Joint relay selection and power control for robust cooperative multicast in MmWave WPANs

摘要

We put forward a joint optimization algorithm for relay selection and source & relay power allocations under line-of-sight (LoS) and non-LoS mixed conditions for both power saving and robustness enhancement in practical cooperative multicast millimeter-wave (mmWave) wireless personal area networks (WPANs). Firstly, a novel beam training protocol capable of overhearing is devised to filter relay candidates with non-LoS links and avoid relays selection for LoS paths through feedback 1-bit channel amplitude information (CAI). Secondly, we provide the problem formulation that minimizes the maximum total transmit power consumption under constraints of maximum tolerable outage probability and limited power. As these two problems are coupled together, the formulation is non-convex with extremely high complexity of finding optimal solution. By introducing relaxation method and Lagrange multipliers algorithm, we propose an optimal approach to reduce complexity while maintaining reasonable performance, and derive a closed-form expression of optimal joint relay selection and power allocation. Lastly, simulation results indicate significant improvements both on outage probability and power consumption performances over the existing sum transmit power minimization algorithm.