Research On Equalization And Detection Technology Forsingle Carrier MIMO Systems

Multiple input multiple output(MIMO)technology has been promoted to achieve high-speed data transmission.Compared to the orthogonal frequency division multiplexing(OFDM)based MIMO systems,the single-carrier scheme(SC)has attracted lots of attention due to its low peak-to-average ratio(PAPR)feature and is quite suitable for longdistance transmission.However,one of the significant challenges of SC-MIMO systems is the vast complexity of channel equalization and signal detection,especially for the intersymbol interference(ISI)caused by the multi-path channel.The single carrier frequency domain equalization-based minimum mean square error(SCFDE-MMSE)algorithm is proved to achieve satisfying performance.However,it involves large numbers of DFTs and large-scale matrix inversions,which is unacceptable for practical systems.In this paper,a low-complexity SCFDE-MMSE(LC-SCFDE-MMSE)algorithm is proposed.Firstly,the characteristics of cyclic matrix and DFT cyclic shift are combined to reduce the number of DFTs.Secondly,SCFDE is transformed into a symbol-wise manner to avoid large-scale matrix inversions.Finally,a frequency-domain interpolation method is proposed to reduce the number of small-scale matrix inversions further.According to the evaluation results,the proposed LC-SCFDE-MMSE reduces the complexity of the traditional SCFDE-MMSE algorithm by more than one order of magnitude with less than 0.1d B performance loss.Matrix inversion is the most critical module in MMSE detection algorithm,which has high computational complexity and high hardware resource cost.Therefore,it is particularly important to select an excellent matrix inversion algorithm.In this paper,an improved LDLT decomposition inversion algorithm with acceptable complexity based on Cholesky decomposition is introduced,which avoided the square root operation and transformed the division operation into the inverse operation.Further,an unified triangular systolic array architecture is used to realize the LDLT decomposition inversion algorithm.This algorithm reduces the number of multipliers and adders required by the original algorithm,improves the utilization rate of hardware and ensures that the time cost can satisfied the high speed transmission standard.Taking a 4-th order square matrix as an example,the triangular systolic array architecture only needs half of the multiplier and 80% of the adder of the original algorithm can complete the matrix inversion within 15 clock cycles and satisfied the engineering requirements.Besides,the performance of liner equalization like SCFDE-MMSE is not satisfied.The Gaussian approximate interference based belief propagation algorithm(GAI-BP)has been proposed as a competitive detection candidate,approaching near maximum likelihood(ML)performance with acceptable complexity.However,there still room for further improving detection performance to near maximum a posterior(MAP)using the extrinsic information provided by channel codes.Besides,GAI-BP suffers from quite slow convergence resulted by complex connections on factor graph.In this paper,a joint detection and decoding scheme(JDD)is proposed for LDPC coded single carrier MIMO systems,which improve performance and speed up convergence simultaneously.Moreover,combining three optimization methods,namely dynamic message damping,early frozen strategy,and layered message updating,a fast convergence GAI-BP detection and decoding algorithm(FC-JDD)is proposed to further accelerate the convergence and reduce the complexity.Evaluation results demonstrates FC-JDD algorithm has about 2d B bit error rate performance gains compared to individual detection and decoding scheme.FC-JDD algorithm has 70% Max-iteration number reduction and 66% Mean-iteration number reduction with respect to JDD scheme.