CGPS: A Collaborative Game in Parking-Lot Search

Abstract

Searching for parking lot can be very complicated due to the rapid movement of vehicles and the change of network topology, thus how to increase the parking efficiency becomes very important to this day. In this paper, we explore a parking-lot discovery algorithm to find the best and most reachable parking lot by evaluating the neighbors’ strategy based on game theory. Without the assistance of central server, vehicles exchange information by inner opportunistic communication based on distributed system of vehicular ad hoc networks (VANET). As the vehicles drive around the parking area, they continuously evaluate the competing neighbors’ cost, with different tendencies of cooperation, they can cooperate with the competitors in some degree, which is also good to raise the parking efficiency of the whole parking area. We formulate the parking-lot discovery problem as an instance of resource selection game, in order to achieve the cooperation between the vehicles, we change the strategy cost using a new pricing policy that each driver’s choice over the competing parking lot should reflect the extra cost of opponent’s lost, based on this principle to derive the Nash equilibrium, which reflects the total efficiency of all competitors. Finally, the simulations of the algorithms are presented to provide hint for parking-lot management that aims to raise the efficiency of the parking area.