Numerical estimation of electromagnetic backscattering from near-zone vehicles for the side-look vehicle-detection radar applications at millimeter waves

The development of vehicle-detection radar system has attracted much attention in recent years due to the increasing interest of smart car and traffic control applications [1, 2]. The success of such radar applications highly depend on reliable estimation of signal echo from the targets, which provide the information of target detection, vehicle identification, and the estimation of locations and speeds. In contrast to the conventional radar applications, where the targets are very far away from the antennas, the vehicles under detection are in the near zone. The detection highly depends on a proper use of antennas, in particular a proper radiation beamwidth for the target illumination. Since the radar detection is directly related to the electromagnetic (EM) backscattering, the estimation of various vehicles is apparently helpful to the radar system design. In this paper, we will present a full-wave numerical simulation over popular vehicle types including sedan and trucks, where the multi-level fast multipole method (MLFMM) of method of moment (MoM) is applied. The vehicles are assumed to be in the near zone of antennas, where the phased array antennas with various beamwidths are used to estimate the EM backscattering fields. To improve the detection accuracy, we also developed a moving signal average technique of signal estimation such that the influence of structure variation in a vehicle on the radar signals can be reduced. Thus in the presentation, the moving signal average method is also applied to the estimation of backscattering fields to exhibit their characteristics.