| As a new radar system,Bistatic MIMO(Multiple Input Multiple output)radar has a series of obvious advantages: unique transceiver diversity technology,flexible station arrangement,high anti-interception and anti-radiation missile capability,and high channel aperture utilization ratio.However,in the face of increasingly severe battlefield situation,conventional bistatic MIMO radar is vulnerable to electromagnetic environment interference.The performance of target estimation of conventional bistatic MIMO radar under the background of high power narrowband jamming is reduced sharply.To overcome the interference problem faced by conventional bistatic MIMO radar,a bistatic electromagnetic environment sensing MIMO(EES-MIMO)radar is presented in this paper,it overcomes the problem that the conventional bistatic MIMO radar has weak ability of anti-jamming in the increasingly complex electromagnetic environment by sensing the frequency characteristics of the space electromagnetic environment and avoiding interference band actively,it is compatible with the unique advantages of electromagnetic environment sensing technology and conventional bistatic MIMO radar.This research can be divided into the following sections: the principle and structure of the basic EES-MIMO radar,the establishment of quaternion model of bistatic EES-MIMO radar,the electromagnetic environment sensing method of basic EES-MIMO radar,angle estimation algorithm of bistatic EES-MIMO radar based on quaternion theory,angle estimation of bistatic EES-MIMO radar based on quaternion in colored noise background,angle estimation algorithm of bistatic EES-MIMO radar based on covariance differential in unkonw noise background.First,the working principle and structure of bistatic MIMO radar are described in detail,a bistatic electromagnetic environment sensing MIMOradar is proposed to solve the problem that the target estimation performance of onventional bistatic MIMO radar is reduced sharply under high power narrowband jamming background and the working principle,system structure and signal model of the new radar areanalyzed,which lays a theoretical foundation for further research.Second,aiming at the problem of the bistatic EES-MIMO radar real / complex data model needs to construct an augmented matrix and parameter matching is easy to lead low accuracy of two-dimensional estimation,two quaternion models of bistatic EES-MIMO radar are constructed: the quaternion model of bistatic EES-MIMO radar based on uniform linear array and the quaternion model of bistatic EES-MIMO radar based on bistatic linear array.It avoids all possible error information generated in the subsequent pairing through fractal dimension,and improves the estimation accuracy and resolution of the subsequent signal processing of bistatic EES-MIMO radar.Third,at first analyzing the jamming environment of bistatic EES-MIMO radar,then laying special stress on analysing two typical interferences:point-frequency interference and sweep interference and an electromagnetic environment sensing algorithm based on these two kinds of interference is presented.For point-frequency interference,cross-spectral RI-MUSIC algorithm is applied to estimate the noise power spectrum of bistatic EES-MIMO radar,the method makes full use of the characteristics of aperture expansion of bistatic EES-MIMO radar array and the rotation invariance of ESPRIT subspace.The two-dimensional parameter estimation problem is transformed into two one-dimensional forms,which reduces the computational complexity and system complexity.A quadratic cross-spectral SVD decomposition method is proposed for frequency sweep interference,the interference signal is transformed from non-stationary random process to stationary random process by quadratic transformation,this method is not only effective in suppressing the influence of obserbation noise,but also reduces the computation and improves the frequency estimation of sweep interference.Fourth,aiming at the parameter estimation of bistatic EES-MIMO radar under the background of gaussian white noise,on the one hand,four algorithms are proposed for the quaternion model of bistatic EES-MIMO radar based on uniform linear array:The parameter estimation algorithm of the bistatic EES-MIMO radarbased on quaternion reduced dimension MUSIC(Q-RMUSIC).On this basis,three new quaternion models are derived through mathematical conversion: quaternion derived matrix,quaternion self-conjugate matrix,quaternion self-conjugate derived matrix.Based on different quaternion conversion models and different data decomposition algorithms,three new angle estimation algorithms are proposed,that is,angle estimation algorithm of bistatic EES-MIMO radar based on QD-RMUSIC、QC-RMUSIC and QCD-RMUSIC.All of these algorithms can realize the angle estimation and parameter pairing,and the dimensionality reduction can effectively reduce the computational complexity.The proposed method does not need 2-D spectral peak search and parameter pairing.On the other hand,quaternion DOA matrix method is proposed based on bistatic linear array,the algorithm overcomes the limitation of large amount of computation of spectrum peak search of MUSIC algorithm and no need pairing.Fifth,three new approachs based on cumulant-quaternion for target parameter estimation are presented in order to reduce the influence of Gaussian color noise:cumulant-quaternion MUSIC algorithm,cumulant-quaternion derived matrix MUSIC algorithm and Root MUSIC based on cumulant-quaternion algorithm.The data modela of bistatic EES-MIMO radar based on cumulant-quaternion is constructed as the fourth order cumulant is insensitive to color noise,which induces the influence of color noise,t hen the model is processed by the cumulant slice,which overcomes the shortcoming of the large amount of cumulant computation.Compared with the traditional algorithm under the background of gaussian white noise,the proposed algorithm need less computational complexity and no parameter pairing and extends the application of quaternion theory in the field of target parameter estimation of bistatic EES-MIMO radar based on gaussian color noise.Sixth,two new approachs based on covariance difference for target parameter estimation are presented in order to reduce the influence of unknown noise.The first method is reduced-dimension music algorithm based on covariance difference.The main thought of this method follows as below : the unknown noise of unknowncovariance form is "eliminated" by using of the methods of difference formulate algorithm,which reduces the influence of background noise on target parameter estimation and overcomes the shortcoming of the traditional parameter estimation algorithm when the background is unknown noise.A spatial differential smoothing MUSIC algorithm based on coherent signal source for target parameter estimation is presented in order to reduce the influence of unknown noise,the main thought of this method follows as below:a spatial differential smoothing MUSIC algorithm based on coherent signal source for target parameter estimation is presented in order to reduce the influence of unknown noise,the main thought of this method follows as below:The original quaternion covariance matrix is processed by forward and backward smoothing,the spatial difference smoothing matrix is obtained by conjugate transformation,then the unknown noise is eliminated and the rank is restored effectively,the problem of coherent source is solved,and there is no need to estimate the original covariance matrix many times.The computational complexity of the feature decomposition process is reduced effectively.At the same time,the dimensionality reduction MUSIC algorithm is used to estimate the angle with high accuracy and automatic pairing. |
PDF Full Text Request