Channel Estimation Technology Research In Massive MIMO System

To further meet the people's urgent requirements of the high quality of mobile communications,massive MIMO,as the key technology for increasing the data transfer rate,has become a hot topic of 5G.However,with the increase of the number of antennas,the pilot length is limited by the coherence time of channel.The limited pilot length not only makes the traditional channel estimation method unable ues in massive MIMO,but also makes the pilot contamination problems in massive MIMO system is particularly prominent.Compressed Sensing(CS)theory is a novel signal rocessing theory which is well known in the wireless communication field.This paper mainly analyzes the SCS(Sparse Common Support)features of the massive MIMO channel and proposes an improved compressed sensing channel estimation method based on the SCS features.Compare to the traditional CoSaMP algorithm,the improved CoSaMP algorithm accelerates the convergence rate,so the algorithm complexity is reduced.In addition to for a pilot contamination problems significant existed in multi-cell massive MIMO system,this paper uses Hadamard matrix to orthononalize the pilots in different cells.Finally,for the proposed optimization method,this paper makes a simulation and analyzes the simulation results.Finally,compared the complexity and performance simulation results before and after optimization,this paper comes to a conclusion that the proposed algorithm time complexity is a lot lower than the previous optimization in computation complexity,but the systems performance is not reduced.The conclusion proves the effectiveness of the algorithm.The simulation results also show that the pilot contamination can be significantly reduced by use the hadamard matrix to orthogonal the different cell pilot.