Reconnaissance using adaptive multi-carrier radar

摘要

This paper discusses the implementation and verification of original methods for performing short-range navigation and detection in common indoor environments. These methods utilize a software-defined ultra-wideband (UWB) orthogonal frequency division multiplexing (OFDM) radar system. Specifically, the system is used to identify the presence of targets while being subjected to realistic noise and clutter by using disturbance models created using experimentally-collected training data. Additionally, the same sensor is utilized to detect and the location and dimensions of doorways using the generalized likelihood ratio test (GLRT) and frequency profile matching (FPM), thus allowing for a localized navigation solution should the system need to move into adjoining rooms. These developments provide functionality that will assist a robotic platform that operates autonomously in unknown indoor environments.