350 rub
Journal Radioengineering №3 for 2014 г.
Article in number:
Digital interpolation algorithms of estimation the frequency of the harmonic signal, comparative analysis
Authors:
S.A. Antipov - Dr.Sc. (Phys-Math.), Professor of chair of the general physics of a radio-electronic profile of the Voronezh state technical university
D.S. Gnezdilovs - Post-graduate department radio engineering. E-mail: GnezdilovDS@gmail.com
V.A. Kozmin - Ph.D. (Eng.), associate professor, Director of research JSC «IRKOS» (Moscow). E-mail: kv@ircoc.vrn.ru
V.M. Stopkin - engineer-software, JSC "IRKOS" (Moscow). E-mail: vmstop@yandex.ru
Abstract:
A Narrowband signal with different types of modulation has been widely used in the transmission of information. Digital processing methods the frequency measurement of harmonic, AM, ASK and PSK signals is one of the typical radio monitoring tasks. The digital measurement algorithms are based on a fast Fourier transformation (FFT) are widely used in modern automated complexes. This article focuses on the comparative characteristics of some of the most common algorithms measurement frequency harmonic signal with the algorithm used in the software by JSC "IRKOS". In the first part of the article discusses algorithms parabolic interpolation (PI), Gaussian interpolation (GI) interpolation Ligges (ILG) and iterative algorithm Reisenfeld (IRZ). To evaluate the potential accuracy of the algorithms and a comparison between them shows plots the mean error measurements within the sampling interval frequency. The second part contains a two-step algorithm developed (TS). In the first stage evaluation of the relative frequency, using the non-rectangular window, on the second - to refine the estimates. To evaluate the efficiency of algorithms carried out an experiment in which a comparison was the mean square error of frequency measurement with of the Cramer-Rao lower bound (CRLB). The experiment showed that the most accurate of the considered algorithms are algorithms IRZ and TS. IRZ algorithm requires more computational resources and loses linearity of about 40 dB signal to noise ratio. Two-step algorithm does not lose linearity considered range from minus 20 to 100 dB signal-noise ratio and has a rating as close to NGKR.
Pages: 42-46
References

  1. Rembovskij A.M., Ashixmin A.V., Koz'min V.A. Radiomonitoring. Zadachi, metody', sredstva // M.: IP Goryachaya liniya ? Telekom. 2012. 217 s.
  2. Marpl.-ml. S.L. Cifrovoj spektral'ny'j analiz i ego prilozheniya. M.: Mir. 1990. 65 s.
  3. Kayukov I.V., Manelis V.B. Sravnitel'ny'j analiz razlichny'x metodov oczenki chastoty' signala // Izv. vuzov. Ser. Radioe'lektronika. 2006. T. 49. № 7. S. 42-56.
  4. Gnezdilov D.S., Matveev B.V. Sravnitel'ny'j analiz czifrovy'x interpolyaczionny'x algoritmov oczenki chastoty' radiosignala // Vestnik VGTU. 2013. T. 9. № 2. S. 37-39.
  5. Reisenfeld S., Aboutanios E. A new algorithm for the estimation of the frequency of a complex exponential in additive Gaussian noise // IEEE Communications letters. 2003. V. 7. P. 549-551.
  6. Gasior M., Gonzalez J.L. Improving FFT frequency measurement resolution by parabolic and Gaussian spectrum interpolation // Geneva 23. Switzerland 2004. CERN. CH-1211. P. 40 - 43.
  7. Beschl B., Ligges U., Weiss C. Frequency estimation by DFT interpolation: a comparison of methods // Signal Processing Magazine. May 2009. 475 p.
  8. Metodika izmerenij NALS. 80064-01 92 02 Izmerenie parametrov izluchenij radioe'lektronny'x sredstv priemnikami izmeritel'ny'mi panoramny'mi «Argamak-M», «Argamak-IM», «Argamak-IS» // M.: ANO «Institut metrologii oborony' i bezopasnosti». 2013. 14 s.
  9. Prikaz №1131 ot 19.12.2011 ob organizaczii raboty' po zakupkam radiokontrol'nogo oborudovaniya i apparatno-programmny'x kompleksov predpriyatiyami radiochastotnoj sluzhby'.
  10. GOST R 53373-2009. Oborudovanie stanczij radiokontrolya priemnoe avtomatizirovannoe. Texnicheskie trebovaniya i metody' ispy'tanij.
  11. Rekomendacziya MSE'-R SM. 377-4. Tochnost' izmereniya chastoty' na stancziyax, ispol'zuemy'x dlya mezhdunarodnogo radiokontrolya.