Reserach On Signal Detection Algorithms Base On Spatial Multiplexing In LTE System

LTE (Long Term Evolution) is a mainstream mobile communication system in the future. LTEsystem uses multiple-input multiple-output (MIMO) technology. The high speed data transfersfrom different antennas which is broken into a plurality of low speed sub-data streams, whichgreatly improves the system capacity. The signal received at each antenna on the receiving side isthe superposition of transmission signal in the frequency domain and time domain. Theperformance of the detection algorithm is an important factor for the actual communication system.Base on theoretical research and Simulation of MATLAB, this paper proposed bi-directionaldetection MMSE algorithm and a new sphere decoding detection algorithm which reduces thecomplexity of the system. Both of the two improved algorithms improve the performance oforiginal algorithm.The main work of this paper includes:Firstly, proposed a bi-directional detection MMSE detection algorithm base on interferencecancellation MMSE detection algorithm in LTE system. In traditional MMSE algorithm forinterference cancellation detection, different layers have different reliability, the first detectionlayer has the lowest reliability; the last detection layer has the highest reliability. In this paper, theBi-directional detection MMSE algorithm detects from the last detection layer to the firstdetection layer, retain the value of the front layers,then detect from the first detection layer to thelast detection layer, the value of the following layers should be reserved, at last, consolidate theobtained results. The simulation results show that this algorithm has better bit error rateperformance and system capacity than interference cancellation MMSE detection algorithm andimproves the system capacity.Secondly, proposed ZF-SD detection algorithm to reduce Sphere Decoding algorithm’scomplexity. ZF-SD detection algorithm uses a fixed threshold as the reliability criterion tocombine the Zero Forcing algorithm and the Sphere Decoding algorithm. Firstly it uses the ZeroForcing algorithm to detect the transmitted data, and then determine the reliability of the ZeroForcing data. If the Zero Forcing data has high reliability and it will be adopted as the final result.If the data is less reliable then it will use the Sphere Decoding algorithm, while taking advantageof Zero Forcing detection results to select the initial radius of the sphere decoding algorithm. Thesimulation shows that in4x4antennas error rate of the algorithm basically unchanged whilegreatly reducing the complexity of the Sphere Decoding algorithm.Finally, base on the analysis of the first two chapters, this paper uses GUI of MATLAB softwarebuild a detection algorithms of LTE system software platform. This platform can get thetheoretical BER curves of LTE detection system under different conditions such as select the channel conditions, chose different modulation schemes, and change the number of transmit andreceive antennas. In the end, this article use specific examples to demonstrate the implementationprocess of this software platform and the simulation results. Using this simulation platform cansimplify the optimization parameters; You can not only achieve the human-computer interaction,but also provides a great convenience for the follow-up study.