Dynamic Q-learning and fuzzy CNN based vertical handover decision for integration of DSRC, mmWave 5G and LTE in internet of vehicles (IoV)

Abstract

Internet of vehicles commonly known as IOV is a newly emerged area which with the help of internet assisted communication provides the support to the vehicles. Due to the access of more than one radio access network, 5G makes the connectivity ubiquitous. Vehicle mobility demands for handover in such heterogeneous networks. Instead of using better technology for long ranges and other types of traffic, the vehicles are using devoted short range communications at short ranges. Commonly, networks for handovers were used to be selected directly or with the available radio access it used to connect automatically. With the help of this, the hand over occurrence now takes places frequently. This paper is based on the incorporation of DSRC, LTE as well as mm Wave on Internet of vehicles which is integrated with the Handover decision making algorithm, Network Selection and Routing. The decision of the handovers is to ensure that if there is any requirement of the vertical handovers using dynamic Q-learning algorithms in which entropy function is used to predict the threshold according to the characteristics of the environment. The network selection process is done using Fuzzy Convolution Neural Network commonly known as FCNN which makes the fuzzy rules by considering the parameters such as strength of its signal, its distance, the density of the vehicle, the type of its data as well the Line of Sight (LoS). V2V chain routing is presented in such a manner that V2V pairs are also selected with the help of jellyfish optimization algorithm considering three metrics – Vehicle metrics, Channel metrics and Vehicle performance metrics. OMNET++ simulator is the software in which system is developed. The performance evaluation is done according to its Handover Success Probability, Handover Failure, Redundant Handover, Mean Throughput, delay and Packet Loss.