Comparison of Reference Clutter Removal Techniques in the Presence of Movement: A Case Study on 5.8 GHz Radar

Estimating vital signs through radars has gained attention in recent years because measurements are obtained non-invasively and without compromising people’s privacy. However, the use of radars for home care monitoring is still limited because the transmitted signal is affected by clutter, obstructions, and the movements performed by one or more people in the scenario. Sophisticated signal processing algorithms must be defined to address the effects of these phenomena, with Clutter Removal being one of the first steps in the estimation chain. Several algorithms have been described for removing clutter, among which Background Subtraction and Moving Target Indicator are frequently utilized. In this study, a comparison is conducted between these two algorithms, using the False Alarm Rate as a performance metric. The Radar signatures of human activities public dataset from the University of Glasgow was utilized for the experiments, which contains raw data from volunteers walking individually toward and away from a 5.8 GHz Frequency- Modulated Continuous-Wave radar in a cluttered scenario. It is shown that the Moving Target Indicator outperforms Background Subtraction in the tested scenario and verified the statistical significance of this result using the Wilcoxon Signed-Rank Test. Further research is needed to confirm whether the results obtained can be generalized to different home care scenarios.