A three-stage robust approach for minimum power multicasting in wireless sensor networks

In this paper we study the minimum power multicasting problem in wireless sensor networks, where the transmission power required for one network terminal to transmit to another is subject to uncertainty. We propose a three-stage approach based on linear programming. Our novel approach adds two polynomial time stages to a first stage corresponding to a classical approach well known in the literature and based on a mathematical programming formulation. Stages two and three are introduced to adjust the transmission power of nodes in such a way to achieve a better protection against uncertainty, without increasing the total power consumption. Experimental results show that our three-stage approach increases the robustness of the solutions obtained using the classical approach.