Research On Array Parameter Estimation Methods Based On Space-Time Information Fusion

With the continuous development of science and technology and the improvement of communication network,the demand of target signal location information is increasing.Array signal processing can realize super-resolution location parameter estimation,which is widely used in military and civil fields.However,with the increasing of signal bandwidth,it is necessary to study a special array location parameter estimation method for the characteristics of wideband signals.At the same time,the array based parameter estimation is gradually developing from direction of arrival(DOA)estimation to multi-dimensional parameter joint estimation,providing more accurate information for the location of the target.Time delay is one of the important location parameters.The joint estimation of time delay and direction of arrival can effectively improve the estimation accuracy and reduce the number of nodes in the system.It has advantages in performance and efficiency,and becomes the research focus of array-based location parameter estimation.Therefore,it is of great theoretical significance and practical value to study the joint estimation method of space-time parameters for wideband signals.At present,the theory of joint estimation of space-time parameters of wideband signals based on array is not complete,and the algorithm is not suitable for coherent multipath channel and colored noise channel.In order to solve the above problems,firstly,from the perspective of signal characteristics,broadband signals are divided into wideband OFDM signals and general wideband signals.Then,the Joint DOA and delay estimation of wideband signals in coherent multipath and color noise channels is studied.The corresponding high-precision parameter estimation methods are designed to effectively solve the influence of signal coherence and color noise interference on parameter estimation.The main contents include:1.The joint estimation method of space-time parameters for wideband OFDM signals is studied.To solve the problem of the failure of existing algorithms in coherent multipath channels,a fast DOA and time delay joint estimation algorithm based on uniform linear array is proposed.Firstly,the algorithm uses the characteristics of multi carrier modulation(MCM)of wideband OFDM signal to narrow-band processing,combines the frequency-domain response of all carriers,constructs the frequency-domain response model of the extended channel;Then it smoothes the data of each carrier to obtain the smooth covariance matrix;Finally,TLSESPRIT and one-dimensional peak search algorithm are applied to DOA and delay respectively,and the estimated parameters are obtained.This method can decoherence by smoothing,and realize the joint estimation of parameters in coherent multipath channel.Because the extended channel frequency response is constructed,it is equivalent to virtual expansion of the physical array,which effectively improves the accuracy of parameter estimation.In addition,through the complexity analysis of the proposed algorithm,the complexity of the algorithm is far lower than the multidimensional parameter spectrum search method.2.The joint estimation method of space-time parameters for general wideband signals is studied.Considering that the frequency response of the channel cannot be obtained from the general wideband signal,and the complexity of the existing array multi-dimensional parameter estimation methods is relatively high,a frequency-domain smooth fast DOA and time-delay joint estimation algorithm based on the uniform linear array is proposed.Firstly,this algorithm constructs the general wideband signal time domain receiving model under the uniform linear array,and then establishes the frequency domain receiving model through the discrete Fourier transform(DFT);Then it smoothes the frequency domain receiving model to obtain the smooth covariance matrix;Finally,the TLS-ESPRIT method is applied to realize a joint estimation algorithm of DOA and delay parameters automatic matching.Through the complexity analysis,we can see that this method can reduce the dimension of joint parameter estimation by solving the closed solution of DOA,avoid the multi-dimensional parameter spectrum search,and significantly reduce the computational complexity.At the same time,frequency domain smoothing can solve the rank defect problem of covariance matrix caused by signal coherence and realize joint parameter estimation in coherent multipath channels.The simulation results verify the effectiveness of the proposed algorithm in coherent multipath channels,and show that the algorithm has robust performance in low SNR scenarios.3.The joint estimation of space-time parameters of wideband signals in colored noise channel is studied.The existing parameter estimation methods usually assume that the noise has the characteristics of Gaussian white,but the color noise actually exists in the wireless channel.To solve the problem that the performance of joint estimation algorithm is seriously deteriorated under the interference of color noise,a joint estimation algorithm of DOA and time delay based on iterative noise suppression and smoothing in frequency domain of uniform linear array is proposed.Firstly,the received signal model in time domain is established in the channel of colored noise,and the frequency domain model is established by DFT transform;Then,the iterative noise reduction method is used to estimate and cancel the color noise components to eliminate the influence of color noise;After that,the frequency domain model after denoising is smoothed to get the smooth covariance matrix;Finally,the two-dimensional spectrum function of DOA and delay is constructed to realize the joint estimation of parameters.The simulation results show the effectiveness of the proposed algorithm in the colored noise channel,and the accuracy is higher than that of the coherent signal subspace method.At the same time,the rank of covariance matrix of coherent signal can be recovered by frequency domain smoothing,which makes it possible to establish two-dimensional spectrum function.Simulation experiments also show that the algorithm can be applied to coherent multipath channels.