The Study, Based On The Statistical Properties Of Wideband Radar Signal Detection Method

In the field of fighting against the Four Threats to radar (electronic warfare, low-altitude/low altitude penetration, anti-radiation missiles and stealth targets) and recognizing radar target, wideband radar has very significant advantages and its applications are more extensive. However, with the radar bandwidth increases, the range resolution cell is also getting smaller and smaller, and the target becomes distributed target. At this point, the traditongnal theory of detecting signal based on narrow-band radar is not becoming to wideband radar, so the detection theory of wideband radar is need to do in-depth study. There are so many ways to detect the distributed target in the broadband system, but this paper is mainly on the research of detecting methods in single-pulse based on the differences in statistical propertier. The principal tasks are as follows:1. Analysed the target model of scattering point and echo characteristics in the broadband and high-resolution system, which focused on analyzing the statistical properties of magnitude and modeling methods for the one-dimensional range profile of a distributed target. Described the range model of clutter in broadband radar, including the latest generalized complex distribution model (GC).2. Presented a method which simulates the real terrain using SAR images to simulate radar clutter and introduced the implementation of processes and structures of the software for simulating clutter. Then, useing this simulationg software generated some kinds of clutters in a variety of typical terrain and did experiments using a variety of experience in distribution model to fit to them, so the results proved the correctness of the clutter generated by this software. Next, did the fitting experiments with the distribution of GC, so it proved that the distribution of GC is more superiority than the traditional empirical distribution. Finally, by fitting simulation data, which indicates a target, will be to change the shape of the distribution of data in inspection unit, namely, the size of the distribution tail.3. Using the differences between the statistical properties of the present unit and the distribution of background clutter, we studied the method of detecting target under the high-resolution distributed using goodness of fit test. Analysed the detection performance under various types of clutter through the simulation, and then compared with the effects of traditional ways. Experimental results show that: when it's acute that the distribution of clutter in the background, GoF detection performance has a very big advantage compared to the traditional methods. And the more acute of the distribution of the background, the greater advantages of GoF.4. Studied the method of the weak signal detection based on independent component analysis, completed the separation of the radar echoes and clutter into the separation using FastICA algorithm, and obtained some preliminary results. The results showed that: After the separation to be seized by the FastICA, the SCR of present unit slightly increased, but ineffective. When the initial SCR is low (<15dB), the increase value of SCR is usually much smaller than 1dB after the separation.