Artificial intelligence based handover decision and network selection in heterogeneous internet of vehicles

Abstract

Internet of vehicles (IoV) is an emerging area that gives support for vehicles via internet-assisted communication. IoV with 5G provides ubiquitous connectivity due to the participation of more than one radio access network. The mobility of vehicles demands to make handover in such heterogeneous networks. Usually, the previous works will directly select the network for handover or it connects with available radio access. Due to this, the occurrence of handover takes place frequently. In this paper, the integration of DSRC, LTE, and 5G mmwave on IoV is incorporated with novel handover decision making, network selection and routing algorithms. The handover decision is to ensure whether there is a need for vertical handover by using a dynamic Q-learning algorithm that uses an entropy function for threshold prediction as per the current characteristics of the environment. Then the network selection is based on fuzzy-convolution neural network that creates fuzzy rules from signal strength, distance, vehicle density, data type and line of sight. V2V chain routing is proposed to select V2V pairs using the jellyfish optimization algorithm that takes into account of the channel, vehicle characteristics, and transmission metrics. The proposed algorithms are then validated and compared with the existing state-of-the-art using OMNeT++ that combines with SUMO which gives real-time map-based architecture. The results indicate the superiority of the proposed algorithms in comparison with the existing state-of-the-art.