Distributed multiple access with a general link layer channel

Abstract

This paper investigates the problem of distributed medium access control in a time slotted wireless multiple access network with an unknown finite number of homogeneous users. Assume that each user has a single transmission option. In each time slot, a user chooses either to idle or to transmit a packet. Under a general channel model, a distributed medium access control framework is proposed to adapt transmission probabilities of all users to a value that maximizes an arbitrarily chosen symmetric network utility. Probability target of each user in the proposed algorithm is calculated based upon a channel contention measure, which is defined as the success probability of a virtual packet. It is shown that the proposed algorithm falls into the classical stochastic approximation framework with guaranteed convergence when the contention measure can be directly obtained from the receiver. On the other hand, when the contention measure is not directly available, computer simulations show that a revised medium access control algorithm can still help the system to converge to the same designed equilibrium.