This paper investigates the compress-and-forward scheme for an uplink cloudradio access network (C-RAN) model, where multi-antenna base-stations (BSs) areconnected to a cloud-computing based central processor (CP) viacapacity-limited fronthaul links. The BSs compress the received signals withWyner-Ziv coding and send the representation bits to the CP; the CP performsthe decoding of all the users' messages. Under this setup, this paper makesprogress toward the optimal structure of the fronthaul compression and CPdecoding strategies for the compress-and-forward scheme in C-RAN. On the CPdecoding strategy design, this paper shows that under a sum fronthaul capacityconstraint, a generalized successive decoding strategy of the quantization anduser message codewords that allows arbitrary interleaved order at the CPachieves the same rate region as the optimal joint decoding. Further, it isshown that a practical strategy of successively decoding the quantizationcodewords first, then the user messages, achieves the same maximum sum rate asjoint decoding under individual fronthaul constraints. On the jointoptimization of user transmission and BS quantization strategies, this papershows that if the input distributions are assumed to be Gaussian, then underjoint decoding, the optimal quantization scheme for maximizing the achievablerate region is Gaussian. Moreover, Gaussian input and Gaussian quantizationwith joint decoding achieve to within a constant gap of the capacity region ofthe Gaussian multiple-input multiple-output (MIMO) uplink C-RAN model. Finally,this paper addresses the computational aspect of optimizing uplink MIMO C-RANby showing that under fixed Gaussian input, the sum rate maximization problemover the Gaussian quantization noise covariance matrices can be formulated asconvex optimization problems, thereby facilitating its efficient solution.
展开▼
