Research On The Key Techniques Of Sparse Channel Estimation In MIMO-OFDM System

In a wireless communication system, MIMO-OFDM can effectivelyincrease the capacity of transmissionsystem and enhance the rate of data transmission, while suppress multipath fading and interference.At the receiver of MIMO-OFDM system, correlation detection and channel equalization often require precise channel state information(CSI), which is acquired by the channel estimation techniques.Wireless channel can often be modeled assparse channel model.Traditional channel estimation algorithms commonly couldn’t use inherent sparsity of the channel and require a large number of pilots to obtain accurate CSI, which seriously reduce the utilization efficiency of the channel.Extensive literatures prove that applying the compressed sensing in sparse channel estimation can greatly reduce the number of pilot and improve spectrum efficiency.The key technique of sparse channel estimation based CS in MIMO-OFDM system is the focus of the paper. The main work and contributions are listed as follows:1. We describe the model of traditional channel estimation method and the traditional channel estimation algorithm. At the same time, the mathematical model of CS theory and the three elements of CS theory are introduced, then based on the traditional channel estimation model deduce the sparse channel model, and the CS algorithm is applied to the sparse channel model. Finally make comparison between the traditional channel estimation method andthe sparse channel estimation method based on CS. 2.An improved algorithm Regularized Orthogonal Matching Pursuit(ROMP) Algorithm is proposed. The algorithm avoids choosing the wrong group and makes the group of atoms’ energy more concentrated which is selected by modified regularized principle, then adds the secondary screening of estimation results, achieves an accurate reconstruction of original signal. 3.The sparse channel estimation in MIMO-OFDM systems has been examined. Regularized Adaptive Matching Pursuit(RAMP) was applied to the channel estimation, and the algorithm’s iteration termination condition has been improved by taking the difference between the residual energy less than the set threshold to terminate the iterative process, eliminate part of the noise interference.Ultimately under the premise of unknown sparsity the improved algorithm reach the same MSE performance with Orthogonal Matching Pursuit(OMP). 4.We research on the problem of pilot position design based on CSin OFDM system. By minimum mutual coherence of the measurement matrix, a generalized optimal pilot placement scheme is proposed. The basic idea of search procedure isthe new placement is generated by replacing n pilot positions of the previous one and the pilot placement with smaller mutual coherence is selected at each iteration. After iterations, the pilot placement with smallest mutual coherence of measurement matrix is the ultimately optimal pilot placement.