Study Of Spatial Modulation In MIMO Wireless Communication System

With the development of wireless communication technique,energy efficiency has attracted more and more attention nowadays as well as spectral efficiency in practical systems.Multiple-input-multiple-output(MIMO)transmission technique achieves high spectral efficiency by simultaneously transmit symbols with multiple antennas,while appears problems such as complicated design structure,high implement cost,etc.Spatial Modulation(SM),as an arising green and low-complexity MIMO transmission technique,achieves both high energy efficiency and spectral efficiency by conveying spatial informantion through different activated antennas.Its flexible design and simple implements,especial in its general form Generalized Spatial Modulation(GSM),attracts worldwide research attentions in wireless communications.Due to its special transmission structure,multiple techniques in spatial modulation such as transmit signal design and detection in receivers,appears huge difference compared with those in conventional MIMO system.For the aim of better improving the performance of this new MIMO technique so that it can be adjusted to practical application,this thesis considering the special structure of spatial modulation system and the need of present wireless communication,focused on studies of two main aspects: low-complexity detection and effective precoding design,with contributions as follows:1)Focusing on the problem of overhigh-complexity in all existing detectors in GSM-MIMO systems,in this thesis we proposed a low-complexity near-ML detection algorithm.Considering the special sparse structure and sparsity in GSM signals,we modeled the signal nonzero distribution characteristic based on Bayesian Compressive Sensing(BCS)theory and located the activated antennas,then re-estimated the TAC index set with a so-called PAPR principle and self-corrected the possible estimation error.We controlled and reduced the computational complexity during the process.This thesis quantized the computational complexity of typical detection algorithms and simulated the BER performance in several cases,which validated the effectiveness and practicability of the proposed algorithm.2)Focusing on the deficiency of existing precoding design in SM-MIMO systems,in this thesis we proposed a low-complexity effective precoding algorithm.We theoretically proved the computational redundancy of existing MMD precoding algorithm and reform the optimization goal to shrink the cvx dimensions,so that it can significantly reduce the required complexity at the meanwhile achieve equivalent error improvement performance.3)To complete the precoding design in GSM-MIMO systems,in this thesis we proposed a full-matrix precoding algorithm that could be utilized in any generalized spatial modulation systems.Under the premise of retaining the system transmission structure,we designed full-matrix precoding structure by reconstructing the formulations of channel matrix and signal vector and provided a general expression for precoding design in all spatial modulation systems.This algorithm breaks the limit of diagonal precoder design and achieves significant error improvement performance with acceptable complexity.This thesis thoroughly analyzed its design characteristic and performance advantages and validated those conclusions through simulation.At the end of this thesis,we summarized the above works and put forward some research directions for future improvement.