Research On Link Adaptation Technology And Detection Algorithm Based On MIMO-SM Systems

Multiple antenna spatial modulation?SM?is a newly proposed technique based on Multiple-input multiple-output?MIMO?transmission techniques in recent years.This technique activates only one transmitting antenna during every transmitting timeslot to transmit data.For the input information bits,one part is mapped to active antenna position which corresponding to the spatial dimension,the other is mapped to conventional amplitude phase modulation?APM?symbols which are transmitted by the active antenna.By using antenna's active position status as the carrier of a part of transmitting information,SM technique can effectively avoid some problems in the conventional MIMO techniques,such as inter channel interference?ICI?,inter-antenna synchronization?IAS?and the costs of radio frequency?RF?link.Therefore,SM technique will have a good application prospect.The researches of this thesis are based on multiple antenna spatial modulation techniques and the main contents and innovations are as follows:1.The theory foundations of MIMO-SM technique are expounded.At first,this thesis tries to analyze and compare the rationale of MIMO system and SM system and point out the advantages of SM technique compared to MIMO techniques.Second,this thesis introduces the link adaptation technology of adaptive spatial modulation?ASM?system and analyzes the optimal adaptive algorithm's rationale which is used for choosing the optimal candidate modulation scheme of ASM.Third,this thesis introduces the SM signal detection algorithms,mainly analyzes the characteristic of SM signal demodulation and introduces four conventional SM signal detection algorithms.2.Aiming at the high costs of computational complexity and lots of redundant computing of the optimal ASM adaptation algorithm,an improved adaptation algorithm which is fit for M-ary phase shift keying?MPSK?modulation is proposed,called MPSK-EAQ algorithm.By using the characteristic of constant amplitude of constant envelope modulation,and combining the idea of equal amplitude and quantization?EAQ?algorithm,this algorithm replaces the optimal algorithm on a simple and equivalent computation achieved by transforming the computation of minimum Euclidean distance into the computation of modulation skewing which is sequentially equivalent to the computation of modulation constellation's number.In this way,lots of multiply operations are reduced,and redundant computing of the optimal algorithm is avoided.The complexity of proposed algorithm is analyzed and the performance of bit error rate?BER?with different antenna numbers and different transmission speeds is simulated.Results show that the performance of MPSK-EAQ algorithm is roughly the same as the optimal algorithm in different conditions,but the costs of complexity are much lower than the optimal algorithm.In a 2?2 system and when m=2bit/symbol,the complexity of improved algorithm is just about 20.5%of the optimal algorithm.3.Aiming at the problem of high computational complexity in the optimal maximum likelihood?ML?detection algorithm for spatial modulation?SM?system,an improved detection algorithm with low computational complexity based on QRD-M algorithm is proposed.By combining QR-decomposition theorem and using tree-searching stage of M algorithm which only computes the optimal M branches at each layer,the improved algorithm avoids searching all nodes which not only reduces the computational complexity but also make the hardware implementation simpler.The complexity of proposed algorithm is analyzed and the performance of BER with different M is simulated.Results show that the performance of QRD-M algorithm is close to ML algorithm with lower computational complexity costs.In the conditions of 4?4-MIMO system by using 4-ary quadrature amplitude modulation?4QAM?,when M=4 and BER=10^-2,the performance of QRD-M algorithm differs by 8dB from ML algorithm with about 75%computational complexity costs reduced;and when M=8 and BER=10^-2,the performance of QRD-M algorithm differs by3.5dB from ML algorithm with about 19%computational complexity costs reduced.4.Aiming at the problem that the execution time and detection performance of QRD-M detection algorithm are affected by transmitting and receiving antennas,an improved QRD-M detection algorithm by using the idea of parallel detection is proposed,called Parallel QRD-M?PQRD-M?detection algorithm.By disassembling the tree into M sub-trees and using QRD-M algorithm to search each sub-trees synchronously and independently,the improved algorithm not only reduces the hardware requirements,saves the execute time of algorithm,but also reduces the deletion probability of the optimal point.The complexity of proposed algorithm is analyzed and the performance of BER with different antenna numbers is simulated.Results show that at a cost of tiny amounts of computational complexity increasing,PQRD-M algorithm can observably improve the detection performance of QRD-M algorithm.In the system of 4QAM,whenN_t?N_r=4?4,M=4 and BER=10^-2,the performance of PQRD-M algorithm increases 3dB compared to QRD-M algorithm;and whenN_t?N_r=8?8,M=8 and BER=10^-2,the performance of PQRD-M algorithm increases 7dB compared to QRD-M algorithm.