Покращення роздільної здатності при спектральному аналізі радіолокаційних сигналів
| Jazyk: | angličtina |
|---|---|
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | Системи озброєння і військова техніка, Iss 2(66), Pp 137-143 (2021) |
| ISSN: | 2518-1580 1997-9568 |
| Popis: | The location of air objects in the ultra-high frequency range of radio waves is associated with a significant distortion of the shape and, consequently, the spectrum of radio signals. The problem of increasing the efficiency of spectral analysis of radar signals against the background of noise by improving the resolution is solved. The analysis of linear and nonlinear methods of deconvolution is carried out, the conditions of their application are defined. It is based on the classical operation of signal spectrum deconvolution, which is performed using a Wiener filter. Measures of suppression of parasitic low-frequency Gibbs oscillations are applied. Detection of individual signals is carried out by comparison with the adaptive threshold, which is formed by taking into account the background component for signal and noise separation, and the local component for removing interfering satellite signal peaks caused by signal processing in the receiving path of the radar station. The formation of the background component of the threshold is performed using the median filter procedure, the local component is calculated as a reaction to the echo signal spectrum of the Gaussian smoothing filter with the selected core width. A nonlinear operation of differentiation of the threshold processing result is used to increase the resolution. In order to assess the quality of detection and separation operations in the selected way, a signal-noise model of a complex manipulated broadband signal is constructed. The possibility of suppressing its satellite petals as a result of coordinated filtration by the receiving path of the radar station was investigated. The application of the chosen method requires a much smaller amount of mathematical calculations and, consequently, less time to achieve the ultimate goal compared to the classical maximum likelihood method. The presented results of simulation modelling confirm the increase of detection efficiency and separation of useful signals at joint use of linear and nonlinear deconvolution methods. The application of the considered method of radar signal processing will affect the quality of Doppler filtering, the accuracy of measuring the radial velocity of targets. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|