Spherical In Wireless MIMO Detection Algorithm And FPGA-based Implementation

Multiple-input multiple-output (MIMO) system is a challenging research topic, which caneffectively solve the problem of shortage of communication channel capacity.The MIMOcommunication system has a high frequency spectrum utilization, high transfer rate can beobtained in a certain frequency band, and theoretically is proportional to the number of antennasof the MIMO in the same transmit power and transmission bandwidth, multi-antenna systemsthan single antenna The system’s channel capacity by40times. The receiving end of the signaldetection part MIMO system is the most important part of the entire MIMO system, signaldetection section can be done accurately detect the BER performance of the system will greatlyenhance system reliability and spectrum eiffciency, also can be improved.In the MIMO detector, Spherical detection algorithm is a bit error rate performance close tothe best ML detection, and can also signiifcantly reduce the effective methods of the complexity.Spherical Detection Algorithm for MIMO systems are mainly of two types, one is depth-ifrstalgorithm, the algorithm can achieve the same performance with ML，the drawback is that thenumber of iterations in the search process more throughput is not high and complex is not ifxed，such as SESD (Schnorr-Euchner SD) algorithm list ball algorithm (LSD). The other is thebreadth-ifrst algorithm, algorithm performance relative to the ML, some loss，but the popularfeedforward structure, has a hardware-friendly features in each layer of the detection signal,leaving only the smallest of the K metric corresponding to the path to determine the expansionconditions, such as K-Best Algorithm.This paper presents an improved K-Best detection algorithm and its implementation, whichis based on MIMO detection algorithm theory and its implementation, to combine and improvethe proved algorithms. The algorithm uses a combination of prediction techniques and Parallelsort method, reduces the computational complexity; adopts parallel pipelined structure to saveprocessing time.MIMO detection algorithm for MIMO detector design，in-depth research andanalysis of the K-Best detection algorithm based on hard decision can be drawn, the complexityof the algorithm is lower, while the performance at a better level. The the MIMO detectorproposed structure as well as improved K-Best algorithm research MIMO detector certainreference value.In this paper, For the realization of the hardware detection algorithm, the MIMO detectordesigned has carried for comprehensive implementation, based on the FPGA platform in Xilinxcompany, In order to verify the detector can work correctly, the detector is simulated in thecondition of Modelsim environment, the ultimate proof, the detector designed in this paper canmeet the design requirements.