mmWave Mapping using PHD with Smoothed Track Confirmation and Multi-Bounce Suppression

Paper i proceeding, 2022

The development of integrated sensing and communication systems together with increased carrier frequencies, larger bandwidth and massive antenna arrays are the key driving forces of future high resolution sensing services. This trend will turn the user device into a mobile radar, thereby opening interesting application scenarios in remote sensing. In this paper, we present a labeled probability hypothesis density filter for feature-based mapping in a mobile radar scenario. Smoothed track confirmation and multi-bounce signal suppression are proposed to enhance mapping quality and mitigate adverse effects of false landmarks. The mapping algorithm is evaluated using realistic ray-tracing data and the results imply that the mapping accuracy improves 31 % compared to a benchmark mapping algorithm.