Joint Precoding And Decoding Design In C-RAN With Imperfect CSI

Currently,cloud-radio access network(C-RAN),as one of the key technologies of the 5th generation mobile communication system(5G),has received extensive attention.The main components of C-RAN include three parts: A baseband processing unit(Baseband Unite,BBU)pool,remote radio heads(Remote Radio Heads,RRH),and fronthaul links.Many works have shown that this technology has obvious advantages in suppressing interference,saving energy consumption and allocating network resources.However,due to the influence of channel estimation error and feedback delay,the channel state information(CSI)obtained by the BBU is often inaccurate.Due to the limited capacity of the fronthaul links,it is necessary to compress and then send the signal.This processing leads to the impact of quantization noise on the compressed signal.In response to the above two issues,this article carried out the following works:First,the research background of C-RAN is introduced.Precoding overview is described combining with the current research.We study the joint precoding and decoding design in C-RAN with imperfect CSI.Combined with the limitations of the existing access network architecture,the advantages and challenges of C-RAN are further introduced.The common precoding schemes are introduced with imperfect CSI.Secondly,for the C-RAN with imperfect CSI,the precoding and decoding matrices of the transceiver are jointly designed.Considering the joint correlation Rice channel model,the total noise including interference is calculated.Its covariance matrix expression is derived.Using the tight lower bound of mutual information as a design criterion,a non-convex problem is transformed into an equivalent convex problem.Using the convex optimization method,the expressions of the joint precoding and decoding matrices and the structure of the optimal transceiver are derived.The optimal transmission covariance for the lower bound is further derived.Based on the analytical expressions of the matrices,an iterative algorithm for optimal precoding and decoding matrices is proposed.The simulation results verify the performance of the proposed joint precoding and decoding design and reveal the impact caused by the correlation between transmitter and receiver.Finally,for the C-RAN with limited backhaul capacity,the precoding and decoding matrices of the transceiver are designed jointly.Based on the imperfect CSI,the expression of the covariance matrix including the total noise with the quantization noise is calculated.Similarly,using the tight mutual information lower bound as a design criterion,the expression of the tight mutual information lower bound affected by quantization noise is deduced.The covariance matrix of the quantization noise is calculated using the bisection method,and then the equivalent optimization is solved by convex optimization convex problem.The expressions of the optimal precoding and decoding matrices are derived.An iterative algorithm of optimal precoding and decoding matrices is proposed.The simulation results verify the performance of the proposed joint precoding and decoding design and reveal the impact of the correlation between transmitter and receiver.