Receiver Design Based On Structured Prior Information

Massive MIMO-OFDM is an important achievement of the fifth-generation communication system combining Orthogonal Frequency Division Multiplexing(OFDM)technology and massive Multiple-input Multiple-output(MIMO)technology,which provides high-speed and reliable data transmission services for a large number of users,and significantly improves system performance.However,the substantial increase in the number of antennas in a massive MIMO-OFDM system also increases the difficulty of receiver design.On the other hand,interference has always been a key factor limiting wireless communication systems’ performance improvement.Aiming at the difficulty of receiver design,this paper considers the two problems of channel estimation in the inter-cell interference scenario and data detection in the burst interference scenario.We focus on exploiting the structured prior information to eliminate interference and improve the receiver’s performance.In the inter-cell interference scenario,the known interference pilot can estimate and eliminate the interference.However,the pilot resources are limited,and the interference estimation will increase the number of variables to be estimated and lead to performance deterioration of channel estimation.To this end,we propose a delay-calibrated block-wise linear model to reduce channel variables.The model in this paper considers the principal delay of the interfering channel as a key parameter and extracts it from the channel coefficients,and then applies the recently proposed block-wise linear model to approximate the residual channel.Based on the delay-calibrated block-wise linear model and the channel angle-domain sparsity,we further design a delay-calibrated Turbo compressed sensing algorithm to solve the channel estimation problem.Numerical results demonstrate that the proposed algorithm outperforms the state-of-the-art algorithms.Compared with inter-cell interference,burst interference is more difficult to handle.In particular,burst interference may only occur on data symbols and not on pilot symbols,which means that the pilot cannot measure the interference information.To cancellate the burst interference,we revisit the uplink multi-user system and build a matrix-form system model based on the frequency-domain block-wise model,and further discuss the low-rank property of the interference matrix.Then,we propose a Turbo message passing based burst interference cancellation algorithm to solve the data detection problem,which fully exploits the constellation information of the target data and the low-rank property of the interference.Besides,we design two denoisers to deal with the leaked singular values caused by channel frequency selectivity.Numerical results show that our proposed algorithm effectively mitigates the adverse effects of burst interference.