A Lifi Approach Using Dynamic Q learning in Vehicular Networks

Vehicular Ad hoc Network (VANET) allows communication between vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications that support road safety as well as intelligent transportation systems (ITS) to avoid road hazards and share safety alerts.Even though traditional handover methods considering Wi-Fi and Light Fidelity (Li-Fi) technologies have seen significant improvement, dynamic network conditions experienced in VANETs need adaptive solutions.This paper presents a Li-Fi-based handover approach with a dynamic Q-Learning algorithm for deciding on the handover decision.The approach uses utilises reinforcement learning for vehicle traffic, vehicular mobility, network occupancy, and signal strength as parameters, thereby optimizing handover performance in highmobility scenarios.The simulated results show that the handover mechanism outperforms other techniques over multiple criteria's such as latency, handover success rate, network throughput and performs more decisional handover.