A Linear Model for Classification of Electronic Devices using Harmonic Radar

Hayvaci H. T., Shahi M., İLBEĞİ H., Yetik I. S.

IEEE Transactions on Aerospace and Electronic Systems, cilt.57, ss.3614-3622, 2021 (SCI-Expanded, Scopus) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 57
  • Basım Tarihi: 2021
  • Doi Numarası: 10.1109/taes.2021.3079572
  • Dergi Adı: IEEE Transactions on Aerospace and Electronic Systems
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-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
  • Sayfa Sayıları: ss.3614-3622
  • Anahtar Kelimeler: Harmonic analysis, Radar, Power system harmonics, Integrated circuit modeling, Radar detection, Electronic circuits, Radar tracking, Classification, estimator, harmonic radar, k-nearest neighbors, linear model, power series model, vector of parameters, TRANSPONDER, TRACKING, DESIGN
  • Ankara Üniversitesi Adresli: Evet

Özet

A new approach to exploit nonlinear reradiation of electronic circuits under test (ECUT) for classification of electronic devices using harmonic radar is proposed in this article. Unlike prior approaches, we develop a linear model to relate the measurements to the unknown parameters representing the nonlinear characteristics of the ECUT. Each nonlinear circuit under test has a distinguishable response to a power-varying single-tone or two-tone incident wave. As a result, each ECUT possesses a unique unknown deterministic vector of parameters derived from the proposed linear model. We use maximum likelihood estimator in the presence of complex white Gaussian noise based on the newly constructed linear model. The statistical features of the normalized estimated vectors of parameters are employed as distinguishing factors in classification of different nonlinear electronic devices using K-nearest neighbors classification method. Simulation results show that the proposed method of linear model with power-swept signals and estimated vectors of parameters successfully classifies nonlinear devices.