A stochastic non-cooperative game for energy efficiency in wireless data networks

TL;DR

This paper introduces a stochastic adaptive algorithm for energy efficiency in wireless CDMA networks, enabling users to optimize their transmission strategies with minimal prior information.

Contribution

It presents a novel stochastic algorithm combining RLS and LMS adaptations for real-time implementation of a non-cooperative game in wireless networks.

Findings

Performance approaches that of the non-adaptive game

Requires less prior information for adaptation

Achieves energy efficiency with satisfactory accuracy

Abstract

In this paper the issue of energy efficiency in CDMA wireless data networks is addressed through a game theoretic approach. Building on a recent paper by the first two authors, wherein a non-cooperative game for spreading-code optimization, power control, and receiver design has been proposed to maximize the ratio of data throughput to transmit power for each active user, a stochastic algorithm is here described to perform adaptive implementation of the said non-cooperative game. The proposed solution is based on a combination of RLS-type and LMS-type adaptations, and makes use of readily available measurements. Simulation results show that its performance approaches with satisfactory accuracy that of the non-adaptive game, which requires a much larger amount of prior information.