The music algorithm-based electromagnetic target classification for isolated targets from incomplete frequency domain data

This paper investigates the performance of a new electromagnetic target classification method to recognize isolated targets in the presence of scattered frequency domain data which may be severely incomplete at low frequencies. The suggested method utilizes multiple signal classification (MUSIC) algorithm to construct approximate pole location maps of the targets without determining the exact pole values. In this work, this method is validated for small-scale aircraft targets modeled by thin, conducting wires with incomplete data, which brings additional difficulty in target classification problems. It is shown that the method provides high accurate classification rates even when incomplete frequency domain data with low signal-to-noise ratio values are utilized while it needs only a few different reference aspects and small memory storage in classifier design.