Throughput maximization method for SWIPT DF multi-relaying network with low computational complexity

Abstract

In this paper, maximum achievable throughput of downlink decode-and-forward (DF) multiple input multiple output (MIMO) simultaneous wireless information and power transfer (SWIPT) relaying system is investigated. In the considered network, a/two MIMO relay(s) can harvest power and transfer information simultaneously. In this paper, three possible scenarios are taken into account. In the first scenario, there is no direct link between the base station (BS) and the users; indeed, the information transfer is done between the BS and the users via a relay. In the second one, some users get service directly via the BS and other part of them receive the BS’s transmitted information via the relay. In the third scenario, there are two relays alongside the direct link. In the all scenarios, the allocated powers to the spatial sub-channels of the BS-relay, the relay-users links and the BS-users links as well as power splitting ratio (PSR) at the relay(s) are optimized. Based on the complicated form of optimality conditions in the all scenarios, two novel power allocation algorithms proposed which can find the optimal solution with a low convergence time. Indeed, the throughput maximization problem is separated into two sub-problem. A novel algorithm couples this sub-problems based on the optimal conditions, which are obtained through convex optimization framework. It should be noted that PSR optimization is carried out based on the sub-gradient method. Moreover, the power allocation and PSR optimization are successively done. Simulation results validate the efficiency of the proposed methods in terms of optimality and convergence time.