FPGA-Based Wigner-Hough Transform System for Detection and Parameter Extraction of LPI Radar LFMCW Signals


Guner K. K. , Gulum T. O. , ERKMEN B.

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, vol.70, 2021 (Journal Indexed in SCI) identifier identifier

  • Publication Type: Article / Article
  • Volume: 70
  • Publication Date: 2021
  • Doi Number: 10.1109/tim.2021.3060584
  • Journal Name: IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT
  • Journal Indexes: Science Citation Index Expanded, Scopus, Academic Search Premier, Aerospace Database, Applied Science & Technology Source, Business Source Elite, Business Source Premier, Communication Abstracts, Compendex, Computer & Applied Sciences, INSPEC, Metadex, Civil Engineering Abstracts
  • Keywords: Field-programmable gate array (FPGA), linear frequency-modulated continuous waveform (LFMCW), low probability of intercept (LPI) radar, parameter extraction, Wigner-Hough transform (WHT), TIME-FREQUENCY-DISTRIBUTIONS, DISCRETE-TIME

Abstract

Low probability of intercept (LPI) radars using linear frequency-modulated continuous waveforms (LFMCWs) is rapidly proliferating in the military domain and posing a serious challenge to conventional electronic support receivers (ESRs). In this article, we provide a practical approach that can be used in digital ESR systems for real-time LFMCW radar signal detection and parameter extraction under challenging SNR levels. We proposed a modified version of the Wigner-Ville distribution (WVD), namely, short-time Wigner distribution (STWD), postprocessing steps, e.g., column-based thresholding of STWD combined with the Hough transform (HT), and a field-programmable gate array (FPGA) design as an embedded solution with low computational complexity. The performance of the proposed system is tested with MATLAB and hardware simulations using a signal database generated with different signal-to-noise ratio (SNR) levels. The system is then embedded into a Virtex-7 FPGA where the performance evaluations showed satisfactory detection results down to -6-dB SNR level, and parameter extraction results down to -8-dB SNR level.