Cloud and edge processing by precoder design for secrecy rate maximization

Abstract

In this paper, we investigate the ergodic sum secrecy rate of cloud and edge processing in multi-cell downlink system by adopting the artificial noise precoding (ANP). To elaborate, in cloud processing, we consider the outdated channel state information (CSI) inevitably induced by fronthaul link which connects the cloud and radio units in return for centralized processing. In contrast, the radio unit (RU) in edge processing only manages its cell using timely but local CSI. Furthermore, under per-antenna power constraints, we obtain a sub-optimal precoder to maximize the ergodic sum secrecy rate for both processing. With simulation results, we compare the ergodic sum secrecy rate of the cloud and edge processing, and show that the sub-optimal precoder performs well.