A Game-Theoretic Strategy for Pareto-Optimal Cooperation in Resource Constrained Vehicular Ad-Hoc Networks

摘要

Vehicular Ad-hoc Networks may be used for multitude of services including those related to human safety, traffic management and providing comfort to the users as applicable to Intelligent Transportation Systems, Smart Communication Systems, Internet-of-Things etc. In most scenarios, vehicles are owned by different entities and therefore it may not be in the best of interest of VANETs nodes to forward data belonging to different entities indiscriminately. This is so because it involves absorbing the incurred cost in terms of available resources. Such strategy may lead to packet drops and consequent delays leading to overall performance of the VANET-in-question. On the other hand if vehicles are readily agrees to forward any data that they receives, it would lead to a complete cooperation scenario, but might prove costly in terms of resources for the involved nodes. Both of these described conditions represents extreme scenarios and are thus undesirable. This opens up an opportunity for devising a new solution that could be built upon a balanced cooperation enforcement scheme involving all the participating nodes in a resource constrained VANET. In this paper, we present an approach for mathematical modeling of VANETs in form of noncooperative game. Using this model probability by which a particular vehicular node would accept the relay request of other vehicular nodes has been calculated by involving a combination of an available degree of resource and a resilient reciprocating strategy that we call V-TFT i.e. Vehicular-Tit-for-Tat. The presented simulation result demonstrates that the V-TFT converges to pareto-optimal solution.In addition, it is capable to perform so in presence of non-cooperative nodes and variable sized network.