Research On Knowledge-Aided MIMO Radar Waveform Design

Due to traditional radar is limited to cope with complex battlefield situations and perform multi-task and,it is one of the necessitated issues to study how to effectively improve the multi-function ability and intelligent level of modern radar system.With the rapid development of military technology,the emergence of knowledge-aided radar signal processing technology and multi-channel radar concept,are of great significance to improve radar performance in complex and valatile battlefield situations.Therefore,on the basis of knowledge-aided MIMO radar system,by taking full advantage of diversify knowledge of radar working environments,this dissertation makes in-depth study at the issues about transmit beampattern waveform design for MIMO radar,the design of the sparse frequency waveform for MIMO radar with electromagnetic conflict suppression,and the design of MIMO radar waveform to improve target detection performance in clutter environments.This study strives to satisfy multi-function and intelligent requirements of modern radar system,its main contributions are as follows:1.In order to implement the function of oriented irradiation for MIMO radar,and focus the electromagnetic energy on interested spacial region.With the approximate direction of target taken as aided knowledge,to solve the problem that existing transimit beampattern design methods for narrowband MIMO radar suffer from high sidelobes,a transmit beampattern sidelobe suppression strategy based algorithm is proposed.Firstly,by using the beamspace weighting matrix signal model,the sidelobe suppression optimization problem is established.Since constant modulus is a nonconvex constraint for sidelobe suppression strategy,which leads three optimization models to be nonconvex.Then,semi-definite relaxation technique is used to convert the nonconvex optimization problems to convex ones.Signal rotational invariance property is guaranteed according to duality of the beamspace weighting matrix,and the beamspace weighting matrix is solved by using Gaussian randomization method,which leading to optimal solution to the original optimization problem.Finally,direction-of-arrival estimation of the target with ESPRIT method is conducted at the receiver,and the Cramer-Rao bound of direction-of-arrival estimation is presented.Simulation results show that all the three proposed strategies maintain the desired mainlobe shape and effectively reduce the sidelobe level of the beampattern,which improves the estimation accuracy and resolution of the target's direction-of-arrival.2.The advantage of wideband radar signal is that it possesses a large amount of information and enjoys high resolution.However,the frequency unconsistency of exsiting wideband MIMO transmit beampattern synthesis methods are poor.Therefore,a conjugate-gradient based algorithm for wideband transmit beampattern design is proposed.Firstly,the power spectral density matrix is used to form the matching criterion for wideband transmit beampattern,and an optimization model is established under the constraint that elemental power is uniform across each antenna.Then,a multi-dimensional spherical mapping method is adopted to convert the constrained optimization problem into an unconstrained one by eliminating all the constraints.Finally,a conjugate-gradient based iterative algorithm is utilized to solve the optimization model.Simulation results show that the proposed algorithm can suppress fluctuations in the mainlobe and sidelobe regions,mitigate the frequency unconsistency issue of wideband transmit beampattern effectively,resulting in a better wideband transmit beampattern synthesis performance compared with exsiting methods.3.In order to implement the function of avoiding electromagnetic conflicts for MIMO radar,and let MIMO radar do not interfere with other wireless equipment with proper functioning when their work frequency band are congested.A cyclic iterative based algorithm for designing sparse frequency orthogonal waveforms is proposed.To aware the spectrum knowledge of congested frequency band,first of all,a real-valued Root-MUSIC based power spectrum estimation algorithm for interference signal is presented,which effectively solves the problem of acquiring the spectrum knowledge.On these grounds,the desired spectrum is used as an auxiliary variable,the objective function is constructed based on both the mean square error of the spectrum of transmitting waveform with the desired one and the integration sidelobe levels of each elemental transmit waveform.The optimization model is established under the constraint that the waveform is constant envelope as well as the spectrum magnitude lies between the pre-established upper and lower bounds.Then,under the framework of cyclic iterative algorithm,fast Fourier transform and spectral decomposition techniques are used to solve the optimization model.Simulation results show that the designed waveform has better notching performance,and it can avoid the congested frequency band,with a guarantee of low autocorrelation sidelobes and cross correlation between each elemental transmit waveform.4.In order to implement the function of suppressing clutter interference,and improve the detetion performance of colocated MIMO radar in the clutter environment.Taking the direction and scattering amplitude of the target along with clutter as aided knowledge,a signal to clutter and noise ratio criterion based algorithm for MIMO radar waveform design is proposed.Firstly,the maximization of output signal to clutter and noise is utilized as the objective function.On the basis that the transmit waveform satisfies constant modulus constraint,as well as the radar waveform satisfies a similarity constraint involving a radar waveform(reference waveform)which has good pulse characteristics,a joint optimization model of finite alphabet waveform and receiving filter weight is established.Then,by utilizing the framework of alternating iterative algorithm,the optimization problem is decomposed into two sub-optimization problems,which are solved by using Lagrange multiplier method and semi-definite relaxation technique respectively.The algorithm results in both transmit waveform and receiving filter weight.Simulation results show that the proposed algorithm has high output signal-to-noise ratio,and can improve the radar dectection probability for targets with an eye on pulse compression property of the transmit waveform.5.In order to improve the detetion performance of distributed MIMO radar in the clutter environment.Taking the second-order covariance statistics of the clutter environment as aided knowledge,a Kullback-Leibler(KL)criterion based algorithm for MIMO radar waveform design is proposed.Firstly,according to the signal model in the presence of clutter and colored noise interference,the asymptotic relationship between detection probability,false alarm probability and relative entropy,which addressed by Stein's theorem in information theory field,is apdopted.With the total power constraint of radar system,the waveform optimization model is established by using KL divergence as the objective function.Then,on the basis of conducting eigenvalue decomposition of the target,clutter and noise response covariance matrix,the closed-form solution is deduced for the optimal detection optimization problem.Simulation results show that the proposed algorithm can effectively suppress the clutter interference,as well as improve the target detection performance of distributed MIMO radar under clutter and colored noise situation.At the same time,the proposed algorithm has certain robustness when there are errors with the covariance statistics knowledge.