Implicit cooperation in distributed energy-efficient networks

TL;DR

This paper explores how distributed wireless transmitters can cooperate to improve energy efficiency through repeated game strategies, without direct cooperation links, by analyzing fast and slow power control scenarios.

Contribution

It introduces a novel approach to stimulate cooperation among selfish transmitters using repeated game frameworks and receiver signals, without requiring cooperation links.

Findings

Proposes an efficient cooperation plan for fast power control scenarios.

Derives equilibrium utility regions for slow power control using recent game theory results.

Demonstrates that cooperation can be achieved with mild information assumptions.

Abstract

We consider the problem of cooperation in distributed wireless networks of selfish and free transmitters aiming at maximizing their energy-efficiency. The strategy of each transmitter consists in choosing his power control (PC) policy. Two scenarios are considered: the case where transmitters can update their power levels within time intervals less than the channel coherence time (fast PC) and the case where it is updated only once per time interval (slow PC). One of our objectives is to show how cooperation can be stimulated without assuming cooperation links between the transmitters but only by repeating the corresponding PC game and by signals from the receiver. In order to design efficient PC policies, standard and stochastic repeated games are respectively exploited to analyze the fast and slow PC problems. In the first case a cooperation plan between transmitters, that is both efficient and relies on mild information assumptions, is proposed. In the second case, the region of equilibrium utilities is derived from very recent and powerful results in game theory.