Radar Data Filtering Technology Research

Radar signal often corrupted by different kinds of noises in the processing of collection and transmission, which make it more difficult to do some subsequent signal processing, such as signal detection, threshold division, character distill, so noise detection and removal becomes part and parcel in signal processing, and has arrows widely researching interest in recent years. In this article, aiming at the descendant of filter performance due to the big infection of impulse noise and gauss noise in the local part of signal as to be difficult to separate the impulse dot from the gauss dot, some correlation research are done and use the fuzzy filter algorithm to solve the problem of fuzzy separation of impulse dot and gauss dot through the research of scale parameter geneβin the whole and local domain of the signal. Therefore, better results are expected by experimental comparation.Item background and the source of research task are introduced simplely in the beginning of the article, then the model of signal and some traditional filtering algorithm and their evaluating methods are introduced secondly. Imitation experiment and analysis of performance of the filter of radar signal using the traditional filtering algorithm are done in the next chapter and pay more attention on the not good filtering performance in the environment of the variety of local noise style using the traditional filtering algorithm. In the following chapters aiming at the worse filtering performance due to the limitation of equal-authority algorithm, introduce the dot-change-authority fuzzy filtering algorithm, and aiming at scale parameter geneβdo some research in the whole and the local of signal. Lastly put forward the improve algorithm on the FWA using the character of protecting the identity of the noise type while the scale parameter gene is at low lever and research better results compared with traditional algorithm. In conclusion, chapter 5 sums up the whole article and gives some new ideas and opinions on some questions in the paper.