Landscape Detection by Leveraging Millimeter Wave Communication Signals

Abstract

Directional transmission is a prominent feature of millimeter wave communication systems, due to its near-optical propagation and the capability of beam forming. When the narrow beam provides high signal-to-noise ratio (SNR) due to the focus of power, it also implicitly hides bonus information for other purposes. One of the possible applications is the landscape detection by leveraging the reflection of beams: when a mobile user moves, the incident angle of beam may be changed; if it is non-line-of-sight (NLOS) beam, which can be determined due to the locations of base stations, the reflecting surface (building, constructions, trees, et al) can be estimated from the trajectory of the mobile user, the location of the base station, and the record of incident angles. This is in principle similar to the synthetic aperture radar (SAR). In this paper, the ideal case of infinitesimally narrow beam (as a straight line) is first studied. The estimation for the smooth curve of reflector is obtained. Then, the analysis is relaxed to the practical case of limited but nonzero beam width. The uncertainty in the direction is tamed by statistical modeling and filtering. Experiments in a 60GHz millimeter wave testbed are carried out to collect measurements and verify the proposed algorithms.