ECHO SIGNAL SEPARATION USING ICA PROJECTION PURSUIT ALGORITHM

Sanchi Kansra, J Valarmathi, Vikas Pandey

Abstract: In any RADAR system, the foremost challenge is to identify the target echo signal from the mixture of unwanted echoes called clutter. RADAR equipment generates linearly frequency modulated pulses which when reflected back gives the information about the target’s range, location and velocity of approach. A significant amount of clutter introduced in RADAR signal is generated by the surface of sea. This surface type of clutter is modelled as Weibull noise source which is mixed with target echo. These are treated as two independent sources. Assuming that the echo signals are independent from each other and are non-Gaussian in nature we use Independent Component Analysis’s Projection Pursuit Algorithm to separate the components so that these components can be analyzed further for target detection an ranging.

Keywords: RADAR, ICA,Blind Source Separation, Weibull Noise.

DOI: https://doi.org/10.15623/ijret.2016.0506052