This paper investigates the linear precoder design maximizing the harvested energy in a multiple-input multiple-output (MIMO) system with simultaneous wireless information and power transfer (SWIPT), where the transmitter is assumed to know perfect instantaneous channel state information (CSI). Different from existing studies with the assumption of Gaussian inputs, this paper considers a more practical scenario; that is, input signals are taken from finite-alphabet constellation sets, such as quadrature amplitude modulation (QAM). The optimal precoding is formulated as an optimization problem, where the objective is to maximize the harvested energy under the constraints of achievable mutual information level and transmit power. Although the formulated problem is NP-hard, by employing its structure, our analysis shows that it can be transformed into a semidefinite programming (SDP) problem. Based on this, a general algorithm for near optimal precoder is proposed. Besides, for the SWIPT system with co-located receivers, by decomposing the precoder, a low-complexity algorithm with almost the same performance as the general algorithm is developed. Simulation results validate the efficacy of proposed algorithms.
Linear Precoder Design Maximizing Energy Harvesting in SWIPT Systems with Finite-Alphabet Inputs
2019-09-01
168792 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch