Detection and Analysis Algorithms of punctual road events from vehicular data in the search for the optimal route An Overview of the AMNAM platform

Beaconing in vehicular networks (VANETs) is a greatly developed research field able to support safety applications like warning the drivers of potential dangers. Nevertheless, service messages, that uses different channels than beaconing, are also a core part to improve worldwide driving conditions and to minimizing traffic. This paper provides data analysis tools and algorithms for the detection and the analysis of road events. In particular, we develop a Java platform called “AMNAM” [1]. Its purpose is transposing and obfuscating raw simulation data into realistic information to, ultimately, detect the real world position of these events e.g., potholes, roadblocks, etc. Additionally, AMNAM computes the optimal route according to the detected events. Through extensive simulations, we demonstrate the efficiency of our proposed algorithms for the detection of punctual events. We also propose a new heuristic called “DirectionnalWeight” (DW) which adds a supplementary weight to the edges guiding the search towards the destination. Moreover, AMNAM's modified A* algorithm enforced by DW outperforms classic pathfinding algorithms found in the literature in terms of required iterations for its completion.