Target shape estimation using the frequency domain response of ultra-wide band pulses

Abstract

This report aims to study the feasibility of Ultra-Wideband (UWB) for the classification of different shapes. This report also summarizes the findings and evaluates the accuracy of the technique by studying how the UWB pulses travel and is being reflected.

The UWB signal waveforms are captured using the Agilent Oscilloscope through a MATLAB program. These signal waveforms are analyzed and trimmed for poles extractions. The poles are derived using the Hankel Matrix’s singular value decomposition technique. These poles then serve as inputs for neural networks classifiers for shapes.