Potential Games for Energy-Efficient Resource Allocation in Multipoint-to-Multipoint CDMA Wireless Data Networks

TL;DR

This paper models noncooperative resource allocation in multipoint-to-multipoint wireless networks as potential games, demonstrating convergence and significant performance improvements through simulations.

Contribution

It introduces a game-theoretic framework for energy-efficient resource allocation in complex wireless networks, ensuring convergence to Nash equilibria.

Findings

Games converge reliably in simulations.

Resource allocation improves network performance significantly.

Potential game approach applies to various network scenarios.

Abstract

The problem of noncooperative resource allocation in a multipoint-to-multipoint cellular network is considered in this paper. The considered scenario is general enough to represent several key instances of modern wireless networks such as a multicellular network, a peer-to-peer network (interference channel), and a wireless network equipped with femtocells. In particular, the problem of joint transmit waveforms adaptation, linear receiver design, and transmit power control is examined. Several utility functions to be maximized are considered, and, among them, we cite the received SINR, and the transmitter energy efficiency, which is measured in bit/Joule, and represents the number of successfully delivered bits for each energy unit used for transmission. Resorting to the theory of potential games, noncooperative games admitting Nash equilibria in multipoint-to-multipoint cellular networks regardless of the channel coefficient realizations are designed. Computer simulations confirm that the considered games are convergent, and show the huge benefits that resource allocation schemes can bring to the performance of wireless data networks.