Characteristic Analysis And Suppression Of The First Order Sea Clutter In HF Hybrid Sky-surface Wave Radar

In this paper, studies are carried out to focus on the characteristic analysis andsuppression of the first order sea clutter through clutter modelling and signalprocessing in high frequency (HF) hybrid sky-surface wave radar. Specifically, theattention has been paid to explicate the clutter feature of broadening, statistics andradar cross section (RCS) as well as the method based on which to suppress the firstorder sea clutter existed in such novel propagation mode.In terms of the issue of first order sea clutter broadening, the spatial geometrydistribution of which is firstly studied when the echo time delay is fixed. Eithertheoretical derivation or simulation could prove that the distribution can bedescribed with an ellipse equation. Furthermore, an expression has been yieldedsimultaneously to calculate the Bragg scattering frequency. Additionally, thebroadening extent is also calculated quantitatively on account of the bistatic anglevariance within a clutter resolution cell, which is disproportional to the receivingbeam bearing. Meanwhile, a research has also been conducted to study thebroadening effect on the first order sea clutter triggered by the nonlinear phase pathdistortions and multi-path propagation based on the ionosphere rough surface model.The results illustrate that the major factor attributing to the broadening is related tothe coherent integration time (CIT)， the transmitting beam width and theionospheric relevant distance etc.Afterwards, as for the problem of statistical test on the first order sea clutteramplitude, a method based on the high order cumulant is put forward to turn aGaussian test into a dualism hypothesis problem. Specifically, data segment on thesea clutter echo is firstly implemented to yield a piece of approximately stationarydata. Based on which, an AR model is then constructed so that the third and fourthorder cumulant covariance matrix can be calculated to yield the calculation result ofstatistical test variable. According to the test result, the amplitude of the first ordersea clutter submits to the Gaussian distribution for most range bin cells with theconfidence prabality of95%.As far as the radar cross section of the first order sea clutter is concerned, anexpression to calculate the ionospheric dispersive coefficient is yielded on conditionthat the ionosphere is spherical stratified and the electron density of ionospheresubmits to the quasi-parabolic model. Meanwhile, simulation has also beenconducted to illustrate the influence on the envelope of HF narrow bandtransmitting signal. Furthermore, the effect on the range resolution due to theionospheric dispersion is studied when considering the output of matching filter so that the impact on the area of clutter resolution cell in reality can be obtained.Eventually, a method to simulate the first order sea clutter cross section is generatedafter taking the limited clutter resolution cell as the standard to divide the clutterresolution cell in reality in conjunction with the existent model derived by Gillapplied in the bistatic HF surface wave radar. The simulation results explicate thestraightforward radar cross section variance when taking the effect on the clutterresolution cell triggered by ionosphere into consideration.For suppressing the broaden first order sea clutter, a cascaded processingscheme based on both empirical mode decomposition (EMD) and clutter cancellingis derived after conducting a deep analysis on the construction of the scattered echosignals. Simulation results are present to verify the rationality and validity of theproposed method to separate the main broaden clutter so as to provide a refer ence toextract the ionospheric phase contaminated function to compensate the distortionsimposed by the ionosphere. However, allowing for the fact that the sharpen firstorder sea clutter is still dominant, therefore an improved method based on forwardand backward linear prediction (FBLP) algorithm and clutter multi-dimensionalfeature is proposed to cancel the dominant first order sea clutter. According to thecancelling result, the first order sea clutter has been suppressed to a large extent andthe target could be also detected simultaneously.