Study On DOA Tracking Algorithm For Array Signal Based On Random Finite Set

In general,the Direction of Arrival(DOA)estimation is to estimate the source Angle at a certain moment by using the data received by the sensor,which has an important application in many fields such as mobile communication.But in the actual monitoring scenario,the detected information source is usually on the move,so real-time DOA tracking is required.However,the existing DOA tracking algorithms all use the covariance matrix obtained from multiple consecutive snapshots,the received signal is affected by impulse noise and it is difficult to accurately estimate.Therefore,starting from the single snapshot measurement and acoustic vector impulse noise model,this paper applies the stochastic finite set theory to the field of time-varying DOA tracking,and the application of this method has important research significance for time-varying DOA tracking.Therefore,this paper applies stochastic finite set theory to DOA tracking problem from two perspectives,and the research contents are as follows:1.Aiming at the DOA tracking problem of multiple maneuvering sources in impulsive noise environment of acoustic vector array,and an interactive multiple-model multiple model with fractional low-order moment multi-bernoulli filter algorithm is proposed.The algorithm first introduce interactive multiple model(IMM)into the multi-Bernoulli filter to predict the multi-maneuver sources.Secondly,using the α stable distribution to model the impulse noise environment,and use the FLOM matrix eigen decomposition to construct the noise subspace to form the MUSIC pseudo-likelihood function based on the spatial spectrum,and it is realized by particle filter.Experimental investigation shows that this method is very good for tracking maneuvering targets under impulse noise,and can track target state and estimate target number effectively.2.Aiming at the multisource time-varying DOA tracking problem in the case of single snapshot in array signal processing,a multi-Bernoulli filter DOA tracking algorithm based on single snapshot spatial smoothing is proposed.Firstly,the Bernoulli random finite set is used to characterize the randomness of the state process,and get a single snapshot measurement directly from the sensor array.Secondly,the spatial smoothing technique is used to process the single snapshot measurement.And the pseudo-covariance matrix is obtained and singular value decomposition is performed.Finally,the MUSIC spectrum function is used as pseudo-likelihood function for multi-Bernoulli DOA tracking.Simulation results show that the proposed algorithm can effectively track the DOA status of time-varying sources in real time and estimate the number of sources accurately under single snapshot measurement.3.DOA tracking of multisources using an array of sensors is a well-known problem in signal processing.Generally,superpositional measurement model can be used to describe the relationship between the multisource DOA and the signal received by the sensor array.Since the superpositional measurement is the sum of the signals generated by the unknown time-varying number of targets in the monitoring scene.It is difficult to obtain an analytical solution of the likelihood function of the standard multi-Bernoulli filter algorithm.So we propose a superpositional measurement multi-Bernoulli filter for multisource DOA tracking.In this algorithm,each Bernoulli component is defined as a conditional PHD.And the pseudo-likelihood function of the multi-Bernoulli conditional PHD update is derived.Thus the parameter set of the multi-Bernoulli filter is updated and then the multisource multi-Bernoulli DOA tracking is realized by the auxiliary particle filter.The advantage of this algorithm is that it does not need to process measurement information and to know the number of signal sources during tracking.This algorithm directly uses the prediction prior and current measurement information to track DOA in real time.The simulation results show the effectiveness of the paper proposed algorithm.