Parameter Estimation Of Array Radar In Alpha Noise Environment

DOA estimation is one of the important research contents in the field of array signal processing.It is widely used in navigation,communication,sonar detection,radar monitoring,and sound source positioning of microphone arrays.Traditional angle estimation algorithms are based on Gaussian noise as the background for research.But due to the influence of ground clutter or man-made factors,the noise often has impact and long tail characteristics,they conform to S?S,and called alpha noise.In the alpha noise environment,the performance of traditional signal processing algorithms based on second-order and higher-order will be severely degraded or even invalid.At present,the far-field angle estimation algorithm under the alpha noise has been widely proposed.But in some special occasions,such as using a microphone array to mesasure sound source,the measured signal source may be near-field signals or far-near field mixed signals.Meanwhile,it is necessary to estimate the source angle and near-field source distance.Then,the previous algorithm will no longer be effective.But the existing near-field source positioning under the alpha noise has the disadvantage of only suppressing weak impulse noise or high complexity;And the mixed source parameter estimation under alpha noise also has the problem of using subjrctive criteria to distinguish the target source type and calculating complexity.The article will conduct research from these two aspects,the specific content includes:First,the differences between near-field and far-field signal sources are introduced separately.Establish a model of symmetrical uniform array antenna receiving signals.Describe super-resolution angle estimation algorithm,including MUSIC and ESPRIT.Explained the S?S theory,the definition and properties of alpha noise and GSNR,laid a theoretical foundation for subsequent algorithm research.In addition,three suppression algorithms for suppressing alpha noise are analyzed,including FLOM,PFLOM and MCC.Mainly introduce the basic principles,analyze and explain realization process,compare and verify through simulation,to pave the way for subsequent research.Then,research on the near-field target setting algorithm under the background of alpha noise.Combine MCC with the dimensionality reduction MUSIC.The algorithm first uses MCC to obtain the covariance matrix,then reduces the dimensionality of the steering vector,and combines the noise subspace to construct a new spectral search function to estimate the angle,and finally applies the least square rule pairs the distance and angle.It can effectively suppress alpha noise and accurately estimate the target angle.Because the dimensionality reduction MUSIC algorithm is compared with the traditional near-field source positioning algorithm,only one search is required,which reduces the amount of calculation to a certain extent.Finally,research mixed source location algorithm.A propagation algorithm based on the covariance difference is proposed.It first uses the array receiving covariance matrix to obtain the angle of the far-field source,then using the covariance difference idea to separate near-field source covariance matrix,and introducing the propagation operator to obtain near-field target parameters.It not only accurately estimates the target position,but also reduces the eigendecomposition of the matrix.When the environment contains alpha noise,creating a covariance matrix based on PFLOM,and combining the proposed algorithm for target location.Through the algorithm complexity,GSNR,the number of snapshots and other parameters on the estimation error to carry out simulation comparison verification.It shows that the algorithm can better suppress the alpha noise,and estimation the far-field signal and the near-field signal separately,the parameter estimation accuracy is higher.