Analysis And Improvement Of Detection Algorithms In V-BLAST Wireless Systems

As the continuously improve pursue of the communications, achieve higherspectrum efficiency has become a key question in wireless communication. Comparedwith the single input single output (SISO) system, multiple-input multiple-output(MIMO) system can improve spectrum efficiency and system capacity obviously. Andtherefore it is regarded as one of the key technology of the next generation wirelesscommunication.The signal detection algorithms in the receiver play a very important role in theperformance of MIMO system. How to design a detection algorithm with highperformance and low complexity has been an important issue of MIMO system.V-BLAST (Vertical Layered Space-Time) system, which is the most typical applicationof the spatial multiplexing in the MIMO system, has a simple and practical structure andhigh spectral efficiency, so it is attracted widespread attention.On the background of scientific and technological projects of Chongqing(CSTC,2011AB2044),this paper focused on research and analysed of signal detectionalgorithms for V-BLAST system. The main research contents are as follows:①Starting from the characteristics of channel and the model of MIMO systems,analyzed the capacity of MIMO channel simply and simulated it by Matlab. Finally,analyzed the principle of the typical space-time coding, and compared their performanceby simulation.②Analyzed the detection algorithms of V-BLAST structure in detail, includingML, ZF, MMSE, QR, etc, and compared their advantages and weakness, finally verifiedthis view by MATLAB simulation.③Focused on the sphere decoding algorithm, described its flow process, andanalyzed the two choice methods of initial radius, finally verified sphere decodingdetection performance by Matlab simulation. This method can achieve a nearperformance of Maximum Likelihood (ML) detection by less calculation, and thereforeit has a great applicable value.④Against the existing questions of QR decomposition and the traditionaliterative QR decomposition, mainly analyzed a low complex iterative QRdecomposition algorithm. In the first time of QR decomposition, we just save the lastlayer’s result temporarily, but not detected it, and only eliminate its interference to other layers. At the same time, we used the reduced channel matrix H to iterationdetection.Simulation results show that the algorithm can get a reliable performance witha lower complexity.⑤Proposed a new joint detection method which combined iteration QRdecompostion and the PIC algorithm. Based on the original QR decomposition, throughparallel interference cancellation, we got a new receiving signal vector, and then usedthe iterative QR decomposition method to detect again. This algorithm has avoided thepseudo inverse of the channel matrix, and made a full use of the signal with the highestdiversity gain, and reduced the error propagation. It has improved the detectionperformance with a slight increase of the complexity.