Local Vehicle Density Estimation on Highways Using Awareness Messages and Broadcast Reliability of Vehicular Communications

The density of vehicles on the road is an important metric of the traffic state. It can be used to estimate traffic congestion, predict travel times, optimize traffic flow, implement dynamic routing strategies, etc. Thus, vehicle density estimation plays a critical role in traffic management. At the same time, vehicles connected via vehicle-to-everything ( V2X) technology can exchange information by broadcasting wareness messages (AMs), making them naturally suited to estimate vehicle density. However, a vehicle might underestimate density due to the loss of some awareness messages from its neighbors. To address this, we propose a scheme that enables vehicles to locally estimate vehicle density based on the received AMs and the broadcast reliability of V2X. The scheme proposes an innovative method to estimate the ratio of the sensed vehicles to the total number of vehicles based on observed reliability metrics. Experimental results show that, compared with a method based solely on received messages, the proposed method improves the density estimation accuracy by up to 37% at an actual density of 0.28 Vehs/m, while not requiring additional communication overhead. We also present an example application of the estimated density, in which the Basic Safety Message (BSM) rate is dynamically adjusted.