Coverage Game in Wireless Sensor Networks

Abstract

In a wireless sensor network, we deploy a number of energy constrained nodes in an area to fulfill some monitoring task for a period. There exists a trade-off between network lifetime and coverage performance. We investigate this tradeoff by maximizing the average coverage performance during a given network lifetime. Our main idea is to divide all the nodes in the network into several disjoint groups, and at each time only one group are in the active mode. To maximize the average coverage, we want the overlapped sensing area in each group to be minimized. Furthermore, to achieve this objective distributively we formulate this group division and coverage maximizing problem into a game theory model and our desired solution in this model is a nash equilibrium strategy profile. Finally, we apply a hill-climbing nash equilibrium convergence idea into our coverage game and let each node converge to an approximate nash equilibrium in a coherent local way. Our simulation results show that from a random start point we can realize convergence in only 20 to 45 iterations when number of nodes in the network increases from 80 to 170. What's more, the coverage performance provided by nash equilibrium solution is 96.43% and 95.49% to a derived upper bound in the high density grid and random networks respectively