Game Theoretic Model for Cluster-based Content Distribution in Vehicular Networks

Peer-to-peer based cooperative schemes significantly improved the efficiency of content downloading in vehicular networks. However, uncooperative behavior of peers can severely degrade the system performance, e.g. some vehicles may avoid downloading original data chunks to save the cost of 3G connection, and choose to be a free-rider. On the other hand, if too many nodes connect to the 3G/LTE, the 3G/LTE network may encounter serious congestion, and it would be a waste of network resource if many of them are actually downloading the same content. Faced with such dilemma, we propose a cluster-based content downloading scheme and model it with game theoretic tools. By analyzing the utilities of the game, we conclude that the Nash Equilibrium changes over different network parameters, such as packet loss rate, number of cluster heads, etc.