MAS based Resource Provisioning in Vehicular Cloud Network

Modern day’s vehicles are equipped with various features, like-storage space, sensors and computing capabilities. They have the flexibility to communicate each other as well as with Road-side unit with wireless transceiver. Therefore by the combination of such resources meaningfully will create a momentous influence on the domain of communication network. The underutilized vehicular resources can be rented to customers as a conventional cloud infrastructure. In this position paper, we propose a multi-agent system based information exchange in vehicular cloud network. The problem we address is that to select a useful solution from a number of information gathered from other vehicles as well as from RSU is intrinsically difficult. Our (proposed) solution is a three-layer architecture which integrates cooperative agents (responsive vehicles), autonomous agent (RSU) and learning agents, the idea being to select a smart solution by combining information from all three agents using multi-objective fitness function. NSGA II algorithm has been used as a multiobjective method.