Late time response analysis in UWB radar for concealed weapon detection : feasibility study

Remote detection of body-worn concealed weapons or explosives (CWE) is a field of ongoing research in order to provide security against terrorist attacks that can potentially cause a threat to the security of the society. So far there is a gap on detection systems able to function at extended stand-off distances and which operate robustly under harsh environmental conditions. In this Thesis the feasibility of CWE detection via the usage of UWB radar is explored as a system able to fulfil such requirements. The hypothesis that the CWE detection is realised is based on the processing and analysis of the Late Time Response (LTR) of the human body which has been illuminated by the UWB transmitted signal. A specific set of LTR parameters characterizes the unique signature of the target. Therefore the existence of a CWE attached on the human body will influence the LTR characteristics and give the composite object i.e. human-CWE a different signature than the simple object i.e. human. The CWE detection methodology is verified by theoretical analysis, modelling and extensive laboratory experimentation. The aim of the research is to demonstrate the methodology of the approach and its ability to extract and characterise the LTR parameters-signature of simple object i.e. human or composite object i.e. human-CWE. In this research the grenade is considered as the CWE as it is a weapon that could be easily used during a terrorist attack. Throughout the thesis, the grenade is approximated as a conducting sphere of 0.0325 m radius since the sphere and the grenade have similar shapes, and since the reflected field of sphere is theoretically known. Most importantly experiments are performed with the sphere appended at the front, the back and the side of the human. Investigation of the way the LTR parameters are influenced by the existence of the CWE signifies the differences of the LTR signature between the human and humanCWE. So the resolution of the differences in the LTR of a human with and without a sphere as the main objective of the research, are presented in the Thesis. The results verify that CWE detection with the use of LTR is feasible under the experimental conditions presented. However a generalisation about CWE detection concerning all possible detection scenarios cannot yet be made. At this early stage of research, the identification of a specific CWE on the human and the robustness of the method to the human movement were out of scope of this Thesis and are considered for further research.