The Research Of Lattice Reduction Aided Detection Algorithms Based On MMSE Criterion In MIMO System

Multiple-input multiple-output (MIMO) technology is an important field of wireless communications technology breakthrough. Spatial diversity MIMO systems and spatial multiplexing make full use of space resources, thereby suppressing the fading between channels. MIMO technology has improved the spectral efficiency and channel capacity of the system without increasing the transmit power and system bandwidth conditions. However, signal detection technology often directly affect the MIMO transmission in wireless communications. Therefore, the new algorithm has been improved for a more in-depth research on the based of some traditional detection algorithm, which made a good trade-off between performance and complexity. The main work of this thesis is summarized as follows:Firstly, we briefly describe the basis principles of MIMO system, analyse the MIMO channel model and these corresponding traditional signal detection algorithms. Then deep study is on the detection algorithms with improvement solution. Traditional detection algorithms are not able to achieve great trade-off between detection performance and computational complexity, then we further introduce the lattice reduction(LR) algorithm which includes DLR algorithm. When they aid the linear detection algorithm and the nonlinear detection algorithm, they can obviously improve detection performance and reduce computational complexity.Secondly, the ordered successive interference cancellation (OSIC) algorithm has repeated MP inverse and the overall complexity becomes higher. To solve this problem, the ordered successive noise projection cancellation (OSNPC) algorithm is proposed. Furthermore, when the OSNPC is applied to the DLR algorithm, the computational complexity is further reduced. Since each iteration of the algorithm OSNPC preferred symbols to detect the maximum signal to noise ratio, and Signal correctness should depend on the current noise samples, so an improved Improved-OSNPC algorithm is proposed. When it is combined with the DLR, the detection performance has improved obviously and the complexity remains unchanged.Thirdly, simulations are conducted to confirm the proposed detetion algorithms in terms of symbol error rate (SER) as well as computational complexity. Then simulation performance comparison and complexity analysis are carried to lattice reduction aided the traditional detection algorithms. And finally simulation comparison and complexity analysis are carried to the improved detectors. The two improved algorithms can make an appropriate trade-off between detection performance and computation complexity.