Multi-User Symbol-Level Precoding for Downlink Reconfigurable MIMO Communication Systems

Abstract

Pattern reconfigurable multiple-input multiple-output (PR-MIMO) can improve the communication quality compared with the traditional MIMO via modifying the wireless propagation environment actively. In this paper, we discuss the design of the constructive interference (CI)-based symbol-level precoding (CI-SLP) in the multi-user pattern-reconfigurable multiple-input single-output (MU-PR-MISO) communication systems. Firstly, the joint design problem of SLP and the radiation pattern is formulated aimed at maximizing the worst user's service quality. Based on that, the optimization problem is further decoupled and an iterative alternating optimization algorithm is proposed to obtain the desired radiation pattern and the corresponding precoding matrix. After simplifying the joint design problem in the single-pattern case where all the antenna elements adopt the same radiation pattern, the convexity of the pattern design problem can be ensured, leading to an optimal pattern design. Numerical results demonstrate that the reconfigurable radiation pattern can manipulate the wireless channel purposefully so that the performance gains by exploiting CI can be further increased.