Mobility Management for Critical Time and Delay Tolerant Applications in Vehicular Networks

Abstract

In Vehicular Networks, the major desire of end-users satisfaction is towards guaranteed Quality of Service (QoS) when they pay for their required services. Many attempts have been taken to ensure seamless connection services, where users can move from one region to another region. Here the moving vehicles are deployed with wireless sensors that would be of heterogeneous or homogeneous technologies. Here the communication is among the vehicles nearby and between the vehicles and the fixed equipment in the road. Mobility management is one among the most demanding research issues in vehicular networks, which supports a variety of Intelligent Transport System (ITS) applications. ITS applications enable users to make transport more coordinated, safer and smarter. Vehicular networks are highly active, where the moving paths and network topologies are often changed which leads to time-varying optimal solutions. The critical time-sensitive applications are applications which cannot withstand delay and data are lost when there occurs delay. While the delay tolerant applications can withstand delay, where no data are lost. Both critical and delay tolerant applications have sparse communication interface which suffers from interrupted node connectivity or from delayed data delivery. When the nodes move in lieu of staying in a same position, the throughput is efficiently because of a large number of contacts. It is looked forward that Reinforcement Learning (RL) algorithm serve as an effective solution to the dynamic network topology, which learns to interact with the unknown environmental changes and take proper actions. Simulation evaluation shows that $\epsilon$ greedy performs much better in selecting the routes in which the vehicle to travel to get a seamless internet connectivity.