Problem statement

Nowadays the organization of mass events is faced with heavier constraints related to safety. Many of these constraints can readily be addressed either during event preparation (choice of the event area, or of contingency plans), or during the event (appropriate fencing), but efficient crowd management can only be roughly achieved despite its importance for safety. Crowd safety management is closely linked to the monitoring of crowd dynamics. This project investigates the use of Wi-Fi-based passive radars to monitor crowd dynamics. Compared to the use of cameras which first comes to mind, the radar technology is appealing as it is robust to low lightning conditions and less prone to privacy issues. While radar systems usually implement both the transmitter and receiver, passive radars make use of third-party transmitters in the environment. The project aims at conceiving and assessing analytically, numerically and experimentally a multi-static passive radar based on the observation of emerging Wi-Fi signals to track individuals or groups of individuals outdoors.

Contributions

  • Study of interferences on Wi-Fi-based passive radar
  • Super-resolution algorithms for range and speed estimation
  • Crowd dynamics simulator
  • Multi-antenna experimental passive radar setup based on software-defined radios
  • Static clutter removal on range-Doppler map
  • Data association and tracking of humans