Research On Clutter Suppression Method In Mixed Propagation HF Over-the-horizon Radar

The mixed propagation High-frequency(HF)over-the-horizon radar(OTHR)combines the respective advantages of the traditional sky-wave and surface-wave HF radar.The new working mode can overcome the inherent shortcomings of conventional HF radar.It is beneficial to further improve the performance of the HF OTHR and expand its application scenarios.In recent years,the mixed propagation mode has become a hot research direction in the field of HF OTHR.The new radar system also inevitably brings new problems.Among them,the suppression of non-stationary clutter in the background of mixed propagation mode is an important factor restricting the performance of the new radar to target detection.The sea clutter signal in the mixed propagation mode will be affected by the phase modulation of the ionospheric transmission channel,the glancing angle and the bistatic angle caused by the system layout,and the non-stationary background environment.These factors will broaden the sea clutter spectrum.The clutter characteristics fluctuate greatly,and it is easy to submerge low-velocity vessel targets,which seriously affects the detection performance of the radar system.Therefore,the study of the clutter suppression in the mixed propagation mode has important impetus and significance for accelerating the practical application of the mixed propagation HF OTHR.To this end,this dissertation aims at the core problem of non-stationary clutter suppression algorithm design in mixed propagation HF radar.Based on the measured data,the geometric model of the new system,multi-mode targets matching,sea clutter broadening mechanism and multi-dimensional characteristic analysis have been carried out in-depth research.Using the relevant statistical characteristics of clutter in range domain,spatial domain and space-time two-dimension domain,the corresponding clutter suppression algorithms are proposed which provide feasible solutions to the problem of non-stationary clutter in practical HF radar system.This dissertation first analyzes the geometric model and clutter characteristics of the mixed propagation mode,as the basis of the subsequent clutter suppression algorithm design.The geometric model of the mixed propagation HF radar system is given in theory,and its feasibility and propagation attenuation of different modes are then analyzed.Aiming at the problem of false track generated by multi-propagation mode echoes in the mixed propagation mode of surface wave,a multi-mode targets matching method is proposed to eliminate false targets and tracks.Meanwhile,the verification of this phenomenon also illustrates the rationality of the surface wave based mixed propagation mode.The mechanism and theoretical Doppler frequency of broadening sea clutter in mixed propagation mode are analyzed and deduced.Using the measured data of the mixed propagation mode,the range,spatial and space-time characteristics of the sea clutter are analyzed,and the corresponding conclusions are given.In the mixed propagation mode,sea clutter exhibits two characteristics with different degrees of stability in range domain.Under the low impact of non-stationary environment,the sea clutter shows the characteristic of discrete partially stationary;with the high impact of highly heterogeneous environment,sea clutter exhibits extremely non-stationary characteristics in range domain,which is called single data set case.In spatial domain,the measured sea clutter does not have obvious directivity.Meanwhile,sea clutter shows significant space-time coupling characteristic with some concentrated large eigenvalues.The study of the different dimensions characteristics of sea clutter is the basis and guidance for the design of the clutter suppression algorithms in this dissertation.In mixed propagation HF radar,the sea clutter has no obvious directionality,the spatial correlation is strong while the target signal has an obvious directionality.Utilizing this feature,an improved sidelobe cancellation algorithm based on optimized correlation analysis strategy is proposed.Combining the characteristic that sea clutter exhibits partial stability in the range domain,a training sample selection method based on the correlation analysis of the maximum eigenvector of angle-Doppler local data is proposed,and a method for determining the size of the localized processing region is given.The effective training samples are selected to calculate the spatial adaptive weight to achieve better clutter suppression performance.Through simulation experiments and measured data,the effectiveness of the method is verified.For that the sea clutter in mixed propagation HF radar has more significant coupling characteristics in the space-time dimension,two reduced-dimension space-time adaptive processing algorithms based on the selection of training samples are proposed.In practical system applications,the requirements of the full-dimensional space-time adaptive processing on the number of training samples are difficult to meet.The reduced-dimension space-time adaptive processing algorithm is proposed to lower the system calculation burden and the number of samples.Also combined with the non-stationarity characteristics of the training samples in the range dimension,the methods of training sample selection based on geometric barycenter distance covariance matrix estimation and information geometric distance covariance matrix estimation are proposed respectively.Through the covariance matrix data of the training samples,different g eometric mean distances are calculated to obtain the covariance matrix estimation of the range cell under test on the corresponding matrix manifolds.Then selecting the training samples which are close to the clutter characteristic information of the cell under test,and calculating the space-time adaptive weight to achieve clutter suppression processing.The vadility of the proposed methods are verified by simulations and measured data.Aiming at the sea clutter characteristics of the single data set with extremely non-stationary clutter in the mixed propagation mode,combining the spatial and space-time characteristics of the sea clutter,a modified spatial adaptive processing algorithm and a reduced-rank space-time adaptive processing algorithm under single data set are proposed.On the basis of the spatial adaptive processing,we select the effective training samples in Doppler domain to estimate the clutter covariance matrix due to it is not possible to have valid training samples in range domain.Selecting the angle localized processing region around the range-Doppler cell under test,and calculating the generalized inner product between each Doppler sample and the data of interest through the non-stationary detector.Based on this,the effective Doppler domain samples are selected to estimate the covariance matrix and obtain the adaptive weight.Besides,taking the advantage of the space-time coupling characteristics of sea clutter,another single dataset reduced-rank space-time adaptive processing algorithm based on fast approximation power iteration algorithm is proposed.Using half-resolution Doppler domain projection technology,the signal component of interest is removed from the primary data of the cell under test.Combined with the iterative subspace tracking algorithm,the clutter subspace of the covariance matrix of the cell under test is accurately estimated and the adaptive weights of the orthogonal projection space to the clutter are obtained to achieve clutter suppression.Using simulation and measured data,the clutter suppression performances of the two methods under extremely non-stationary environment and single data set were verified.The research results of this dissertation will help to improve the clutter suppression performance and anti-jamming ability of the mixed propagation HF radar,as well as the radar’s ability to work in the background of complex and non-stationary environments.It provides practical solutions with certain universality to the application of HF radar in the non-stationary environments.