Millimetre Wave RADAR spectra simulation and interpretation for outdoor SLAM

Abstract

Millimetre Wave RADARs are more robust than most other sensors used in outdoor autonomous navigation in that their performance is less affected by dust, fog, moderate rain or snow and ambient lighting conditions. This paper describes a method to accurately simulate the range spectra using the RADAR range equation. This is very important in robot navigation (eg. SLAM) for generating predictions of what can be observed from different sensor locations and correspondingly, providing an interpretation for observed targets. To understand the MMW RADAR range spectrum and to simulate it accurately, it is necessary to know the noise distributions in the RADAR spectrum. A detailed noise analysis during signal absence and presence is carried out which shows various sources of noise affecting MMW RADARs. RADAR range bins are then simulated using the RADAR range equation and the noise statistics and are compared with real results in controlled environments. It is demonstrated that it is possible to provide realistic predicted RADAR power/range spectra, for multiple targets down range. Feature detection from the RADAR spectra based on target presence probability is then explained. The detection technique uses binary hypothesis testing. Results are shown comparing a new probability based feature detection with other standard feature extraction techniques such as constant threshold on raw RADAR data and Constant False Alarm Rate (CFAR) techniques. The results show that the proposed algorithm is more robust compared to other detection techniques as it does not require human assistance. This work is a step towards robust outdoor SLAM with MMW RADAR based continuous power spectra.