Automatic Radar Target Identification Using Neural Networks
Automatic radar target identification is a very difficult task because data are very noisy and very few is known to decide which part of the signal is important. Identification techniques could help to build more efficient radars but also to better understand how this signal could be interpreted. We present here some connectionist techniques that have been tested on this task. They offer rather good identification rate but do not yet permit to understand the hidden structure of the signal.
KeywordsAspect Angle Range Profile Hide Structure Radar Domain Efficient Radar
Unable to display preview. Download preview PDF.
- F. Le Chevalier. Principes de traitement des signaux radar et sonar. Masson, 1989.Google Scholar
- T. Martinetz J.A. Walter and K. Schulten. Industrial robot learns visuo-motor coordination by means of “neural gas” network. In T.Kohonen, K. Mâkisara, and J. Kangas, editors, Proc. Int. Conf. Artificial Neural Net-works, pages 357–364. Elsevier Science Publishers B.V., 1991.Google Scholar
- H.J. Li and S.H Yang. Using range profile as feature vectors to identify aerospace objects. In IEEE Transactions on antennas and propagation, pages 261–268, 1993.Google Scholar
- T. Martinetz and K. Schulten. A “neural-gas” network learns topologies. In T.Kohonen, K. Mâkisara, and J. Kangas, editors, Proc. Int. Conf. Artificial Neural Net-works, pages 397–402. Elsevier Science Publishers B.V., 1991.Google Scholar
- M.I Skolnik. Introduction to radar systems. Mc Graw-Hill, New-York, 1980.Google Scholar