Optimization of uplink rate and fronthaul compression in cloud radio access networks

Abstract

In cloud radio access networks (C-RANs), a central unit (CU) and remote radio heads (RRHs) are connected with a wired fronthaul, e.g., a common public radio interface (CPRI). Due to the limitations of the fronthaul bandwidth in 5G systems, some digital baseband processing blocks are moved from the CU to the RRHs. In this case, the uplink data and pilot symbols after digital processing are delivered from the RRHs to the CU for further processing to decode transmitted information. We consider compression of the data and pilot signals at different compression rates. In the compression of signals, as the compression rate is higher (i.e., the fronthaul rate after the compression is lower), the signal is more distorted. Moreover, compression will affect both the uplink user throughput and the fronthaul rate. The effects of the pilot and data signals on uplink throughput are formulated with an achievable rate considering channel estimation. We formulate an optimization problem to address the tradeoff, and find the optimal distortion variances for the data and pilot signals under high signal-to-noise ratio (SNR) conditions. We show that the optimal signal distortion variance is proportional to the number of data and pilot symbols. We provide numerical results that verify our analytical derivations.