Research On Signal Detection Algorithm In MIMO Systems

MIMO technology will effectively enhance the system capacity,signal transmission rate and increase the spatial diversity gain,and do not need to increase the transmission power and system bandwidth.So MIMO technology has become a hot research field of communication with great research value,and also becomes the next generation of mobile communication key technology.However,the signal detection algorithm at the receiving end determines the performance of the MIMO system.Therefore,this paper focuses on the detection performance of the signal detection algorithm at the receiving end of the MIMO system.The specific content and innovation of this paper are as follows:1.Firstly,it introduces the background knowledge of MIMO system.Secondly,the signal model,basic principle and channel capacity of MIMO system are introduced.Then it studies the optimal Maximum Likelihood?ML?algorithm in signal detection technology,Forced Zero?ZF?and Minimum Mean Square Error?MMSE?algorithm in Linear Algorithm and interference cancellation algorithm,and uses MATLAB simulation software to compare the error bit performances of these classic algorithms in different modulation modes.The simulation results show that the ML algorithm has optimal BER performance of but largest complicated.ZF and MMSE algorithm have lower complexity but their BER performance is worst.The BER performance of ZF-MMSE algorithm and MMSE-OSIC algorithm is improved,and the improved amplitude of MMSE-OSIC is the biggest which is close to ML.2.In MIMO systems,the ML detection algorithm offers optimal performance but complicated,and the low complexity of MMSE algorithm with the detection performance is poor.The higher condition number of channel matrix has bad effects on the performance of signal detection.In lattice reduction algorithms,the LLL algorithm is a preprocessing algorithm which can effectively reduce the condition number of channel matrix.Aimed at these problems,this paper provides a signal detection algorithm based on the condition number threshold of channel matrix,which can enhance the performance of traditional detection algorithm when the condition number is high.The algorithm selects the corresponding detection algorithm by comparing the condition number of channel matrix with the preset condition number threshold:When the condition number is lower than the threshold,the algorithm chooses the lower complexity LLL-MMSE to reduce computational load;when the condition number is higher than the threshold,it chooses the sorting and grouping based detection algorithm which combining ML and LLL-MMSE,ensuring the detection performance of the system by increasing a certain amount of computational load.Through the simulation of the BER performance of the algorithm in different threshold,it turns out that the algorithm in this paper is better than the traditional LLL-MMSE detection algorithm.By setting the threshold of the condition number in advance,the algorithm in this paper achieves a better balance between performance and complexity,and finally achieves the goal of optimizing the performance of detection algorithm.3.In the higher dimension MIMO system,the bit error rate of the lattice reduction subtractive linear algorithm is poor.Element-based Lattice Reduction?ELR?aided Algorithm improves detection performance of linear detection algorithm in large dimensional MIMO Systems.However,the new problem is that the generation of the constellation set is not the original constellation set after reduction.In order to improve this problem,this paper introduces the Element-based Lattice Reduction?ELR?aided Algorithm based on convex optimization?C-ELR-MMSE?.Firstly,this new problem is transformed into an unconstrained convex optimization problem by a series of transformations,and then solved by KKT condition in convex optimization method.Finally,the simulation results of MMSE algorithm,LLL-MMSE algorithm,ELR-MMSE algorithm and the C-ELR-MMSE algorithm proposed in this paper are simulated by MATLAB simulation software in different dimension and different number of transceiver antennas.It is shown that the C-ELR-MMSE algorithm can significantly improve the bit error rate performance of higher-dimensional MIMO systems.Especially in the case of 4QAM modulation in 16?16 MIMO systems,the algorithm is about 3dB higher than the ELR algorithm at 10^-4 bit error rate.