Research On Channel Estimation Algorithm Based On Pilot For Massive MIMO-OFDM System

The next generation of wireless communication systems(5G)is faced with challenges such as a sharp increase of wireless data traffic and mobile terminal numbers,and users have requirements for wireless communication of high-speed,low-latency and high-reliability.Massive MIMO technology can multiplies the system capacity and data transfer rate without increasing system bandwidth via equipping dozens or even hundreds antennas in the base station and using these antennas to establish many independent channels,the huge number of antennas brings a high degree of freedom also increases the reliability of the communication,therefore the massive MIMO technology has become one of the key technologies of 5G communication.Channel state information(CSI)is an important parameter in the wireless communication process,so channel estimation is an important part of wireless communication.The traditional blind channel estimation method is not suitable for massive MIMO systems due to the high computational complexity,and the number of orthogonal pilots is insufficient due to the limited coherence interval of Time Division Duplexing,so the pilot-based channel estimation method is affected by pilot contamination,thus restricting the performance of massive MIMO system.Therefore,it is of great significance to study the channel estimation algorithm with less pilot numbers.For this purpose,this paper mainly studies the channel estimation algorithm of massive MIMO-OFDM system.Combining the sparsity of massive MIMO system channels,this paper focus on the channel estimation algorithm based on compressed sensing(CS)because the compressed sensing theory can effectively reconstruct the sparse signal from limited sample.The main work is as follows:(1)Analyzing the models and characteristics of massive MIMO systems,the comparison of TDD and FDD,and the causes of pilot contamination.(2)Studying several commonly used OFDM channel estimation algorithms for flat fading channels,such as least squares,minimum mean square error,liner-minimum mean square error etc.,analyzing their advantages and disadvantages by Matlab simulation.(3)Analyzing the compressed sensing theory,aiming at the problem of insufficient number of orthogonal pilots in massive MIMO systems,proposing a channel estimation algorithm based on compressed sensing and two coordinated channel estimation strategies based on this algorithm,including the sparse representation of the signal,the design of the observation matrix and the design of the signal reconstruction algorithm.The Matlab simulation results show that the proposed algorithm can accomplish channel estimation with less pilot numbers,not only reduces the dependence on the pilot,but also improve the data transfer rate,and the performance is superior to the traditional least squares method.The coordinated channel estimation strategies have high precision and good anti-noise ability,are suitable for small antenna array.