Electromagnetic Target Recognition with the Fused MUSIC Spectrum Matrix Method: Applications and Performance Analysis for Incomplete Frequency Data

The aim of this paper is to apply an electromagnetic target recognition method, which is based on the use of fused MUSIC spectrum matrices, to the case of incomplete frequency domain data. The aforementioned method was suggested recently and succesfully applied to both canonical and complicated targets in the presence of complete frequency domain data [1]. However, most of the real world applications involve the use of severely incomplete frequency data, especially missing low frequency information. In this paper, the performance of mentioned target recognition method is investigated for various incomplete frequency bands and for a group of five small-scale aircraft targets. As a result, the mentioned target recognition method is demonstrated to give still high correct decision rates even when the frequency band is reduced by 300 MHz at low frequency end and at the signal-to-noise ratio is decreased down to -5 dB level.