Abstract
The evolution of Internet of Things (IoT) leads to the emergence of Internet of Vehicles (IoV). In IoV, nodes/vehicles are connected with one another to form a vehicular ad hoc network (VANET). But, due to constant topological changes, database repository (centralized/distributed) in IoV is of spatio-temporal nature, as it contains traffic-related data, which is dependent on time and location from a large number of inter-connected vehicles. The nature of collected data varies in size, volume, and dimensions with the passage of time, which requires large storage and computation time for processing. So, one of the biggest challenges in IoV is to process this large volume of data and later on deliver to its destination with the help of a set of the intermediate/relay nodes. The intermediate/relay nodes may act either in cooperative or non-cooperative mode for processing the spatio-temporal data. This paper analyze this problem using Bayesian coalition game (BCG) and learning automata (LA). The LA stationed on the vehicles are assumed as the players in the game. For each action performed by an automaton, it may get a reward or a penalty from the environment using which each automaton updates its action probability vector for all the actions to be taken in future. A detailed comparison has been provided by analyzing the cooperative and noncooperative nature of the players in the game. The existence of Nash equilibrium (NE) with respect to the probabilistic belief of the strategies of the other players in the coalition game is also analyzed.