Progressive Linear Precoder Optimization for Multicast Retransmission

Abstract

Our investigation concentrates on the linear pre-coder designed for the packet retransmission in the multicast sys-tem. To fully utilize the time diversity provided by the automatic repeat request (ARQ), we consider a sequential linear precoder optimization problem assisted with hybrid-ARQ (HARQ), which aims to maximize the minimum quality-of-service over all users under the peak power constraint of base station. To solve the non-convex problem, we propose a low-complexity high-performance algorithm based on convex-concave procedure (CCP) together with alternating direction method of multipliers (ADMM). The proposed algorithm not only guarantees a sub-optimal solution but also has closed-form update in each iteration. Numerical results show that the proposed algorithm can continuously improve the system performance under multiple retransmissions, including reducing the successful transmission time and the bit error rate.