Interference Suppression And Target Detection Of Multichannel Radar

Radar is a surveillance instrument that can operate all day all weather,so it plays a vital role in terrestrial safeguard.All functions of a radar surveillance system reckon on valid detection of targets and target detection is a critical issue in the radar theory.In practice,detection algorithms should be designed based on both target characteristics and background interference.According to the operation bandwidth,radar targets are often categorized into point targets and distributed targets,whereas background interference includes sea clutter,ground clutter,jamming and so on.It is interesting to design interference suppression algorithms and target detection algorithms according to real radar operation circumstances.This papers aims at designing radar clutter suppression and target detection algorithms in theory at the background of some real applications.First,for colocated Multiple-Input Multiple-Output(MIMO)radar with transmit array widely separated from receive array,a waveform decorrelation algorithm is concerned to alleviate mutual interference between target returns from different spatial directions.Second,in order to suppress wide-band and narrow-band interference,a suppression algorithm based on fractal Fourier transform is proposed and verified through real data.Third,in wideband radar target detection,super resolution of clutters would make clutter returns have statistical distributed deviated from the Gaussian model,so a distributed target detection algorithm based on a compound Gaussian clutter model is proposed with the Rao and Wald test.Last,a bunch of real data from airborne platform is found to have many clutter residues after the Space-Time Adaptive Processing(STAP),so a moving target detection algorithm based on the Robust Principal Component Analysis(RPCA)is studied,which is also robust to system errors.These algorithms spring from real applications and solve in theory.Colocated MIMO radar may still have mutual interference after range compression and poor orthogonality of waveforms in range compression may deteriorate the interference caused by poor condition number.We present a decorrelation method based on the QR decomposition,which can alleviate interference increasing caused by unstable numerical calculation.Through a matrix decomposition and iteration operation on transmit waveform covariance matrix,decorrelation is achieved.We verify the applicability and effectiveness through analyzing direction estimation accuracy of multiple targets for the bistatic radar.For two kinds of typical active interferences,i.e.,Narrowband Interference(NBI)and Wideband Interference(WBI),they will overlap with target returns in both the time dimension and the frequency dimension.Given a bunch of real data contaminated by two kinds of interferences,we analyze the time-frequent characteristic of NBI and WBI and thereby present an interference suppression algorithm based on the fractal Fourier algorithm.It first transforms received data into the time-frequency domain used in the fractal transform and then separate target returns and interference in this domain,followed by optimization of the gain coefficient.In this manner,with interferences effectively suppressed,the time-frequency signature can be maintained to a large extent.At last,the inverse short-time fractal Fourier transform is implemented to obtain useful signals after interference suppression.Both simulative signals and real data are used to verify the effectiveness and the applicability.In multi-channel high resolution radar,target detection often confronts wideband clutter interference that are inaccurate to be modeled by common statistical models;meanwhile,samples with identical distributions to be used for interference covariance matrix estimation is often insufficient to an accurate estimation.Given unknown interference covariance matrix,the Rao and Wald test is formulated in the case of distributed targets based on a compound Gaussian distribution.Meanwhile,in case of interferences with covariance matrix having a special inverse symmetric structure,the estimation accuracy is improved especially with insufficient samples.Simulation results indicate that the algorithm can improve the target detection performance with wideband interference,especially with limited samples.In processing real data from an airborne platform,it is found that conventional Space-Time Adaptive Processing(STAP)suffers from intensive clutter remains.A moving target detection technique is presented based on the RPCA method,which presumes sparsity of targets in scope and then imposes a constraint on the shape of the output.Numerical results indicate that this method effectively clears clutter remains and improves clarity of output figures.Moreover,simulative data are generated to indicate that this method is also robust to system mismatches.