Discrimination of Complex Radar Targets Using the Dominant Poles Determined in the Time and Frequency Domains

Abstract

In this paper, we present and compare two techniques of discriminating complex objects, using the electromagnetic scattering response in the time and frequency domains. The scattering in the time and frequency domains are computed using the Finite Difference Time Domain (FDTD) method and the Method of Moments (MoM), respectively. The dominant Natural Resonant Frequencies (NRFs) of the objects have been identified, using the Matrix Pencil of Function (MPoF) algorithm from the scattering response in the time domain, and the Vector Fitting (VF) method using the response in the frequency domain. The dominant NRFs in both cases have been identified after thresholding the power contribution of each pole. A “Risk” function using the dominant poles of any two scattering objects is then defined, which aids in quantifying the discrimination. Perfectly Electrically Conducting (PEC) objects with minor variations in their geometric shape have been discriminated using the risk function. It is shown that the techniques described in this paper are effective in discriminating targets. Both approaches, using the time and frequency domain scattering response are found to yield identical results. A single run of the FDTD algorithm is sufficient to compute the electromagnetic response required to determine the dominant natural resonances. Hence, the approach of using the time domain scattering response is computationally efficient. Thus a hybrid approach of computing the resonances from the time domain scattering response and discrimination using the risk function defined, is fast enough to apply to targets of realistic size such as an aircraft at operational radar frequencies.

KEYWORDS:

• Comparison of FDTD and MoM
• Discrimination of objects
• Matrix pencil of function method
• Natural resonant frequencies
• Radar target discrimination
• Time and frequency domain response
• Vector fitting

ACKNOWLEDGEMENTS

The authors would like to thank Prof. N. Balakrishnan, Professor Emeritus at Indian Institute of Science (IISc), Bengaluru, Karnataka, India, for his valuable suggestions, guidance and support.

Additional information

Notes on contributors

S. Anuradha

S Anuradha received her master’s degree in electronic science from Bangalore University, Bengaluru in the year 1999. She is currently working as assistant professor in the Department of Electronic Science, Bangalore University, Bengaluru. Her research interests are computational electromagnetics, radar signal analysis for target recognition and signal processing.

G. U. Varalakshmi

G U Varalakshmi received her bachelor’s degree from Maharani’s Science College for Women, Bangalore University in 2015 and master’s degree in electronic science from Bangalore University, Bengaluru, in 2017. She is currently working in the Computational Electromagnetics project at SERC, Indian Institute of Science, Bengaluru, Karnataka, India. Email: [email protected]

Jyothi Balakrishnan

Jyothi Balakrishnan received the MSc degree in physics from Pune University, Pune, India in 1976 and PhD degree in physics from the Indian Institute of Science, Bengaluru, India in 1981. She superannuated from the Department of Electronic Science, Bangalore University, Bengaluru in 2016. Her interests are in material characterization, device physics, microwaves and signal processing. Email: [email protected]