Study On Signal Detection Algorithm In Differential Spatial Modulation System

In recent years,spatial modulation(SM)technique becomes a key research directions of multiple input multiple output(MIMO)technique.SM adopts single RF module,and at any time slot it only activates single transmitting antenna to avoid inter channel interference(ICI).And SM introduces the information of antenna's index as an additional dimension to improve the spectrum efficiency.SM technique requires channel state information(CSI)used for the receiver to detect,so it must be on channel estimation.The process of channel estimation is very complex,and requires a lot of resources,and channel estimation error will cause serious loss of detection performance.In this case,the differential spatial modulation(DSM)technique was put forward,the DSM technique is a novel SM technique with great potential,it retains the advantages of SM technique and avoids the uncontrollable process of channel estimation.DSM technique is applicable to any constant energy constellation,has no requirement for the number of transmitting antenna and receiving antenna.This thesis studies on the signal detection algorithm in the DSM system,and aims to find some detection algorithms with near-optimal performance and low complexity.The thesis found out the maximum likelihood(ML)detection algorithm in DSM system has high complexity,in order to reduce the complexity,a few detection algorithms based on the idea of reduces the number of candidate information blocks and time sharing calculation then combined detection are proposed.(1)In this thesis,the M algorithm is applied to the DSM system,and a detection algorithm based on M algorithm(D-MML)is proposed.D-MML algorithm sees the detection process as a tree search structure,and reduces the complexity by reducing the number of information blocks of each layer of the tree.(2)Based on D-MML algorithm,an improved D-MML(D-IMML)detection algorithm is proposed to further reduce the number of information blocks using the branch and bound technique.(3)On the basis of D-MML algorithm,a dynamic D-MML(D-DMML)detection algorithm is proposed.This algorithm estimates the noise existing in the testing process of DSM then setting a threshold to further reduce the number of information blocks.(4)This thesis found there is a lot of repeated calculation in the traditional detection algorithm of DSM system.A time-division combination(TDC)detection algorithm is proposed.The algorithm separates the information blocks into vectors to calculate,then detect,in order to reduce the computation redundancy.Accoding to the simulation result,the improved detection algorithm provides a substantial reduction in computational complexity compared with the ML detection algorithm while retaining the almost same bit error rate(BER).The TDC algorithm is the optimal performance detection algorithm,and the other three are near-optimal detection algorithms.