Electromagnetic and Statistical Approach to Improve Radar Performance in Multipath Environments

2014-04-15T00:00:00Z (GMT) by Harun T. Hayvaci
In this work, multipath exploitation in radar target detection and parameter estimation is considered. Multipath is a natural phenomenon present in almost every propagation environ- ment and affects radar problems such as urban area sensing, through the wall radar imaging, and low altitude target detection and parameter estimation. In the past thirty years multipath has been viewed as a negative effect that needs to be mitigated; researchers have developed many signal processing techniques in that regard. Recently however, our view on multipath is changing: with the proper knowledge of the environment, multipath may be considered to be an opportunity to obtain additional degrees of freedom by providing virtual radar sensors that observe the target from different aspects. Thus, multipath exploitation has been receiving more interest lately in order to achieve more robust performance and degrees of freedom in detection and estimation problems. Although there is strong evidence that one can improve detection and estimation performance in radar through multipath exploitation, the amount of improve- ment still depends on the radar-target geometry, the corresponding multipath structure and the available signal noise ratio (SNR) levels. Thus, a quantitative analysis of the target probability of detection and parameter estimation accuracy with and without multipath exploitation, the topic of this thesis, is needed in order to fully assess the performance improvement.