| Massive multiple-input multiple-output(MIMO)is one of the key technologies in physical layer of the 5th generation mobile communication system(5G),by deploying a large number of antennas at base stations,massive MIMO has the ability to help multi-ple users communicate with the base stations simultaneously at the same time-frequency resources,which greatly improves the system spectral efficiency and energy efficiency.Moreover,massive MIMO can increase system capacity by increasing the number of base station antennas instead of increasing the density of base stations and bandwidth.It has been proved that in the limit of an infinite number of antennas,the effects of fast fading and uncorrelated noise will vanish,and the performance of the simple linear precoding algorithms in massive MIMO can be nearly optimal.On a separate track,in order to meet the exponentially increasing demand of mobile data,by overlaying the macro base stations(MBSs)with another tier of densely deployed lower-power small access points(SAPs),heterogeneous networks can effectively enhance the system capacity and provide superior user experience.The integration of heterogeneous networks and massive MIMO is able to provide a reliable solution in improving the overall system performance of 5G network.The accurate acquisition of channel state information(CSI)is critical to maximize the system performance in the massive MIMO systems.By utilizing the channel reci-procity,the desired CSI can be obtained through uplink training in time division duplex(TDD)mode.However,the accuracy of channel estimation heavily depends on having perfectly orthogonal pilot sequences allocated to the users in the uplink training.Lim-ited by the channel coherence interval,it is impossible to allocate the orthogonal pilot sequences to every user,the same orthogonal pilot sequences,or nonorthogonal pilot se-quences must be allocated to the users of the neighboring cells.Thus the results of chan-nel estimation obtained in the desired cell is contaminated by pilot sequences transmitted by users in other interference cells,which greatly degrades the accuracy of channel esti-mation.In addition,massive MIMO based heterogeneous networks are facing technical challenges in designing and deploying.On the one hand,with the dense deployment of S-APs in heterogeneous network,the serious interference and the limited backhaul capacity will become the main bottleneck in improving the system performance of 5G heteroge-neous networks,on the other hand,with the increasing number of base station antennas,the fully digital precoding algorithm has a high cost,which is difficult to implement.In order to solve the above problems,this thesis investigates the pilot contamination reduction in massive MIMO and the data transmission in massive MIMO based hetero-geneous networks.The main works are as follows:1)The effect of pilot contamination on system performance of massive MIMO in TD-D mode is analyzed,and the analytical expressions of downlink spectral efficiency under different pilot reuse factors is presented.Moreover,the analytical expressions are presented on the normalized mean square error(NMSE)of the minimum mean square error(MMSE)channel estimation algorithm.Based on the obtained NMSE,an optimal pilot assignment strategy to minimize the effect of pilot contamination is proposed.2)In order to reduce the large overhead of pilot sequences in the uplink training of the conventional massive MIMO system,a Chu-based pilot design scheme is proposed.Chu sequences have perfect auto-correction and cross-correlation properties.By us-ing the Chu sequences as the basis pilot signal,the proposed pilot design scheme obtains the system pilot matrices through rotation and the parameter design.The de-signed pilot sequences can not only effectively reduce the pilot overhead,but also minimize the effect of pilot contamination.3)For the problem of serious interference and unbalanced network load caused by the dense deployment of SAPs in heterogeneous networks,massive MIMO is introduced to reduce the effect of downlink interference caused by the small cell range expan-sion.Firstly,different cell range expansion boundaries and the values of cell range expansion basis are given,then a downlink precoding scheme with limited backhaul capacity is proposed.According to the number of antennas deployed at MBS and SAPs,the cell-edge-aware zero forcing(CEA-ZF)algorithm adopted in the downlink transmission of MBS exploits the spatial degrees of freedom from large antenna array to suppress the inter-cell interference,and the block diagonalization(BD)precoding algorithm is introduced at SAPs to eliminate the multi-user interference.4)A new hybrid precoder for millimeter wave massive MIMO based wireless backhaul in heterogeneous network is proposed to simultaneously support multiple SAPs with multi-streams.Compared with the conventional precoding scheme,the proposed hy-brid precoding scheme does not require the number of radio frequency(RF)chains to be comparable with that of antenna elements,which effectively reduces the cost and implement complexity.At the same time,the proposed precoder with lower complexity can achieve a sum-rate performance that approaches the one under fully digital precoder.To sum up,this thesis investigates the pilot contamination problem in massive MI-MO and the massive MIMO based downlink data transmission in 5G heterogeneous net-works.The effect of pilot contamination on the performance of massive MIMO systems is deeply analyzed,and a practical solution is given.Moreover,the application of massive MIMO in small cell range expansion and the downlink data transmission of 5G hetero-geneous networks is explored,which provides a theoretical method with reference value for the research on 5G network capacity enhancement. |
PDF Full Text Request