The Research Of Array Signal Processing Methods In Complex Environment And Its Application

Compared with traditional radar,phased array radar is more capable in agile beam control by electronically controlling its beam forming,which meets the demand of detecting while tracking in multiple areas or upon multiple targets.Furthermore,the aperture synthesis effect in phased array radar largely increases the detection range and cross-range resolution of radar.Therefore,the phased array radar has already become one of the important development trends for modern radar.Because the introduction of a new space dimension,the array signal processing applied in phased array radar could achieve higher target angle resolution and the ability to suppress interference and clutter.Therefore,the research on array signal processing is of great theoretical and realistic significance in radar system design.This paper studies the application of array signal processing methods in realistic complex environment in order to improve the performance of array signal processing in these complex environment.Although the array signal processing techniques could bring phased array radar higher precision in target estimation and the ability to suppress the interference and clutter,the realistic complex environment in which the radar works in raises new problems for the current array signal processing methods.The paper will investigate the application of array signal processing in complex environment from three aspects.1 For the non-ideal factors in the phased array radar system,the paper will study array signal angle estimation methods that are robust against two of the most common non-ideal factors.The difficulty lies in the complexity of the estimation of coefficients in the non-ideal models.Therefore,the paper will study the array signal angle estimation methods that are not dependent on specific non-ideal models.2 To meet the demand of phased array radar in adaptive interferencesuppressing,the paper will study accurate and real-time estimation methods to acquire target angle from interfered signals.The difficulty is to solve the problem that adaptive interference suppression technique would deteriorate the precision and accuracy of array signal angle estimation.3 The paper will introduce array signal processing technique into the study of blind expert's echolocation system.Specifically,the paper will use the array signal processing and the perception-action cycle from the blind expert on a two-sensor array to recognize the target shape.The difficulty is to establish the model of signals used by the blind and find the target feature that can be used by the blind for target shape recognition.The main content of the paper is introduced as below.In the presence of spatially correlated noise and array mutual coupling,the performance of signal estimation algorithm would deteriorate to a large extent.Therefore,the paper has proposed a robust signal estimation algorithm against the spatially correlation noise and mutual coupling.To improve the estimation in the presence of spatially correlated noise,the paper proposed a signal angle estimation algorithm called sub-array weighting UN-MUSIC.The algorithm uses the characteristic of the noise that it is not correlated between sub-arrays,and by the general correlation decomposition of the signal mutual covariance matrix of sub-arrays,the algorithm could get the noise subspace of each sub-array.Then,through mathematical derivation based on the minimum variance principle,the paper proposes a signal angle estimator that optimally combines the two noise subspaces and the performance analysis of the proposed estimator is given.Mathematical simulation shows that compared with traditional array signal estimation methods,the proposed algorithm gives significant performance improvement.Finally,with the study of array mutual coupling models,the paper proposes a modified sub-array weighting UN-MUSIC algorithm which is applicable in the presence of both spatially correlated noise and array mutual coupling.The mathematical simulation results show that the proposed method can outperform the traditional estimation methods in the presence of the non-ideal factors mentioned above.Radar targets are usually accompanied by interference.The paper investigates fast and real-time angle estimation methods in the presence of strong suppressing interference near the target.The paper proposes a novel method that combines the adaptive interference-suppressing technique and monopulse technique.Specifically,the new method proposes a robust adaptive difference beam forming algorithm that can compensate for the distortion brought by adaptive interference-suppression,which guarantees the linearity of monopulse ratio upon target deviation angle.The paper gives detailed derivation of the proposed estimator's mean and variance and demonstrates the advantages over other traditional methods.The paper also analyzes the computational cost of the proposed method and shows that the proposed method can meet the demand of a real-time system.Because the proposed method utilizes the fastness of monopulse estimation,it can be applied in some radar systems that have very high data rate.The mathematical simulation demonstrates the effectiveness of the proposed method and the advantages over the traditional methods.Some mammals,such as bats,dolphins,can perceive the surrounding complex environment using self-generated sound.They can use this ability to search and recognize preys.This ability is highly desirable in modern radar and sonar system.Recent researches find out that some of the blind people also possess the ability to use echolocation,just like the bats and dolphins.The processing of the signals received by two ears provides essential cues for echolocation.Currently,there has been little report on how the received two-channel signals are processed by the blind in order to achieve target recognition.The paper studies the perception-action cycle in the mammal's echolocation system by experimental cooperation with blind experts and experimental data analysis.From the study,the paper proposes a method that combines array signal processing with mammal's perception-action cycle for the task of target shape recognition.Specifically,the paper first conducts time-frequency analysis on the sound signals used by the blind experts in echolocation and formulates a mathematical model for the signal which is validated by the experimental data recorded from the experts.Then the paper introduces a novel target shape recognition method which uses the channels' signal amplitude difference as the feature and incorporate the information of sonar platform positions.The method has been validated through target shape recognition experiment based on a two-channel microphone array in the anechoic chamber.The proposed method is inspired by the principle of perception-action cycle observed in mammal's echolocation system,which hopefully might provide insights that are useful for the development of array signal processing in the future.Furthermore,the paper has studied the application of array signal processing technique.A design of array signal processing machine based on highly integrated chips is introduced.The design of this signal processing machine has made breakthrough in multi-channel synchronous sampling technique and connection technique for high-speed processing nodes.Using this signal processing machine,the proposed array signal estimation methods in the presence of strong interference is validated by the experimental data collected in the anechoic chamber.