Aircraft identification from RCS measurement using an orthogonal transform

A comparative study on target identification using the radar cross section (RCS) signature of an aircraft in both the frequency domain and the range d omain is conducted. A maximum likelihood method is employed to perform the identification process. Generalised likelihood identification when the rece ived RCS signal is attenuated by an unknown amount is also examined. Target identification could be quite computationally intensive since a large numb er of library reference signatures may have to be searched to declare an id entification. The use of an orthogonal transform is proposed to reduce the computational requirement. It is found that the discrete cosine transform i s very effective in compacting the RCS signature in the frequency domain, a nd the Haar transform is more efficient in the range domain. The applicatio n of orthogonal transforms can reduce the computational complexity by at le ast 50% while maintaining the same identification accuracy.