Mobility Based Improved Q-Learning Approach for RPL Routing Based Vehicular Adhoc Networks

Vehicular ad hoc networks (VANETs) are a futuristic technology where more innovations are introduced to achieve effective communication in the Intelligent Transportation System (ITS) based applications. Nowadays VANETs becomes huge in numbers and it maintains high speed communication in complex and dynamic environment which ensure reliable delivery and effective traffic monitoring in it. To achieve congestion free communication in VANETs in this research Mobility based Improved Q-Learning Approach for RPL Routing (MIQAR) is proposed which is the combination of initial RPL routing, Mobility based RPL routing and Improved Q-Learning based RPL routing. During the process of initial RPL routing the network topology is identified transmitting the hello packets where the general operations of RPL are performed. Through mobility based RPL, mobility is created for all the vehicles which record its migration and data forwarding history so that the network congestion can be identified and reduced. Through improved Q-learning process the contention window (CW) of each vehicle is analyzed that reduces the collision probability that results in the increase of successful packet delivery ratio during the process of data transmission in the network. The proposed MIQAR network model is simulated in NS2 where the vehicle mobility is generated in real time using the SUMO mobility generator. The parameters which are considered for the process of comparative analysis are packet delivery ratio, end to end delay, energy efficiency and routing overhead and as well it is compared with the earlier routing protocols such as QLAC-VANET and DQLDH-VANET. From the outcome analysis it is proven that our proposed MIQAR-VANETs achieve maximum delivery ratio and energy efficiency during the process of communication in the RPL based VANET networks.