Wlodek Dobosiewicz

Classification of fairness measures for local and metropolitan area networks

Abstract Fairness of very high-speed networks became a focus of interest as soon as the Distributed Queue Dual Bus (DQDB) gained popularity. Several papers exposed its unfairness; at the same time, many researchers argued that this unfairness is purely hypothetical. In our opinion, this controversy resulted from the absence of any systematic approach to what fairness actually is. We introduce a systematic classification of fairness measures in local and metropolitan area networks. A ‘points of view’ approach to fairness is proposed: depending on the point of view, the observers perception of fairness may differ significantly. This phenomenon is illustrated with the example of a simple (and apparently unfair) protocol.

On a MAC Protocol Based on Distributed Cycles

This paper introduces a novel transmission scheduling scheme applicable to high-speed ring networks. The essence of the proposed scheme is a special marker, called the terminator, circulating in the ring. The maximum normalised throughput, of our protocol does not degrade with the increasing propagation length of the ring and the stations endure small access delays under light load. Our protocol also offers a full-scale isochronous service without any need for fixed reservations, negotiations, or other preparatory stages.

Towards a gigabit FDDI

A possible network architecture suitable for a next-generation follow-up to FDDI Is presented. The solution can be scaled-up to arbitrary transmission rates (limited only by the technological barriers) and arbitrarily long networks. The protocol allows every station to transmit when the ring is silent, even when the station is not in possession of the token. The resulting token-passing protocol yields a capacity-1 network irrespective of the value of TTRT. Additionally, stations endure a very small access delay under light load. The need to single out synchronous traffic is eliminated, guaranteeing each station the right to transmit at fixed intervals. This modification allows graceful station insertion and deletion; moreover, the recovery of a lost token can be done on the fly without disrupting the operation of the network.<<ETX>>

Fault-tolerant capacity-1 protocol for very fast local networks

A substantial amount of attention has been paid recently to DQDB--a proposed bus architecture and MAC-level protocol for fast local and metropolitan area networks. The main advantage of this solution over previous concepts is in the fact that the performance of DQDB does not degrade with the increasing value of a--the ratio of the packet length to the propagation length of the bus expressed in bits. The big value of a characterizes networks that are either long geographically or very fast, or both. Thus, at the threshold of the forthcoming era of very high transmission rates and increasing demands for wide-area networks with the functionality of LANs, DQDB has been enthusiastically received by the networking community. DQDBs disadvantages can be stresses in the following two points: (1) The flexibility of the network is limited: each station must know the relative location on the bus of every other station. (2) The network is susceptible for faults: the failure of one of the extreme stations or disconnection of one bus segment makes it totally inoperable. In this paper, a capacity-1 network inspired by the DQDB concept which attempts to eliminate the above disadvantages of original DQDB is proposed. The solution is based on the UU-BUS topology, i.e., a network consisting of two separate, folded, unidirectional busses.

Dynamic recognition of the configuration of bus networks

Abstract The configuration of most local area networks changes in time, either as a result of adding or removing stations, or as a result of stations failing for one reason or another. Many situations require that all the stations connected to the network must possess a certain amount of knowledge about its configuration. As the network configuration changes, such information has to be updated, preferably automatically, i.e. by the stations themselves without external intervention. A family of algorithms for automatic recognition of the current configuration is presented for a number of bus networks. The algorithms presented are useful in medium access protocols based on partial (or total) knowledge about the configuration.