Review of radio frequency spectrum analyzers and proposals for dynamic implementation of energy detector
| Autor: | M.V. Buhaiov, O.A. Nahorniuk, R.V. Shapar |
|---|---|
| Jazyk: | English<br />Ukrainian |
| Rok vydání: | 2021 |
| Předmět: | |
| Zdroj: | Технічна інженерія, Vol 2, Iss 88, Pp 95-102 (2021) |
| Druh dokumentu: | article |
| ISSN: | 2706-5847 2707-9619 |
| DOI: | 10.26642/ten-2021-2(88)-95-102 |
| Popis: | Nowadays the main approach of organizing the radio frequency spectrum monitoring is fully automatic continuous analysis of a given frequency range. The overwhelming majority of software products of the world leaders specializing in the development of radio monitoring instruments have an open architecture, which allows users to implement their own signal processing modules. Therefore, improving the software and algorithms of existing spectrum analyzers will increase the efficiency of detecting occupied frequency regions and forming receiving channels. In this paper, it is proposed to use a dynamic energy detector in the frequency domain to detect frequency channels. The dynamism of the detector consists in the sequential calculating of the detection threshold depending on the value of the test statistics. The coefficient of variation of frequency samples is used as a test statistic. The essence of the proposed method is calculating the coefficient of variation using the samples of the power spectral density of the received signal and comparing it with the threshold for test statistic, calculating threshold in frequency domain for dividing samples into signal and noise. Then the procedure is repeated until all signal samples are detected. For operation of the algorithm, it is necessary to set only the parameters of the Welch periodogram, the vector of the threshold values for the test statistics, and the probability of a false alarm. The software implementation of the proposed approach will allow detecting and determining the frequency boundaries of signals in a wide dynamic range, with unknown values of the bandwidth occupancy and noise power. The increase in the speed of the developed algorithm in comparison with similar ones depends on the level of occupancy of the radio frequency spectrum, the dynamic range of signals and the length of the fast Fourier transform and can be tens of times. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|