Research On Low Complexity Decoding Algorithms On OVXDM With High Spectral Efficiency

Various mobile Internet services such as virtual reality,ultra high-definition 3D video,mobile clouding provide life-like user experiences to mobile users.In the meantime,these services bring urgent demand for ultra-high rate transmission.However,since available spectrum resources are limited,it is difficult to achieve the requirements of a large number of connections,ultra-fast speed,ultra-high reliability,and ultra-low latency for future wireless communication based on present communication theory and technical framework.In that context,a subversive innovation encoding which is called overlapped x domain multiplexing(OVXDM)is innovatively proposed by Prof.Li Daoben.OVXDM is a novel encoding scheme which adopts multiplexing waveform to carry transmitted information with shifting and weighted transmission symbols in x domain(time or frequency domain)to obtain high spectral efficiency and coding gain.Back in 2006,the principle of OVXDM encoding discovered by Prof.Li Daoben indicates that the overlapping between adjacent symbols is not just interference but a beneficial coding constraint relation offering coding gains.Obviously,OVXDM subverts the design concept which must follow the Nyquist criterion in the traditional communication theory.Though maximum likelihood sequence detection(MLSD)decoders can obtain the optimal decoding performance,their computational complexities which increase exponentially with spectral efficiency in OVXDM are unbearable for practical implementations.Therefore,this thesis studies low complexity decoding algorithms and novel encoding structure which can achieve high spectral efficiency in OVXDM encoding.The main research work includes the following aspects:(1)Coding gains of OVXDM can be obtained by shifting and overlapping transmission symbols with multiplexing waveform.This thesis firstly analyzes the impact of different multiplexing waveforms on the performance of OVXDM encoding under additive Gaussian white noise(AWGN).Then combined with the characteristics of OVXDM encoding structure,a multi-bit sliding stack decoding(Multi-Bit SSD)algorithm is proposed,whose performance is analyzed in consideration of sliding window size.Simulation results show that compared with existing fast algorithms,the proposed Multi-Bit SSD algorithm can achieve 1?3dB improvement in decoding performance and has lower decoding complexity.(2)Considering the practical application of OVXDM encoding technology,this thesis firstly studies the performance of OVXDM coding in fading channel,then proposes a multiple signals joint detection(MSJD)algorithm with low complexity.Moreover,the relationship between performance of the proposed algorithm and number of received signals under different OVXDM encoding parameters including multiplexing waveform and overlapping fold is indicated by theoretical analysis.And a method is provided to select the proper number of received signals for decoding to optimize decoding performance.Simulation results show that the proposed MSJD algorithm can achieve 5?10dB decoding performance improvement over the existing fast decoding algorithms with lower complexity.(3)Turbo and multi-carrier techniques are widely used to improve spectral efficiency in wireless communication systems.If OVXDM encoding technology can combine with them,the spectral efficiency of OVXDM can be further improved in theory.With this in mind,two novel encoding structures of OVXDM are proposed in the thesis.Through interleaving two identical OVXDM encoders in parallel concatenation,a Turbo-Like OVXDM encoding structure is proposed,and a maximum likelihood based iterative decoding(MLID)algorithm is proposed correspondingly.Moreover,combined with multi-carrier technology,a multistream OVXDM structure is proposed.In order to realize the multistream OVXDM signal decoding,a two-dimensional Trellis graph search algorithm is firstly proposed based on the state transition theory of time-frequency domain signals under Trellis graph proposed by Prof.Li Daoben.Secondly,to reduce decoding complexity,a minimum mean square error based decision feedback equalization(MMSE-DFE)algorithm is proposed.Simulation results demonstrate that performance of two proposed encoding structures is obviously better than that of the traditional OVXDM.Two-dimensional Trellis graph search algorithm can achieve the optimal performance in multistream OVXDM.And MLID and MMSE-DFE algorithms have 2?5dB performance improvement and lower complexity than traditional fast algorithms.