Partner Selection Based on Optimal Power Allocation for Lifetime Maximization in Cooperative Networks

This paper investigates the joint partner selection and power allocation problem for two different network power management policies, i.e, group lifetime maximization (GLM) and minimum weighted total power (MWTP), with the aim of lifetime maximization in symbol error rate (SER)-constrained pairwise cooperative wireless networks. With the optimal power allocation solution derived according to each policy, we show that the optimal partner selection can be obtained using the bottleneck matching (BM) algorithm and the maximum weighted matching (MWM) algorithm for GLM and MWTP policies, respectively. Since partner selection with the BM algorithm is not power efficient, we propose a novel minimum BM (MBM) algorithm to solve the optimal partner selection problem for the GLM policy. We further propose a suboptimal partner selection (SPS) algorithm for these policies with lower time complexity, compared with the optimal partner selection schemes. Simulation results demonstrate that the proposed partner selection and power allocation strategies based on the GLM policy have better network lifetime performance over the strategies based on the MWTP policy and achieve significant network lifetime improvement over conventional strategies of minimizing the total power per transmission, with and without an equal power constraint. Compared to the MBM and SPS algorithms, partner selection based on the BM algorithm for the GLM policy has inferior network lifetime performance at a low update interval. We show that partner selection based on the MBM algorithm achieves maximum network lifetime performance and that partner selection with the SPS algorithm performs very close to it.