Research On Clutter And Interference Suppression Based On Joint Multi-dimension In HFSWR

High Frequency Surface Wave Radar(HFSWR) can detect targets at distances beyond the line of sight by exploiting the surface wave mode of vertical polarization electromagnetic wave propagating over the sea water. It can also detect the steal th designed target due to its working band on 3~30MHz. It has drawn much attention for its notable features of large scale, long distance and all-day worked adaptability. However, due to the crowded working band and quite complex clutter and interference background, how to suppress the clutter and interference in HFSWR keeps one of hot topics all around the world.In this thesis, we focus on the characteristics of the clutter and interference in HFSWR for further understanding them and propose two methods f or range correlation analysis: sub-space based method and Compressed Sensing(CS) based method. Sub-space based method is used for analyzing the range correlation by the obtained clutter and interference subspace and it’s effective for the clutter and interferences which have highly statistical property. CS-based method can separate the clutter components in single snapshot and the separated clutter components are used for the range correlation analysis. In this thesis, the range correlations of sea clutter, ionospheric clutter and meteor trail are analyzed utilizing CS-based method based on measured data and the range correlations under different coherent integrate times are also analyzed.Sea clutter and ionospheric clutter can be treated as space spread c lutter due to the space distribution analysis in this thesis. And a novel method that the clutter in the main-lobe can be cancelled by the clutter in the side-lobe is proposed due to the analysis of space correlation. Firstly, this novel method is verified by the sea clutter in theory and further proved by the simulations of four typical space distributions. However, this method demands strict amplitude and phase consistency and is not robust for the situation that the direction of target is not just at the beam pointing of radar system. In this case, a knowledge-based space spread clutter canceller(KB-SSCC) is proposed based on the equivalence of space and frequency. According to the principles of target detection and maximum output Signal to Clutter Ratio(SCR), two principles for space blocking filter designation are proposed with the key thought of designing the space blocking filter adaptively based on maximum energy of the pre-detected target and the parameters of the radar system. So the robustness of the method is improved.KB-SSCC is not widely suitable of ionospheric clutter which is much more complex distributed. So a joint multi-dimension processing method based on space-time adaptive processing is proposed according to the analysis of ionospheric clutter distribution in angle-doppler frequency domain. Traditional space-time adaptive processing method can’t be used in HFSWR due to the heavily computing load and weak adaptability for complex clutter background. According to the comparisons of several partial space-time adaptive processing method and the freedom analysis in HFSWR, a principle for the size of local processing region determination is proposed. Furtherly a set of joint multi-dimension processing method is proposed based on the range correlation analysis in Chapter 2 which contains the equal weight diagonal loading method for homogeneous clutter and interference, selecting threshold method for non-homogeneous clutter with abundant secondary data, variable weight loading method for limited secondary data. This set of method is effective which is verified utilizing the measured and robust for array error and the bias of target direction.In this thesis, we not only propose a set the methods for clutter characteristics analysis and clutter suppression, but also propose the framework of joint multi-dimension processing method for implementation. The computing load is also analyzed. Due to the heavily computing load, the common Digital Signal Processor(DSP) platform can’t meet the requirements for computing ability. In this case, a novel platform based on Graphic Processing Unit(GPU) is proposed in this thesis. By processing the measured data, the computing ability and performance of target detecting and tracking have been shown to verify the superiority of joint multi-dimension processing method on GPU based platform. And the pr oposed framework is quite practical in HFSWR.