Clutter Suppression Method In HF Hybrid Sky-surface Wave Radar

In the past decade,high-frequency hybrid sky-surface wave radar(HFSSWR)is a new development direction in the HF radar research areas.It takes the ionosphere as the transmission medium.The forward reflected electromagnetic waves are irradiated to the interested target area and then diffracted to the receiver through the sea surface.This kind of radar system theoretically overcome the problem of the close distance blind zone and the ionospheric modulation effect in HF skywave radars,and the common vertical ionospheric clutter problems in HF surface wave radar.The special working principle overcomes the above shortcomings but also introduces new problems.It is found that the frequency of the first order Bragg peak is coupled with the bistatic angle,the grazing angle,and the ionospheric modulation.This leads to the broadening of the Bragg peaks in one resolution unit.Compared with the HF skywave radar,the electromagnetic waves in HFSSWR are subjected to only one ionospheric reflection.But the ionospheric modulation effect cannot be ignored.It is still the main factor of the echo spectrum broadening.In addition,when the multimode conditions occur,the multiple sets of clutter superimposed on each other,it is greatly exacerbated the difficulty of processing.In this paper,we start from the analysis of clutter characteristics based on the measured data,and then study the clutter suppression algorithm and target detection technology in a complex environment.The characteristics of clutter in distance domain,temporal domain,spatial domain and eigen domain are explored.Two clutter suppression algorithms and a detection algorithm are designed based on these characteristic analysis results.In this paper,we firstly introduce the research background of the HFSSWR and then explain the purpose and significance of the research.Then we summarize the research status quo,broadening characteristics and model research status of clutter and clutter suppression method.At the end of this chapter,we explain the research ideas and the structure of this thesis.In the second chapter,the working principle and system structure of the HFSSWR are introduced firstly.The working principle of the radio wave environment diagnosis and the ionospheric monitoring sounder is introduced emphatically.Then,we introduce the linear frequency modulation continuous wave and the traditional processing flow adopted by the radar.The three-dimensional data block of radar signal processing is shown.The first-order Bragg peak frequency formula is deduced.The influences of the bistatic angle,the pitch angle and the ionospheric phase path disturbance on the spectral broadening are analyzed.For the first time,we systematically analyzed the characteristics of sea clutter in the measured data of HFSSWR.The distance domain characteristics,spatial domain characteristics and eigen domain characteristics are studied.We obtain the conclusion that the sea clutter is non-homogeneity in the distance domain,the sea clutter has poor directionality and has obvious spatial-slow time coupling.The stationarity of the echoes are analyzed using the surrogate data method and the conclusion is obtained that target echo is stationary and clutter data is stationary in segment.Finally,according to the results of the characteristic analysis,the difficulties of clutter suppression are summarized.In the third chapter,we study the adaptive spread clutter localized processing(SCLP)algorithm for spreading clutter suppression under single dataset.As a spatial-frequency cascading algorithm,it overcomes the disadvantages of single spatial processing that cannot effectively suppress the clutter with poor directivity.The non-stationarity of clutter in the range dimension leads to that only the current range cell data is available when estimating the statistical properties of clutter.We call this case a single-dataset.According to the poor spatial characteristic of clutters and good spatial characteristic of targets,the spatial orthogonal projection matrix is used to construct the clutter sample from only one range bin.In order to reduce the need for samples for the adaptive processor,we design a frequency domain localized processing algorithm.The transformation matrix is used to transform the time domain data to the region of interest in the frequency domain,and then the clutter suppression is performed using a minimum variance distortionless response filter.The problem of interval size in the frequency domain is studied.We obtain the conclusion that the processing interval should be smaller than Doppler resolution.Finally,the validity of the algorithm is verified by simulation and real data processing.In the fourth chapter,we propose a spatial-temporal joint processing algorithm based on the subspace signal model.The algorithm realizes sea clutter suppression and better spectral peak holding ability for spread target signal simultaneously.The algorithm is also a single-dataset algorithm and utilizes another conclusion that the clutter has a spatial-slow time coupling property.Firstly,the basic principles of amplitude and phase estimation(APES)algorithm and STAP algorithm are introduced.Considering that the target may also be broadened by ionospheric contamination,a subspace signal model is introduced to characterize the potential target signal,which is more robust.We design an eigenvector basis and use the eigenvector corresponding to the large eigenvalue of the average covariance matrix of the potential frequency range signal as the base vector.When deducing the APES-STAP algorithm,we firstly introduce the signal model and method of constructing the space-time data in the case of single-dataset.The realization of the subspace signal APES-STAP algorithm is shown to be a two-step cascade algorithm: the first stage is a time domain projection process and constructs the sample data;the second stage is a classical STAP algorithm that computes the adaptive weighting.Finally,the effectiveness of the algorithm is verified by the simulation data and the measured data.In the fifth chapter,we studies the target detection algorithm based on segmented AR model generalized likelihood ratio test.The algorithm provides a new idea for the problem of target detection in a non-stationary environment.Traditional target detection process is generally a two steps process: the first step is anti-jamming which improves the signal to noise ratio,and then cascade a constant false alarm detector.The generalized likelihood ratio detection theory redefines the target detection problem by using the hypothesis test theory,combines the anti-jamming and constant false alarm detection into one step.However,the problem of estimating the covariance matrix is still inevitable.To estimate a large covariance matrix under single-dataset,the clutter is modeled as a time-varying autoregressive model.The clutter is segmented according to the nonstationarity characteristics obtained in the second chapter.This reduces the difficult to estimating the time-varying AR model coefficients.Afterward,the relationship between the model coefficient and the covariance matrix is used to estimate the large covariance matrix The test statistic of the detector is deduced.Finally,the validity of the algorithm is verified by simulation and measured data.Finally,we summarize the whole paper,expounds the innovation of this paper,and points out the further possible research direction.The research results of this paper will improve the anti-interference ability of the HFSSWR,improve the detection probability of the target in the complex environment,and lay a solid foundation for the final engineering practice of the radar.The single-dataset clutter suppression algorithm has a general significance and can be extended to other radar applications that face the same complex background.