Optimal Energy Allocation to Maximize Network Utility of Wireless Sensor Networks Based on Data Fusion

Abstract

Wireless sensor network (WSN) is energyconstrained, the data fusion techniques used in wireless sensor network can significantly reduce energy consumption and prolong the lifetime of network. In this paper, we study the optimization of performance data fusion in sensor network as a network utility maximization (NUM) problem, and we argue that energy need to be allocated for data transmission and fusion. Different from the classical NUM model which usually takes the simple physical quantity as independent variable, our optimization scheme attempts to find a better energy allocation strategy and a more accurate model for data fusion in WSN. To deal with this issue, we developed a heuristic solution by introducing a quadratic term into the objection function on energy allocation ratio. Despite the difficulty, based on optimal model and primal-dual analysis, we develop a decentralized algorithm to solve the performance optimization problem. Extensively experimental results demonstrate that the proposed algorithm not only converge to the Pareto optimal tradeoff between rates, power and energy allocation ratio for all users, but also provides energy allocation strategy.