Frequency Monitor And Frequency Selection Method Research Of High-Frequency Software Defined Radar

High-Frequency Software defined Radar (HFSR) is the combination of the Software Radio technology and High-Frequency Radar. It is an important research to develop High-Frequency Radar with Software Radio technology which can make the Radar a miniaturized system and improve its mobility and flexibility. Software managed Radar can also improve the maintainability of the Radar system. The High-Frequency software defined Radar project is one of the 211 National Subject Building Project.However, HFSR operates in shortwave band, where the frequency is very crowded, and electromagnetic environment is very complex, at the same time distributes a lot of interference. The purpose of this paper is, through analyzing the specific HFSR frequency monitoring system hardware platforms and the frequency monitoring system signal processing procedures, to achieve the frequency monitoring system software design and engineering realization, which implements real-time and continuous monitoring the HFSR frequency spectrum occupancy, and chooses'quiet'band which is suitable for high-frequency radar to work with, so that the HFSR can work normally in the shortwave band with terrible electromagnetic environment.Firstly, the paper introduces the HFSR frequency monitoring system's hardware component, working sequence and signal processing procedures, and finds the problems that frequency monitoring system software design should solve. Secondly, the paper analyzes the time-frequency characteristics of lightning interference in HFSR system, and then according to the impulsion interference's singular features in the time domain, researches the method of detecting and eliminating the impulsion interfere in time domain. After analysis of previous methods'shortcomings, the paper puts forward two simple and effective methods of impulsion interference detection: Detecting Impulsion Interference by Morlet Wavelet Transform and Maximum Likelihood Estimation and Impulsion Interference Detection Based on the Hilbert-Huang Transform (HHT). And the simulation results demonstrate the effectiveness of the two methods. Thirdly, the paper researches the methods of the two parts of Quiet Frequency Selection: frequency spectrum prediction and sliding window frequency selection. Frequency Spectrum Prediction Method Research: EMD method is particularly applicable to non-stationary signal analysis and Support Vector Machine (SVM) has incomparable advantages, comparing to other prediction methods .This paper proposes a new time series prediction method by integrating EMD and SVM, and the actual data simulation results show that this method is more effective than single SVM prediction method and traditional AR model prediction method. Sliding Windows Frequency Selection Method Research: Extract two selected frequency factor: sliding window average power spectral and undulate degree, then score them respectively in accordance with certain criteria, after that, synthesize two factors'scores to calculate the total score of each sliding window, ultimately, the frequency monitoring system selects the best frequencies according to the final scores.Finally, the paper accomplishes the frequency monitoring system software design, and engineering realization of'quiet'frequency selection of multi-band multi-carrier frequency radar.