Edge Enabled Identification of Vehicle Emergency Situation using Data Aggregation and Filtering Approach

Abstract

      Alert messages should to be transmitted to the vehicles instantly to save people’s life and time, while they commute. Vehicular cloud network is slow to deliver the messages, since transmission delay between cloud and vehicle is high.; Hence edge computing has taken new paradigm shift in cloud computing to handle delay sensitive applications. The end devices like vehicles, road infrastructure, Base stations can act as edge nodes/servers. Each edge node is capable of computation, storage and communication. Configured edge nodes communicate each other for data sharing to reduce data redundancy in turn efficiently utilize the bandwidth of the network. This work proposes a context aware data aggregation technique at edge nodes to identify accidental vehicles. The objective is to send accident alert messages to vehicles in the vicinity so save lives and congestion on roads with minimum delay. Mathematical model for data aggregation and filtration are simulated at both vehicular cloud and vehicular edge networks using EdgeCloudSim. Simulation results shows, edge nodes identify accidental vehicle and alerts others to change the route who are on that lane. Hence informed vehicles takes other route. Later edge node displays number of alert messages sent to the vehicles. Simulation results also shows performance graph between edge and cloud nodes with different parameters. Edge nodes takes half the time of cloud nodes for message transmission.  Cloud nodes takes 15% more execution time than edge nodes. Edge nodes are 16% more accurate than cloud nodes. Hence edge computing is emerging technology for delay sensitive applications in all domains especially transportation sector.