Research On Detecting Technology In Uplink Mu-MIMO System

As multimedia data services are becoming the dominant services in wirelesscommunications, the demand of high transmitting rate is growing rapidly. Meanwhile, theavailable frequency resources are rarer and rarer. The Multiple Input Multiple Output(MIMO)technique can significantly raise the transmitting rate without adding bandwidth and transmittingpower and is already at the core of the next-generation wireless communication systems.Compared with the Single-User MIMO (SU-MIMO) system, the Multi-user MIMO(MU-MIMO) system can make full use of spatial resources to enhance the throughput of systemand has attracted more and more attentions. The dissertation has concentrated on the research ofthe detection technique of the uplink MU-MIMO system. The main contents and innovation aresummarized as follows:1.In the classic Block diagonalization(BD) scheme, most of the spatial resources areconsumed to direct nulls and a lot of diversity gain is lost. So a linear detection algorithm basedon block QR decomposition(BQRD) is proposed, in which the co-channel interferences(CCIs) issuppressed by projecting the channel matrix of each user to the zero space of channel matrixs ofother users. Then the MU-MIMO channel matrix is decomposed into an orthogonal matrix and ablock upper triangular matrix. With the result, users’ signals can be detected with interferencecanceling method. Theoretical analysis and the simulation results show that the BQRD algorithmachieves better detection performance with lower computational complexity than the BDscheme.2.In the BD scheme, the interference of noise is not taken into consideration. So a lineardetection algorithm based on maximizing signal-to-noise-plus-interference-ratio(MSINR) isproposed. The channel state information is made full use of to design filters to suppress theinterferences among users and noise.The design of algorithm under imperfect channel stateinformation(CSI) is provided. Theoretical analysis and the simulation results show that theMSINR algorithm can acheive much better detection performance than the BD scheme.3. Combining the space-time code with MU-MIMO system, the detection performance canbe greatly improved. A space-time linear detection algorithm is proposed, in which the linearrecombination is imposed on the received signals. Then the MU-MIMO system with space-timecode is turned to a equivalent MU-MIMO system without code. So the linear detectingalgorithms can be applied rather than the ML(Maximum-Likelihood) scheme of high complexity.The simulation results reveal that the space-time linear detecting algorithm can acheive dominantdetection performance even with imperfect CSI. 4.User scheduling scheme is an important part of radio resource management in MU-MIMOsystem. A user scheduling scheme of maximizing the minimum user SINR is proposed toimprove the system detection performance. And the design of optimal algorithm and the fastalgorithm under the scheme are also provided. The simulation results reveal that the detectionperformance is effectively improved with the proposed scheme.