Computational Models and Algorithms for Millimeter Wave Whole Body Scanning for Advanced Imaging Technology (AIT)

Download Project Report (Phase 2, Year 7)

Project Description

Overview and Significance

Active millimeter wave radar is being used for imaging objects concealed on the human body at security checkpoints. Currently employed systems are based on monostatic or quasi-monostatic configurations that can misrepresent areas of the target when the specular reflection is oriented away from the incident direction. Using computer modeling for both forward wave propagation and near field imaging and reconstruction, it is possible to determine the feasibility and limitations for various sensor configurations.

First, we report on the advantages of multistatic sensing (having multiple receivers separated from transmitters) for human body screening. One important result of this work has been the observation that without having several distinct transmitters providing multiple illuminating views of a target object, no receiver array configuration would be able to gather enough signal to reconstruct even one half of the body surface. We present a study of a fully multistatic mm-wave imaging architecture, with transmitters placed off the receiving aperture panel to provide the largest possible body surface reconstruction. A reduced number of transmitters allows for fast imaging, which minimizes the target motion effects. Multiple three dimensional simulation-based examples are presented to validate the proposed system.

Next, we describe the multiple transmitter implementation for the elliptical torus Blade Beam reflector antenna developed in R3-A.1 and R3-A.3. The primary effort is aimed at developing an efficient modeling method and then finding an optimum transmitter/receiver array configuration using the smallest number of antennas that still produces full surface reconstruction. A fast and efficient analysis method is suggested based on Lorentz reciprocity. Taking advantage of this proposed method, the sparse array configuration is found using Simulated Annealing, minimizing a cost function based on the sidelobe level of the combined transmitter/ receiver point spread function (PSF).

We generated simulated data and applied experimentally measured data for our moving reflector-based AIT imaging system. A more efficient way to build the scattering matrix for the reflector antenna system based on Lorentz Reciprocity was conceived and implemented.
Phase 2 Year 2 Annual Report
Project Leader
  • Carey M. Rappaport
    Northeastern University

Faculty and Staff Currently Involved in Project
  • Ann Morgenthaler

Students Currently Involved in Project
  • Mahshid Asri
    Northeastern University