Abstract

Abstract-The problem of optimally choosing message rates for users of a store-and-forward network is analyzed. Multiple users sharing the links of the network each attempt to adjust their message rates to achieve an ideal network operating point or an “ideal tradeoff point between high throughput and low delay. ” Each user has a fixed path or virtual circuit. In this environment, a basic definition of “ideal delay-throughput tradeoff ” is given and motivated. This definition concentrates on a fair allocation of network resources at network bottlenecks. This “ideal policy ” is implemented via a decentralized algorithm that achieves the unique set of optimal throughputs. All sharers constrained by the same bottleneck are treated fairly by being assigned equal throughputs. A generalized definition of ideal tradeoff is then introduced to provide more flexibility in the choice of message rates. With this definition, the network may accommodate users with different types of message traffic. A transformation technique reduces the problem of optimizing this performance measure to the problem of optimizing the basic measure. V I.

Citations

98	Flow control: A comparative survey - Gerla, Kleinrock - 1980
15	The control of congestion in packet switching networks - Davies - 1972
11	Routing and Flow Control in Systems Network Architecture - Ahuja
5	Majithia et al., “Experiments in congestion control techniques - C - 1979
1	2831, Using 2) for the iteration number k at which the user was first saturated at 1, (k < i), all users that share I are either NTISAD781467,May - Beranek, Rep - 1974
1	TYMNET-A terminal-oriented communication network,” in AFIPS Conf. Proc., Springfoint Comput. Conf.., vol. saturated before the kth iteration or become saturated at the 38 - Tymes - 1971
1	SNA multiple-system netously saturated use the same throughput as before the kth working - Gray, McNeill - 1979
1	This continues through iteration i. Since the user is saturated at I, its throughput y satisfies y = x(r(Z)). Also, since all users of I are saturated, ysat(I, i + 1) = y(I) and y(l, i + 1) = x(&) - y(2)) = y. Thus, due to the saturation allocation at I, - iteration - 1976
1	Optimum end-to-end flow control in netTo prove that the user is not overloaded at a link l‘, it suffices to show y < x(c(Z’) - y(l‘)) where y(Z’) is the sum of throughputs of users of I’ after the i + 1st iteration. Consider the iteration (iteration k) at - Bharath-Kumar - 1981