Research On Limited Feedback Technology Of MIMO Wireless Communication System

MIMO (Multi-Input Multi-Output, MIMO) transmission technology can resolveconflicts of wireless communication system between the continued growth in demandand the limited spectrum resources and can increase the spectral efficiency of thecommunication system exponentially without increasing the bandwidth, as radiospectrum resources become tighter nowadays. Thus, that make it became one of thekey technologies for next-generation broadband mobile communication systemsquickly. In the MIMO system, the transmitter channel state information (ChannelState Information of Transmitter, CSIT) and the feedback of CSIT get an importantrole in the system performance.The thesis focus on the study of MIMO systems, limited feedback technology,channel prediction (AR prediction model), temporally-correlated channel andfeedback compression. In the thesis, the problem of feedback delay in MIMO systemswith limited feedback and the problem of temporal redundancy intemporally-correlated MIMO systems were proposed. Both the solutions and thesimulations of the two problems are illustrated too. Channels in the main part of thepaper are temporally-correlated MIMO channel. In dealing with the problem offeedback delay: first, according to the known channel matrices, predicting the MIMOchannel matrices in advance by use of the AR channel prediction model; Second,feeding the optical code word of the predicting channel matrices which is chosenfrom the codebook back to the transmitter for pretreatment on the basis of limitedfeedback technique based codebook; Last, proving the validity of the method byexperimental simulation. In temporally-correlated channel MIMO systems, thecharacter of time redundancy can be used to compress the system feedback overhead,the roughly steps are: First, according to the known channel matrix, predicting thechannel matrices of several successive data frames which are strong correlative;Second, find a matrix which has the strongest correlation with the other predictedchannel matrices among all the predicted channel matrices, and match the optimalcode word from the code book based on the selected matrix, Then, feeding the optimal code word back to the transmitter for transmission preprocessing; Finally, testand verify the validity of above-described method through simulation.For the problems and solutions of feedback delay and time redundancy intemporally-correlated MIMO feedback system, simulations and experiments based ontechnical indicators of channel capacity and bit error rate are done in Matlab platform.In solving the feedback delay problem, the proposed method impacts better on theperformance of channel capacity and bit error rate in temporally-correlated MIMOsystem. In solving the problem of temporal redundancy, although the proposedmethod caused some damage of channel capacity and error rate of temporalcoherence MIMO system, but it can reduce the feedback overhead exponentially, andachieve a good compromise between the feedback rate and system throughput. Theproven methods which designed to solve the problems has some practical values.