Research On Direction Of Arrival Estimation Technology For Mixed Signals In Multi-antenna Systems

Direction of arrival(DOA)estimation,as a key part of array signal processing,is the core technology of the detection system.Traditional DOA estimation algorithms usually only target a single type of signal.However,the signal leakage of hardware devices in a multi-antenna system and the multipath effect in the environment will lead to the mixed existence of incoherent and coherent signals.The mutual interference of the two types of signals leads to the existing method DOA estimation.Technical performance is significantly reduced.In addition,with the development of large-scale arrays,the received signal is a spherical wave,and the position and DOA must be jointly estimated,which makes the complexity of the estimation algorithm increase sharply.To this end,this thesis first analyzes the advantages and disadvantages of the current DOA estimation algorithm,facing the problem of far-field mixed signal and near-field signal estimation,systematically studies the internal reasons that restrict signal DOA estimation,and then proposes a new step-by-step algorithm for farfield and near-field Formula composite DOA estimation algorithm.For farfield mixed signal DOA estimation,the estimation time is reduced by more than 50%on the basis of ensuring the estimation accuracy to support the fast estimation of far-field signals;for near-field sources,the estimation time is reduced on the basis of slightly improved DOA estimation accuracy More than 20%,in order to meet the demand for fast positioning of nearfield sources in practical engineering.The specific content includes:1.The structure of the multi-antenna DOA estimation system is studied,and the basic model of the far-field narrowband signal under the uniform linear array and the derived coherent signal model are derived.Based on the derived model,the basic principles of the beamforming algorithm,the subspace estimation algorithm and the decoherence algorithm are analyzed,and the simulation experiments are compared for the above algorithms,and the applicable scenarios of different algorithms are analyzed from the obtained results.2.Aiming at the mutual restriction of high-precision and real-time DOA estimation caused by the mutual interference of far-field mixed signals,this thesis proposes a step-by-step composite estimation algorithm based on the combination of spatial difference matrix and sparse reconstruction.Firstly,the DOA information of incoherent signals is obtained by spatial spectrum estimation based on the mixed signal receiving model.For the DOA information of the coherent signal,firstly,the spatial difference processing is performed on the original receiving matrix to obtain the spatial difference matrix containing only the coherent signal information,and the interference of the non-coherent signal and noise is filtered out.Then,according to the idea of sparse reconstruction algorithm,aiming at the sub-array division method of constructing spatial difference matrix,combined with spatial smoothing algorithm to construct a new over-complete dictionary,and then combined with Orthogonal Matching Pursuit(OMP)to recover the coherent signal DOA information.Executing the two steps in parallel reduces the time complexity of the overall estimation.The simulation results show that the proposed algorithm is superior to the spatial spectrum estimation algorithm and the traditional sparse reconstruction algorithm in estimation accuracy,and under the premise of practical application,the estimation time is reduced by 50%compared with the spatial spectrum estimation and the traditional sparse reconstruction algorithm.More than%,it is more suitable for practical engineering.3.Aiming at the problem that the traditional DOA estimation algorithm uses too much time for near-field source location,this thesis proposes a composite algorithm combining Rank Reduction(RARE)algorithm and sparse reconstruction.Using the second-order statistical information of the signal,only one-dimensional search is carried out for the angle information of the near-field source,and the angle information of the source is obtained first.Then use the angle information as a known parameter to construct an over-complete dictionary for near-field source estimation,and combine the OMP algorithm to restore the source to obtain the distance parameter of the source.This process completes the automatic matching of parameters and reduces the time consumption of the matching process.The simulation verification shows that the estimation accuracy of the algorithm is improved compared with the traditional near-field DOA estimation algorithm,and the estimation time is reduced by more than 20%compared with the traditional algorithm,which has a significant improvement and is more suitable for the location of near-field sources need.