Research And Implementation Of Low-complexity Mimo-otfs Receiver Algorithm

OFDM is one of the core technologies of 4G,which has the characteristics of high spectrum utilization,anti-multipath interference and frequency selective fading.However,the development of science and technology has brought new challenges to mobile communication,one of which is the faster and faster mobile communication transmission environment.In such fast-variant transmission environment,traditional OFDM technology can no longer meet the needs of people for reliable communication.As a result,OTFS technology was proposed as a communication solution in high-speed mobile scenarios,which transforms signal processing to the Doppler-delay 7plane and converts the fastvariant channel that transmitted signal undergoes into a time-invariant channel with a constant channel gain.Researchers have demonstrated that OTFS communication systems have better transmission performance than OFDM communication systems in fast time-varying fading channels.However,most of the OTFS detection algorithms have high computational complexity,which is not conducive to the implementation of the algorithm on the ground.It is of great practical significance to study how to achieve a good balance between the performance and complexity of the algorithm.In addition,the MIMO technology is also being widely used in mobile communications.Researches on the combination of OTFS and MIMO technology can not only increase the spectral efficiency,but also enhance system robustness,which is of great significance to the development of mobile communication technology.This paper takes the low complexity OTFS detection algorithm as the research focus,and combines MIMO scenario to work from both theoretical and practical aspects.The theoretical research can be divided into two parts:system basis and low complexity OTFS detection algorithm.In the introduction of system basis,key technologies of OFDM and OTFS systems on both sides of transmission and reception are described in detail,and the performance of glsOFDM and OTFS communication systems,using first-order frequency domain equalization and Message-Passing algorithms on the receiver side respectively,is shown by simulation software,which proves the advantages of OTFS technology in high-speed mobile environment.In the low complexity OTFS detection algorithm,based on the Gauss-Seidel iterative solution of the linear Maximum Ratio Combining algorithm,two optimization ideas are proposed.One is to optimize the initial value estimation.The initial value using the first order MMSE time domain equalization result,Newton’s one-step iteration result and the expansion of the quadratic term of Newman series are analyzed and compared.It is found that adopting the first order MMSE time domain equalization result as the initial value makes the algorithm with better balance between complexity and performance.Algorithm of linear Maximal Ratio Combining is optimized by the Untransformed Preconditioned Conjugate Gradient algorithm.The algorithm is compared with the MP and the Gauss-Seidel algorithms by simulation software.It is proved that the proposed Untransformed Preconditioned Conjugate Gradient algorithm converges faster and has lower computational complexity at the cost of performance loss.Finally,the proposed algorithm is combined with Space-Time Block Code,a space-time encoding method commonly used in MIMO technology,to expand its use scenario.In the aspect of low complexity implementation,this paper takes the domestic DSP SB3500 as the hardware implementation platform,and puts forward two low complexity implementation ideas.One is to use multi-threading on multi-core and coroutine technology to parallelize the system,the other is to use RPU inline function to accelerate the processing of matrix operations and so on.By making use of the computing resources of the hardware platform,the system can run more efficiently.