Research On MIMO Signal Detection Algorithm Under V-BLAST Decoding

Multiple-input Multiple-output (MIMO) technology can reasonably use the weak pointof multi-path fading to make the system capacity increasing times in the same bandwidth andtransmitting power. Today, the quantity of spectrum resources is decreasing and the demandof high speed data service is increasing day by day. MIMO technology can work with hightransmission rate and reliability of information transmission in the limited bandwidthresources. So it is great popularity. Signal detection is a key technology of the MIMO systemand is coming to a hotspot of research. This paper reviews the background, the presentresearch situation and the basic theory of MIMO technology, mainly studied the signaldetection algorithm of V-BLAST system and proposed some improved algorithms. The majorcontents are summarized as follows:(1) Making a thorough research on the basic theoretical knowledge of MIMO technologyand master the basic detection algorithm of MIMO signal, such as the maximum likelihood,forced zero, minimum mean square error, serial interference cancellation and parallelinterference cancellation detection algorithm. The sort parallel interference cancellationalgorithm with the opposite test sequence is proposed on the basis of the ordered successiveinterference cancellation algorithm. And then all of the detection performance and complexityare analyzed. Finally, go simulation experiments for each algorithm.(2) The comprehensive consideration of the error propagation of QR decompositionalgorithm and the diversity gain, an improved sorted serial interference cancellation algorithmwas proposed. The improved algorithm firstly rotated left with the channel matrix, to find themost reliable channel matrix of the first detection layer. Then use QR decompositionalgorithm, and take the final testing layer as the test results. And then delete the interferenceof this layer from the received signal. Remove this layer signal corresponding to the columnvector of the original channel matrix and continue to detect until all the layers was known.The improved algorithm reduces the error propagation, and achieves maximum diversity gainof the final detection layer. So, the error rate of signal detection is low.(3) According to the high complexity of maximum likelihood detection algorithm, asplit-step ML detection algorithm was proposed. The algorithm firstly used optimal MLdetection algorithm to detect some signal by construction of smaller MIMO system, and thensubtracted the interference of detected signal, finally the rest signal continued to be detectedwith ML detection. Simulation results showed that the complexity of this algorithm is reducedgreatly with the smaller loss of performance. Complexity is reduced more obviously in the multiple antennas.