Philip G. Potter

The DQDB protocol and its performance under overload traffic conditions

Abstract The Distributed Queue Dual Bus (DQDB) protocol is being considered by the IEEE 802.6 and ANSI T1S1.1 as an important component in their Metropolitan Area Networks (MAN) and Broadband ISDN User Network Interface (BI-SDN-UNI) standards respectively. This paper provides first a description of DQDB focusing mainly on its distributed queue operation. Then, the performance of DQDB under traffic saturation for the cases of concentrated and widely separated sources is studied and discussed with a special focus on the effect of the elimination of the STANDBY state on fairness and throughput. It is shown that the new version of DQDB, without the STANDBY state, provides better fairness than the previous one which include the STANDBY state.

A protocol for eraser node implementation within the DQDB framework

The Distributed Queue Dual Bus protocol (DQDB) has been specified by the IEEE 802.6 committee as part of their proposed standard for metropolitan area networks. It is pointed out that, by allowing erasing capabilities to some stations, the transmission capacity can be considerably increased. A protocol for eraser nodes operation in DQDB based on the erasure of a request for each erased slot is presented. Simulation results show that a DQDB subnetwork with 10 nodes (three of which are erasers) and with a total throughput of 2.556 times the bus speed performs similarly to corresponding DQDB subnetworks that do not have erasers and have a throughput of 0.8 of the bus speed.<<ETX>>

Analysis of a discrete multipriority queueing system involving a central shared processor serving many local queues

The authors consider a multipriority queuing system which involves several distributed local queues and a central server which performs round-robin processor sharing (RRPS). This queuing model, referred to as multiqueue processor sharing (MPS), is an extension of Kleinrocks classical processor sharing model with generalization to multiple priorities and the addition of a buffer at each traffic source for each priority to buffer additional packets queued at that source. The authors present an exact analysis of the MPS model which leads to a closed-form result for the mean packet delay as a function of the packet length for each priority. It is demonstrated by simulation that the result obtained is applicable for accurately computing the average packet delay for a DQDB (distributed queue dual bus) subnetwork, given that stations transmitting on the same bus are not widely separated. >

A discrete shared processor model for DQDB

Abstract The Distributed Queue Dual Bus (DQDB) protocol is being considered by the IEEE 802.6 and ANSI T1S1.1 as an important component in their Metropolitan Area Networks (MAN) and Broadband ISDN User Network Interface (BISDN-UNI) standards respectively. The DQDB protocol is modelled in this paper by a discrete time multi-priority multiqueue system involving several distributed stations and a central shared processor. Exact analysis for the model yields an expression for the mean packet delay as a function of the packet size for each priority. Comparison with simulation results demonstrates that the model can accurately describe DQDB where transmitting stations on the same bus are concentrated.

A generic flow control protocol for B-ISDN

A generic flow control (GFC) protocol is proposed to regulate multiple terminals within the broadband-ISDN customer premises network. The GFC protocol is based on multipriority distributed queuing preceded by a traffic shaping function, and meets the requirements for flexible allocation of service parameters, control of jitter and guarantee of bandwidth. This GFC protocol provides the necessary flexibility to match the constant bit rate (CBR) jitter performance and variable bit rate (VBR) service requirements to the particular service type. It integrates well with usage parameter control without imposing excessive burstiness on each virtual connection cell stream delivered to the network. By allowing connectionless users to access at the lowest priority, only when no higher priority users are queued, there is no reduction in guaranteed capacity available for other VBR and CBR users.<<ETX>>

A shared medium multi-service protocol

A new asynchronous transfer mode (ATM) shared medium protocol which guarantees capacity for constant bit rate (CBR) and variable bit rate (VBR) services is described. The protocol includes a traffic shaping function, multiple prioritized access, a buffer insertion mechanism, destination release and distributed queuing for lower priority traffic. The protocol also allows for mixed priority mode for a connection to provide efficient and fair access for VBR connections requiring a guaranteed rate lower than the peak. It is demonstrated by simulation that the protocol can guarantee capacity for CBR and VBR connections, with minimal cell clumping.<<ETX>>

DQDB performance under sustained overload with bandwidth balancing and multiple requests outstanding

The performance of an enhanced DQDB protocol with Bandwidth Balancing (BWB) and Multiple Requests Outstanding (MRO) under sustained overload is examined by means of simulation. It is demonstrated that MRO may provide a desired bandwidth apportionment without using BWB, for a very specific case of concentrated sources; otherwise. MRO cannot be used without BWB to obtain the desired bandwidth allocation. When used together with, and in the same ratio as, the BWB, MRO speeds up the convergence towards the steady state bandwidth apportionment intended by BWB.

Fairness under sustained overload in a DQDB subnetwork with eraser nodes

Fairness criteria are proposed for the apportionment of throughput among the different stations for a DQDB (distributed-queue dual-bus) subnetwork involving eraser nodes under sustained overload. These criteria are based on maximizing the minimum throughput available to any station. Idealistic values of capacity apportionment for each active station are analytically computed by a recursive procedure. These idealistic values provide a quantitative standard against which the fairness of proposed eraser node protocols may be assessed. A numerical example is used to provide a detailed demonstration of the method for capacity apportionment.<<ETX>>

Broadband network development in Australia

The authors describe the development of the Australian public telecommunications network in preparation for the introduction of broadband services. They consider the application of metropolitan area networks (MANs) and in particular the IEEE 802.6 standard proposals based on the queued packet and synchronous switch (QPSX). The transition from these solutions to a B-ISDN (broadband integrated services digital network) is discussed in detail, with regard to possible use of synchronous transmission networks of B-ISDN transport and compatibility with existing terminals and network products. The longer term implications of service integration based on asynchronous transfer modes are also addressed, with no particular emphasis on the ability of such techniques to offer highly flexible network infrastructures for resource allocation and management. They note that Australia foresees widespread application of the higher rate B-ISDN interface around 600 Mb/s to allow customers full broadband support.<<ETX>>