Multi-Cell Channel Estimation On Compressive Sensing In Massive MIMO Systems

The rapid development of mobile communication systems has not only changed people's life,but also become an important engine of social informatization.MASSIVE MIMO becomes one of the key technologies of 5G by deploying a large number of antennas at transceivers.The mobile communication systems using MASSIVE MIMO technology achieves 3~5 times spectrum efficiency of the traditional base station under the same bandwidth,better network coverage and more customer satisfaction,etc.The algorithm complexity of channel estimation is greatly increased due to the numerous antennas at transceivers.Therefore,it is of great significance to obtain lower overhead of training sequence and increase channel estimation accuracy rate.MASSIVE MIMO system mainly incorporates TDD and FDD transmission modes.We propose an adaptive channel estimation algorithm in TDD transmission mode of Multi-Cell Multi-User MASSIVE MIMO systems where an approximate sparse synthesis channel matrix is built.The Fruit Fly algorithm is introduced into the StOMP algorithm to achieve adaptive search iteration threshold.Simulation results show that the proposed algorithm effectively improve the accuracy rate of channel estimation.We also present a threshold-based adaptive sparseness compressed sensing channel estimation algorithm in FDD transmission mode of MASSIVE MIMO systems where a sparse channel model is established at angular domain.The atomic selection rules of the Backtracking Orthogonal Matching Pursuit(BAOMP)algorithm are introduced into SAMP algorithm,so the initial atoms of the SAMP algorithm are effectively selected by setting reasonable threshold.That means that the maximum approximate coefficient of the original signal is sought.It greatly reduces the iteration number and improves the convergence speed of the SAMP algorithm.The algorithm achieves adaptive channel estimation with higher estimation accuracy rate if the number of users in the cell changes.