rotation time (TTRT). The asynchronous service and traffic load.
allows the use of a token only when the time since a token
was last received has not exceeded the established TTRT.
Each station on the ring measures the time since it last 3. DQDB
received the token. The time interval between two
successive receptions of the token by a station is called the The architecture of a distributed queue dual bus (DQDB)
subnetwork consists of two unidirectional buses as shown
in figure 3. Both buses, referred to as bus A and bus B,
cany data in opposite directions allowing full duplex
communication between any pair of stations connected to
the network. As both buses are operational at all times, the
total available bandwidth is up to twice the bandwidth of a
single bus. The operation of the two buses in the transfer of
data is independent. The head of each bus generates fixed-
length slots which are used to cany data between stations
on their respective buses. All data flow on both buses
terminates at the end of the bus. As specified in the IEEE
802.6 standard, a station cannot remove data and can only
alter it when permitted by the access protocol. The DQDB
layer provides two modes of access control to the dual
I / U -.\ I buses: prearbitrated (PA) and queued arbitrated (QA),
SecondaryRing which are used to provide isochronous and asynchronous
U services respectively. This paper quantifies the performance
of a single DQDB subnetwork for the interconnection of
Figure 2. FDDl network schematic layout. LANs carrying connectionless data only. The following
Slotwithdirectionof flow

Bus A
U--

Figure 3. DQDB dual bus architecture. AU, access unit: 8,busy bit; R, request bit.

128
 A comparison of DQDB and FDDI for the interconnection of LANs

&
Bus A

1
Count down for each
empty QA slot on bus A
. . _ . . . . _ . . . . . . . . .-..
-
. <.,

Transfer RQ count

Counter (RQ) counter (CO)

't Count new requests

on bus B

Bus E

Figure 4. DQDB access control using busy and request bits.

description, therefore. refers to the DQDB protocol for the 4. Transparent bridges
QA access mode only.
The operation of the access control protocol is based Bridges operate at the medium access control (MAC)
upon two control bits contained in the access control field sublayer and are transparent to end stations. A description
of the slot header: the BUSY bit and the REQUEST of the detailed operation of transparent bridges is given in
(REQ) bit. The BUSY bit indicates the slot status whether [3]. A bridge receives and buffers all frames in their
empty or occupied. The REQ bit is used to indicate when a entirety before performing the relaying (interconnection)
segment has been queued for transmission on the opposite function. During normal operation, the MAC address of
bus. Each station, by counting REQ bits from bus B and each end station is leamt by the bridge and retained in a
empty slots that pass on bus A, can determine the number routing table known as the forwarding database. After the
of segments that are ahead of it in the distributed queue for learning phase, when a frame is received at a bridge port,
bus A. The distributed queue uses two counters in each the forwarding database is searched to determine whether
station: a request (RQ) counter and a count down (CD) the destination address is present for this port. If so, the
counter as shown in figure 4. The RQ counter is used to frame is discarded, otherwise it is relayed to the appropriate
maintain a count of the number of stations downstream on port for forwarding. The bridges do not inspect or change
bus A which have requested access to bus A via REP bits the data carried by the MAC frame and end stations
on bus B. A station that has a segment to transmit on bus perceive the bridged network as a single, extended LAN.
A sets a REQ bit on bus B, and transfers the current value
of the RQ counter into the CD counter. The CD counter
thus indicates the number of segments which are queued 5. LAN interconnectionissues
by the downstream stations before the current request was
made and hence the number of empty slots the station The various IEEE 802 LAN standards differ at the physical
must allow to pass on bus A. For each empty slot allowed layer and medium access control (MAC) sublayer, but are
to pass downstream on bus A, the CD counter is compatible at the logical link control (LLC) sublayer [12].
decremented. When the CD counter becomes zero, the The relationship between FDDI, B E E 802.6 and the
station can send its segment in the next empty slot that different 802 LANs within the three lower layers of the
passes on bus A. Each station can place only one segment IS0 reference model is shown in figure 5. Both FDDI and
at a time in the distributed queue. A separate queue is IEEE 802.6 provide a connectionless MAC service to the
operated for each of the two buses, with separate counters LLC sublayer in a manner consistent with other 802 LANs.
at each station for each bus.
A performance study of the DQDB protocol reported in
[ l l ] shows that the basic access protocol has unfair
I Network Laver I
bandwidth sharing for users when there are heavy traffic
demands; especially at higher network speeds and larger
station separations. The proposed IEEE 802.6 standard
specifies a bandwidth balancing (BWB) mechanism to
ensure fair sharing of bandwidth between stations operating
at a single priority. The impact of the bandwidth balancing Figure 5. Position of IEEE 802 LANs, FDDl and IEEE 802.6
mechanism on network performance will be expanded upon in IS0 reference model. LLC, logical link control; MAC,
later. medium access control.

129
R Ahmad and F Halsall

This makes both network types suitable for use as given in the 802.6 standard is the generation of a bus
backbones for LAN intekonnection using MAC bridges. selection table for evely end station on the entire bridged
The maximum frame size specified in FDDI is 4500 bytes network. With this scheme, however, table maintenance can
and in 802.6,9188 bytes, both of which are greater than the become unacceptably large as the size of the bridged
maximum frame size range of 802.3 (CSMAKD) LANs network increases. In general, therefore, this is not used and
(1518bytes). Although the= is no specifiedframe size limit each frame is transmitted on both buses. All bridges then
in 802.5 (token ring) when interconnected via MAC reassemble all segments received back into their initial MAC
bridges the frame size with these LANs must not exceed PDU and, from thejr stored forwarding database, relay
the allowable limits of the backbones. Moreover, FDDI, appropriate frames onto their required LAN subnetwork. As
IEEE 802.6 and each of the 802 LANs use and can can be concluded from the foregoing, DQDB requires
recognise only unique frame formats specific to their MAC additional segmentation/reassembly overheads compared
protocol. Hence, in the same way that frames between with FDDI. Moreover, if both rings in FDDI are used, this
dissimilar 802 LANs cannot be relayed, LAN frames yields a more efficient utilization of transmission bandwidth.
cannot be transmitted directly onto either an FDDI or To enhance the performance of the DQDB, a scheme known
DQDB backbone. The bridge, therefore, must encapsulate a as destination release has been proposed 1131. A special bit,
LAN frame within a FDDI or DQDB MAC frame before called previous slot received (PSR),is defined in the slot
forwarding it onto the backbone. Then, prior to forwarding header which indicates whether or not the previous slot has
the frame onto the destination LAN subnetwork, the MAC been read. Some special access nodes, called eraser nodes,
frame must be decapsulated and retransmitted in the correct 'erase' all the slots that have been marked as 'read' so that
format. they can be reused downstream by other users. In practice,
In order to encapsulate a LAN frame within a FDDI however, because of the similar reasons meddoned before,
MAC frame, 27 bytes are added. For DQDB, at least 28 bytes bridges cannot set PSR bits. Hence, with bridges, the total
are added to form an initial MAC protocol data unit available bandwidth of a DQDB backbone remains equal to
(TMPDU). In FDDI, MAC frames are transmitted on the the transmission rate of a single bus.
medium without segmentation. In the w e of DQDB, each
JMPDU is segmented into fixed length (44byte) units.
6. Simulation environment
Control and message identifier fields are then added to each
segment prior to transmission in a 53-byte DQDB slot. The results presented in the following section have all been
These are used by the destination stations to reassemble derived using a commercial simulation packag-
segments into the original IMPDU. To avoid the unnecessary NETWORK II.5 [14]. For comparison purposes, both
use of both buses, the source bridge must know which of the FDDI and DQDB have been modelled for similar network
two buses to use for transmitting each frame. One suggestion configurations and similar LAN traffic. Both run at the

-0

.-0
90
80

-2
E 70

60 ... _ I _ .. 8. ..I . . . .'. .

8
In
50

-
m
12
m
40

30 ., . .. . _ I . . . . _... . . _._. . . . I
. . . .' . . . . . ,. .
L I ..
v)

s 20 .- .A.. .. . I . . .. _ I - . .. _I.. ~. . . . . - .- .,. . . .

L I _ I _ . ..

5l 10 .__.1 _ _ . _ ~ . _ . _ _
1.. .. _ . , . . . . . ~ . . . . , " .. ,~_ _ _ _ . , _ _ _ _ _
0
1 2 3 4 5 6 7 8 9 10
User location

t 1 0 k m +50km +100hm-+r200km

Figure 6. Unfair behaviour of DQDB for various stations at differentpositions on the bus.

130
 A comparison of DQDB and FDDl for the interconnection of LANs

0 1
1 2 3 4 5 6 7 8 9 10
Node Number

-+Unbalanced + 6W6-M =8 BWB-M = 4

Figure 7. impact of bandwidth balancing mechanism on the fairness of DQDB.

0 10 20 30 40 50 60 70 80 90 100
Total Load (Mbps)

I DQDB(BWB-M = 6 ) +FDDlmRT = 4 ms) X FDDI(TTRT= 8 mS) I

Figure 8. Performance comparison of DQDB and FDDl subnetworks (framesize = 64 bytes, network
size= 100 km).

transmission rate of 100Mbps. In DQDB, each frame is balancing modulus (BWB - M) values for DQDB vary
transmitted on both buses and, to keep the same bandwidth, from 4-8. Each backbone contains 10 bridges and each
a single FDDI ring is used. The physical size of the bridge connects two similar LANs to the backbone. The
backbone is varied from 1!un to 300 b. The TTRT values performance of each backbone type was measured for three
used for FDDI are 4 and 8 milliseconds. The bandwidth frame sizes: @bytes and 1218bytes (to represent

131
R Ahmad and F Halsall

0 10 20 30 40 50 60 70 80 90 100
Total Load (Mbps)

+DODB(BW8-M-6) -e- F D D I ( T R T = 4 m S ) +FDDI(lTRT=BmS)

Figure 9. Performance comparison of DQDB and FDDl subnetworks (framesize = 1218 bytes,
networksize = 100 km).

m
m

0 I
0 10 20 30 40 50 60 70 80 90 100
Total Load (Mbps)

+DQDB(BW5M = 6) +FDDl(rTRTr4 mS) X F D D I W T = 8 mS)

~ ~

Figure 10. Performance comparison of DQDB and FDDl subnetworks (fmme size = 4000 bytes,
network size = 100 km).

CSMA/CD frames) and 4000bytes (to represent token The performance measures are. the mean response time
ring.) The anival of frames at bridges from each of the of the backbone with varying offered load and the
LANs is exponentially distributed and the total traffic over maximum effective throughput for various backbone sizes
the backbones is uniformly distributed between the bridges. and varying traffic. The mean backbone response time is

132
 A comarison of DQDB and FDDI for the interconnection of LANs

defined as the delay between the frame arriving at the limit known as the BWB-modulus, M, the RQ counter is
source bridge and the corresponding destination bridge incremented by one thereby forcing the station to let a free
receiving the entire frame. For the comparison of the two slot pass. The impact of using a BWB - M = 4 and 8 for a
backbone subnetworks, the bridge processing overheads are 200km network is shown in figure 7. The network still
assumed to be negligible. The presented simulation results shows a level of unfairness with a BWB - M = 8 and
are after the steady state network conditions have been completely fair with a BWB - M = 4. From other
reached. simulation results it was found that a BWB - M = 6 for
l O O k m and 8 for smaller network sizes is suitable. Hence
in the rest of the simulation results presented, these BWB
- M values have been used for corresponding network
7. Simulation results sizes.
The performance comparison of FDDI and DQDB
As indicated earlier, a feature of a DQDB subnetwork is backbones for various frame sizes is presented in figures
that, under heavy load conditions, the access delay of 8-10. The backbone size is l O O k m and the total offered
stations varies depending on their position on the bus. This load varies from 1 to 100Mbps. A frame size of 64bytes
is shown in figure 6. Station numbers indicate their relative has been used for the results in figure 8. The effective
position on the bus; station 1 is close to the head of bus A throughput of a FDDI backbone is 60.5 Mbps with a TTRT
and station 10 is close to the bead of bus B. To simulate = 4 and 65.5 Mbps with a lTRT = 8. The DQDB backbone
heavy load conditions, all stations always have a segment gives a throughput of only 39.6Mbps with this frame size.
ready to transmit on bus A. As seen from the graphs the The MAC frame encapsulation overheads are responsible
stations located in the middle experience longer access for the low throughputs in both types of backbone. The
delays. Also the extent of unfairness increases with segmentation overheads in DQDB further reduce the
increasing network size. Stations located next to the bus effective throughput. After encapsulation, the 64byte MAC
heads always show a better performance. Such unfairness is frame size is increased to 92 bytes to form an IMPDU with
unacceptable for backbone networks and it is for this DQDB which, in turn, requires 3 slots. For larger frames,
reason that the bandwidth balancing (BWB) mechanism the impact of encapsulation overheads is relatively small as
has been introduced. shown in figures 9 and 10. These are for 1518byte and
To implement the BWB mechanism, each station has a 4000 byte frame sizes respectively. Similar results have
BWB counter which is incremented after every also been reported in 1151 for a different traffic model. For
transmission. Then, whenever the BWB counter reaches a both FDDI and DQDB, the backbone response times are

0 20 40 60 80 100 120 140 160 180 200

Backbone Size (km)

-6- 64 (byte) + 1518 (byte) -e,- 4000 (byte)

Figure 11. Maximum throughput of DQDB and FDDI subnetworks at various frame and network
sizes.

133
R Ahmad and F Halsall

relatively low. The DQDB has relatively lower delays References

within its effective bandwidth range compared to FDDI.
Using a " R T = 8 , the FDDI throughput is further [I] Draft of Proposed IEEE Standard 802.6 Distributed Queue
increased compared with a TTRT = 4. Dual Buses (DQDB) Subnetwork of a Metropolitan Area
Network (MAN) 1990Media Access Control (MACJand
The maximum throughput obtained for different frame Physical Luyer Protocol Documents Draft P802.6D12
sizes with varying backbone lengths is shown in figure 11. [2] American National Standard 1989 FDDI Token-Ring Media
As can be seen, for a 1km FDDI backbone (TTRT = 4), 99 Access Control (MAC) ANSI X3.139-1989
Mbps throughput can be obtained with a 4000 byte frame [3] Backes F 1988 Transparent bridges for interconnection of
size. Also, the throughput of FDDI decreases with longer lEEE SO2 LANs IEEE Network 2 (1) 5-9
[4] Draft of Proposed IEEE standard 802.1 MAC Bridges 1989
distances whilst there is almost no change 'Uthe throughput DraftP802.1 dD8
with DQDB. [51 Ahmad R and Halsall F 1992 Performance analysis of
bridged LANs IEE Proc. E 139 64-72
[6] Ahmad R and Halsall F 1993 Interconnecting high-speed
LANs and backbones IEEE Network 7 ( 5 )36.43
[q Ross F E and Hamstra I R 1993 Forging FDDI IEEE J.
Selected Areus Commun. 11 181-90
8. Conclusions 181 Jain R 1993 FDDI: Current issues and future plans IEEE
Commun.Mug.31(31) 98-104
This paper has presented simulation results relating to the 191 Dykeman D and Bux W 1988 Analysis and tuning of the
use of FDDI and DQDB networks for the interconnection FDDI media access control protocol IEEE J. Selected
Areas Commun. 6 997-1010
of LANs. Both networks have been used as backbone .1101.Jain R 1991 Performance analvsis of FDDI token rinr
subnetworks and the effect on their performance of v e i n g networks: effect of paramete& and guidelines for setting
operational parameters investigated. In the case of DQDB, TTRT IEEE LTS 2 (2)16-22
the unfair access delay of stations in relation to their [ l l ] Hahne E L, Choudhuy A K and Maxemchuk N F 1992
position on the bus has been identified and the effect of the
DODB networks with and without bandwidth balancinr
IEEE Trans. Commun. 40 1192-204
-
use of a bandwidth balancing mechanism demonstrated. It 1121
. .Halsall F 1992 Data Communications, Comouter Networks
may be concluded from the throughput delay graphs that and Open Systems 3rd edn (Reading, 'MA Addison-
both networks give a low access delay up to their Wesley)
throughput limits. However the FDDI backbone gives a [13] Zukerman M and Potter P 1990 A protocol for eraser node
implementation within the DQDB framework Proc. IEEE
higher maximum throughput for all frame sizes providing GLOBECOM' 90 Vo13 (New York: IEEE) pp 1400
the network size is less than 140km. This is particularly the [14] Network D.5 1991 User's Manual, Version 7.0 CACI
case with the lowest frame size because of the larger Producs Company, La Jolla, California
segmentation overheads with DQDB. For distances greater 1151 Wainwright N and Myles A 1989 A comparison of FDDI
Fiber Optic Network with the Engineering IEEE 802.6
than 140!un, then DQDB gives a better throughput for the Metropolitan Area. Network IEE Coll. on Fiber Optic
larger frame sizes which demonstrates its appeal for LANs and Techniques for the Local Loop (London)
metropolitan area networks. (London: IEE) Digest No. 4

134