Research On Ionospheric Clutter Detection And Identification Method Based On Multi-dimensional Spectral Characteristics

High frequency surface wave radar (HFSWR) can detect the targets over the sea orat low altitude, for it emits vertical polarized wave which can diffract along the seasurface. And compared to traditional microwave radar, HFSWR has its extraordinarysuperiority, such as the long-distance and large-scale detection, strong anti-stealth andanti-Low Altitude Assault capability, which make HFSWR be widely used in bothnational defense and civilian areas. But HFSWR will be interfered by a variety of noiseand clutter at work. As a major interference in Range-Doppler spectrum, ionosphericclutter has seriously affected the performance of radar’s detection. What is worse,ionospheric clutter is difficult to be suppressed for its complex time-varying fluctuationsand similarity to targets. Therefore, it is the key technical problem that must be solvedto extract ionospheric clutter efficiently from the Range-Doppler spectrum. For thisreason, the research focuses on the detection and identification method of ionosphericclutter based on multi-dimensional spectral characteristics by using the measured dataof radar. The dissertation explores the spectral and statistical characteristics ofionospheric clutter based on the extracted results in order to provide the necessaryknowledge for understanding the mechanism of ionospheric clutter, and it statisticallyanalyzes the background of detection and eliminates the false alarms of radar. The maincontents are as follows.1. The mechanism of ionospheric clutter and the characteristics of each componentin the Range-Doppler spectrum are introduced. The generation process, the transmissioncharacteristics and the extended features in the Range-Doppler spectrum are mainlyanalyzed. The generation mechanism and characteristics of other ingredients areintroduced, and the image features of them in the Range-Doppler spectrum aresummarized.2. The detection and identification method of ionospheric clutter are mainlyresearched. The first part introduces a detection method to ionospheric clutter. Firstly, itjudges the possible existence of ionospheric clutter through the threshold method. Thenaccording to the massive and planar features of ionospheric clutter, it determines accurately the presence or absence of ionospheric clutter through thetemplate-convolution algorithm. And it analyzes the effectiveness of the algorithm byusing the measured data. The second part proposes a segmentation method toRange-Doppler spectrum based on Gabor wavelet filter. Firstly it gets the GaborCoefficient Map in Gabor space through the two-dimensional Gabor wavelet transformto the Range-Doppler spectrum. Then it separates the Gabor Coefficient Map by usingthe adaptive threshold method and gets the final segmentation results by mapping thefirst segmentation result to the original Range-Doppler spectrum. In addition,comparisons with some classical segmentation methods are presented. The third partpresents a region identification method to ionospheric clutter based on the results ofsegmentation results. Firstly, it identifies and eliminates the first-order sea clutter byusing the template-convolution algorithm combined with the theoretical position ofBragg peak. Then it eliminates granular noise and small block noise by using thetemplate-convolution and connected-component-labeling algorithm. Next it uses therange and Doppler features of ionospheric clutter to determine the distance ofionospheric clutter. Ultimately it gets the identification results of ionospheric clutter.3. The identification effects measure and evaluation methods of ionospheric clutteris studied. Firstly, by using statistical distribution function fitting and parameterestimation method, it measures the identification effects of ionospheric clutterqualitatively and quantitatively. Finally, it defines several parameters that can describethe degree of interference to evaluate ionospheric clutter.