| Nowadays,the amount of devices connecting to the mobile communication system are becoming huge and the needs for new services are springing up.The time of big data has arrived.As the carrier of large amount of data traffic,mobile communication system can’t meet the need of increasing demands for data transport.To solve the problem,5G has been studied recently.One of the key technologies of 5G is massive MIMO(massive Multiple-Input Multiple-Output,massive MIMO).Scaling up to tens or hundreds of antennas,massive MIMO is capable of achieving higher data rates and fully exploiting spatial diversity.Thus massive MIMO has become an indispensable technology for future wireless communication network.However,with the increasing of antennas and users,the dimension of received vectors becomes large,which makes it difficult for receivers to recover signal.Meanwhile,with the increasing of users,the interferences between cells have to be dealt with.Based on the two problems above,this thesis focuses on the detection algorithms and interference suppression technologies in massive MIMO systems.Firstly,low complexity detection algorithms for massive MIMO have been investigated,such as LAS(Likelihood Ascend Search,LAS),M-LAS(Multistage LAS,M-LAS)and other improved LAS algorithms.To enhance the ability of escaping local minima,genetic algorithm has been applied.With the grouping strategy and fixed number of iteration,an improved algorithm has been proposed.Through simulation,the proposed algorithm achieves better performance than M-LAS.When number of antennas is larger than 100,GGALAS(Grouped Genetic Algorithm LAS)has lower complexity than M-LAS.In addition,Time delay of GGALAS accounts for approximately 10%of that of M-LAS.Secondly,RTS(Reactive Tabu Search,RTS),LTS(Layered Tabu Search,LTS)and R3TS(Random-Restart Reactive Tabu Search)algorithms have been investigated.To enhance the ability of escaping local minima of RTS,reliability has been used to choose right branches,with each branch applying a new escaping strategy.An improved algorithm termed as PTS(Parallel Tabu Search)has been proposed.According to simulation results,the proposed algorithm achieves better detection performance than RTS at 4QAM.At last,pilot assignment and PCP(Pilot Contamination Precoding)precoding for decontamination have been investigated.To decrease the complexity of pilot assignment method,a method based on repetition of pilot usage is proposed.The proposed method defines the pilot repetition table based on sorted order of cell loads,then determines cell groups to be improved.According to results,the proposed method achieves better sum rates than randomly assigned pilot method. |
