Decentralized Pedestrian Density Maps based on Sidelink Communication

Given the present situation brought about by the COVID-19 pandemic, social-distancing and avoiding dense situations for pedestrians is of uttermost importance. However, to know if the current path taken by a pedestrians will lead to crowded locations requires realtime, up-to-date information on the density of pedestrians in the local area. Utilizing this, pedestrians (i.e. route recommending applications) could avoid crowded areas thus mitigating or reducing exposure risks. State-of-the art system relying on centralized backend services do not provide necessary realtime information and might lead to privacy concerns.In this paper, we propose an alternative, decentralized approach to disseminate Decentralized Pedestrian Density (DPD) maps which provide realtime density data in close to medium proximity. The DPD map is based on merging and aggregating position beacons into density measures disseminated by each pedestrian. We use LTE Advanced sidelink multicast communication to disseminate beacons and DPD maps. To evaluate our approach, we simulate an urban scenario modelling a small/medium area located within the city center of Munich, Germany. This requires a realistic modelling of the pedestrians movements and of the communication system. Therefore, we make use of the open-source simulation framework CrowNet, maintained at our group which couples the microscopic pedestrian dynamics simulator Vadere with a system-level simulation of LTE-A (SimuLTE model in OMNeT++).Initial results demonstrate the feasibility of the proposed approach but also indicate that overload situations can lead to the dissemination of outdated information, implying a need for situation- and load-adaptive transmission schemes.