Optimal Design of SWIPT Systems With Multiple Heterogeneous Users Under Non-linear Energy Harvesting Model

This paper investigates the optimal power minimization design of simultaneous wireless information and power transfer systems under non-linear energy harvesting (EH) model, where a multi-antenna hybrid access point (H-AP) simultaneously transmits information and power to multiple heterogeneous users, such as information-energy receivers (IERs), information receivers (IRs), and ERs. Power splitting (PS) receiver architecture is adopted at all IERs. In order to achieve green system design, an optimization problem is formulated to minimize the transmit power of H-AP subject to the required signal-to-interference-plusnoise ratios (SINRs) constraints at IERs and IRs, and the harvested energy constrains at IERs and ERs, by jointly optimizing the beamforming vector at H-AP and the PS ratios at IERs. Since the problem is nonconvex and the employ of the non-linear EH model makes it more difficult to solve, the semidefinite relaxation and variable substitutions are used to handle it. For some cases, we theoretically prove that the globe optimal solution can be guaranteed by using our method, and for rest cases, we discuss its optimality via simulations. Numerical results show that although the traditional linear EH model is feasible for the practical EH circuits in some cases, the corresponding system consumes more transmit power than that under the non-linear model EH. Moreover, the effects of the numbers of users, the required SINRs, and the harvested energy on the system transmit power are also discussed.