Research On Key Techniques Of Joint Processing In Multi-antenna Signals Reception Over Fading Channels

Channel fading is an important factor of restricting the increase of the communication rate,and reducing the effectiveness and reliability of the communication system.With the development of the society,wireless communication techniques have also obtained rapid development.The complex electromagnetic environment and the increasing data rate have made the channel fading effect more and more serious,which brings great challenge to the communication receivers.The demand of efficient and reliable processing techniques with lower signal processing threshold and higher processing performance becomes more and more urgent.As an important approach to cope with the channel fading and improve the signal processing performance,joint receiving and processing techniques for communication signals have received wide attention,and become an important research direction in the field of signal processing in recent years.Within the joint receiving and processing schemes,multiple signal processing steps are jointly considered by making use of the relationship between different signal processing levels or multiple data streams,to provide better system performance than traditional hierarchical or non-joint processing structures.In this paper,the main research is focused on the joint signal processing techniques in multi-antenna arraying systems.The analysis of joint parameter estimation and symbol detection for multiple signal streams in flat fading channels,blind channel estimation,joint channel estimation and symbol detection,and channel equalization with unknown time delays in frequency-selective fading channels are conducted by the use of Bayesian learning theory and methods.The main work and innovation of this paper are summarized as follows:1.For the problem of joint parameter estimation and symbol detection for multi-antenna signals with channel parameters difference over flat-fading channels,a new joint processing scheme is proposed based on the variational Bayesian?VB?method.The proposed scheme directly uses multiple received signals for the estimation of information symbols,restraining the information loss in conventional decoupled scheme of signals combination and demodulation.The problem is modeled as the joint maximum a posteriori?MAP?estimation of information symbols,time-delays,complex channel gains,and noise powers,given multiple observations,and approximately solved by means of VB approach.Based on the criterion of minimum relative entropy,analytical-form of the approximate distributions,i.e.,variational distributions,for all unknown parameters are derived.There is no need to determine accurate point estimates of the parameters.Instead,the proposed scheme proceeds iteratively by alternating between the variational distributions of channel parameters and the information symbols.Simulation results show that the proposed joint processing scheme has significant performance improvements in comparison with conventional decoupled or partly joint processing schemes especially with large array sizes and short signal lengths.2.For the problem of channel estimation in multi-antenna arraying systems,an iterative blind channel estimation algorithm is proposed based on the sparse Bayesian learning?SBL?approach.A linear model is built up by using the cross relation property between the received signals and the corresponding multipath channels.And,the channel estimation problem is transformed into the computation of a linear model.Then,the inherent sparse nature of wireless multipath channels is exploited by introducing heavy-tailed conjugative priors in hierarchical form,based on which the sparse Bayesian learning method is applied for obtaining MAP estimation of the channel coefficients in an iterative manner.Closed-form estimation formulations for both channel parameters and hyperparameters are derived.Under the condition of sparse channels,the proposed scheme does not need to know the exact order of the channel coefficients with the aid of sparsity-promoting priors.It is shown that the proposed scheme can provide significant better estimating performance than conventional non-sparse estimation algorithms.Moreover,compared with present l₁-norm regularization method,the performance loss from high mutual coherence of the signal convolution matrix in high signal-to-noise ratio?SNR?values can also be avoided.3.For the problem of multi-antenna signals combination and equalization in frequency-selective fading channels,an iterative approach for frequency domain equalization and combination is proposed based on the maximum likelihood?ML?estimation model for incomplete data set.Different from the traditional decoupled processing of channel estimation and equalization,all the unknown channel parameters are denoted as missing information,and processed jointly with information symbols under the expectation-maximization?EM?framework.Closed-form expression of the equalization output is obtained,which shows that the problem of signals equalization and combination in multipath channels is converted to the weighted summation of each discrete-frequency signals,eliminating the need of complicated sequence estimation or convolution operations demanded by time domain equalization.The proposed scheme operates in an iterative manner.In each iteration,the equalization outputs and joint conditional posterior of the channel parameters are updated in turn.Simulation results show that the proposed scheme has significant better performance than conventional blind channel identification method and decision directed aided spatial diversity blind equalization method.Moreover,the proposed scheme provides a lower computational complexity than a typical joint channel estimation and symbol detection method,and has lower symbol error rates?SERs?with short signal lengths.4.For the problem of symbol spaced equalization's sensibility to timing errors in frequency-selective fading channels,a new frequency domain equalization framework is proposed based on joint symbol synchronization and channel estimation.The proposed scheme operates directly on the oversampled matched filter output,eliminating the need of fine timing parameter estimation and samples recovery.Firstly,based on the ML estimation model of incomplete data set,an iterative linear equalization?LE?scheme is proposed with the aid of EM algorithm,where unknown channel and timing parameters are denoted as missing information.In each iteration,ML frequency domain equalization is derived first,conditioned on which,joint posterior distribution of the channel and timing parameters is obtained using variational Bayesian approach.By the discretization of timing parameters,closed-form parameter estimation formulations based on the weighted summation of oversampled matched filter outputs are derived.Moreover,an iterative frequency-domain block decision feedback equalization?IBDFE?scheme is presented,to solve the performance degradation of linear equalization due to channel deep fading.Finally,both LE and IBDFE schemes are extended to the applications of multi-antenna arraying systems.Simulation results show that the proposed scheme has better system performance than conventional symbol-and fractionally spaced equalization algorithms.And,the proposed scheme has significant better system SER performance in relative low time intervals than two-dimentional traversal based joint channel and timing parameter estimation method.