Linear approximation for energy and throughput optimization in Wireless Sensor Networks

The focus of this paper is to analyze and assess the performance of various optimization techniques in the effort to determine the maximum lifetime of the nodes and the throughput in a Wireless Sensor Network (WSN) environment while being able to meet the network´s application specific design requirements. We formulated the initial objective of the optimization problem as a nonlinear function and solved an instance of the problem in its original form using LOQO [7]. The objective function is then relaxed to assume simplified formulations based on certain assumptions in order to be solved using Geometric Programming (GP) and Linear Programming (LP) techniques with specific solvers, namely CVX and GGPLAB in Matlab. The main findings of this study reveal that there exists a tradeoff in the solutions obtained between performance accuracy and computational complexity. The resulting energy and delay minimization approach is shown to be able to support the various WSN application requirements.