Doctor of Philosophy, The Ohio State University, 2013, Electrical and Computer Engineering
Automobile safety system has received tremendous attention in the past few years. Radar used in such system must be capable of detecting not only other vehicles but also pedestrian. Automobile radar working at 24GHz has been used in blind-spot detection (BSD) and automatic cruise control (ACC) system to track the distance and relative speed of on-road object. However, existing radars are limited to short detection range (30m) and low spatial resolution, making them less useful for pedestrian detection. A new frequency band 76-77GHz, recently designated by the Federal Communication Commission (FCC), the International Telecommunication Union (ITU) in Europe, and the Ministry of Internal Affairs and Communications (MIC) in Japan for vehicular radar. At such high frequencies, a longer detection range (100m-150m) and better resolution can be achieved. As such, it will enable more reliable detection of pedestrians in front of vehicles with lower false alarm rate. In on-road environment, multiple radar reflections may be generated from clutters like trees, trash cans, road surface, curbs and other vehicles. Therefore, it is necessary to identify unique pedestrian radar signatures in the 76-77GHz band to help discriminate them from clutters.
Experimental characterization of radar response of human targets at such high frequencies is not trivial and often inaccurate due to extremely short wavelength which makes radar measurement very sensitive to uncertainties associated with body position, orientation, and breathing motion. The variations produced by these uncertainties severely affect the reliability of the radar features extracted from measurement. In addition, the almost infinite combination of clothes, accessories and body postures significantly increase the time and resource required by this approach.
In this dissertation, analysing pedestrian radar signatures in the 76-77GHz band via numerical simulations is proposed to overcome the issues with measurements. (open full item for complete abstract)
Committee: Chi-Chih Chen (Advisor); John Volakis (Advisor); Baker Christopher (Committee Member); Olli Tuovinen (Committee Member)
Subjects: Electrical Engineering; Electromagnetism