Research On The Key Technologies Of Massive MIMO

In order to meet the requirements of modern society,5G should have the characteristics of high channel capacity,high spectrum efficiency and high energy efficiency,as the next generation wireless communication criteria to be used in the future.A lot of former works have verified that Massive MIMO could greatly improve the channel capacity,spectrum efficiency and energy efficiency of cellular communication networks,with the increasement of the number of antennas at base stations.Therefore,Massive MIMO is one of the candidate technologies for 5G,and it is also one of the key research topics in recent years.With the increasement of antennas at base stations,Massive MIMO shows many advantages.But it also has some weaknesses,such as pilot contamination,channel estimation and detection is very complex,and so on.In this thesis,we talk about the opportunities and challenges that Massive MIMO meets at first.Then we give a further research on the key technologies that Massive MIMO used to combat its shortcomings.We propose an effective precoding algorithm for the downlink of multicell Massive MIMO network.Another effective channel estimation and detection algorithm has also been proposed in this thesis.Precoding is one of the key technologies which are often used for interference cancellation of cellular networks.An effective Conditional Regularized Zero-Forcing(CRZF)Precoding for the downlink of multicell Massive MIMO network has been proposed in this thesis.We have got the explicit deterministic equivalent expression of the signal-to-interference-pluse-noise(SINR)ratio for each user,with the results of Sebastian Wagner's former work on Regularized Zero-Forcing(RZF)precoding on single cell Massive MIMO network and the random matrix theory.With the results of CRZF precoding,we can learn about the parameters which affect the SINR of users.Then we can decrease the adverse effects of these parameters by setting suitable values for them.Furthermore,in order to decrease the effects of these parameters to the detection performance,we can deduct the inter-cell interference and noise signals before detection.Channel state information(CSI)is very important to MIMO technology.Because of the inter-cell interference which causes the pilot contamination problem,the precise of traditional channel estimation methods for Massive MIMO networks is very low,and the detection performance is also very bad.An enhanced eigenvalue decomposition based(EEVD)channel estimation and detection algorithm is proposed in this thesis.Comparing with the traditional channel estimation methods for Massive MIMO network,we should only get the lower dimensional projected channel matrix with EEVD algorithm,which greatly reduce the computional complexity.EEVD algorithm requires less pilot symbols,which could improve the spectrum efficiency of Massive MIMO network.When the number of antennas at base stations tends to infinity,EEVD algorithm could eliminate the pilot contamination problem completely.When the signal-to-noise ratio(SNR)is 40 dB,the mean square error(MSE)performance of EEVD algorithm is two orders of magnitude better than traditional eigenvalue decomposition(EVD)based algorithm,and the bit error rate(BER)is approximately one percent of that for traditional maximum ratio combination(MRC)detection algorithm.