A Distributed Game-Theoretic Approach for Target Coverage in Visual Sensor Networks

Abstract

Visual sensor networks normally consist of a collection of camera sensors deployed randomly yet densely to fully cover a set of targets. Due to high redundancy incurred, it is possible to both preserve energy and enhance coverage quality by first switching off some sensors and then adjusting the orientations of the remaining ones. The problem is that no global knowledge of the environment is available to be used to decide which sensors should be switched off and which ones should adjust their orientations. In this paper, we propose a new distributed game theoretic approach to full target coverage. A potential game is formulated in which a utility function is designed to consider the tradeoff between coverage quality and energy consumption. In order to solve the game, we present a distributed payoff-based learning algorithm where each sensor has only access to its last two actions played and own utility values. Simulation results show that our proposed game-theoretic approach has greater energy efficiency and can extend the network lifetime, as compared with prior approaches.