Road State Novel Detection Approach in VANET Networks Based on Hadoop Ecosystem

The problem of road congestion is becoming more and more serious in urban areas, which calls for solutions. This makes life in cities uncomfortable and costs a huge budget every year. Several resources are wasted during a bottling of fuel, weather, etc. In the ad hoc network of vehicles (VANET), useful information is exchanged between vehicles and traffic to avoid congestion and ensure easy fluidity. Vehicle-to-vehicle communication (V2V) is a means of transmitting this information in a VANET network. The immense amount of data that can be generated by a VANET network makes processing difficult for traditional tools to take advantage of its generated data. In this paper, we propose an approach based on big data tools to analyse the floating data in a VANET network and to detect the congested roads each based on occupancy rate of the roads, we detect the congested roads in the monitored area. Then, exctract more details about the congestion occurred by identifying the congestion interval and the peak instants in this interval. Simulations are done using the SUMO mobilisation generator and the NS-2 simulator.