Survey on Deep Reinforcement Learning Protocol in VANET

Now a day’s numerous application are based on machine learning (M-L) techniques for enhancing performance of data, M-L techniques consist deep learning, reinforcement learning (RL), deep reinforcement learning (DRL), supervised learning (SL), unsupervised learning(UL), deep Q learning (DQL) etc. Vehicular Adhoc Networks (VANETS) most crucial aspect in modern network, which are decentralized over network. Here the challenges of decision making plays vital role for increasing performance, efficiency and minimizing energy consumption for that M-L techniques are utilized. RL unable to address the problem in a large-scale network. Hence RL is combined with DL and is known as deep reinforcement learning (DRL) for addressing challenges in large-scale network. Here we mainly focus VANET, which is the sub type of Mobile Ad-Hoc Network that offers connection between base stations of street– side and vehicles with an objective of giving secure and efficient conveyance. We also discuss different algorithms of M-L technique like KNN, Deep Q learning, SVM.