Stable coalitions for urban-VANET: A hedonic game approach

Abstract

This paper tackles the problem of improving the coalitions' stability for Quality of Service Optimized Link State Routing (QoS-OLSR) protocol in urban Vehicular Ad-hoc Networks (VANETs). VANETs have introduced communication among vehicles for the Intelligent Transportation System (ITS) allowing several safety and user-oriented applications to be implemented in an ad-hoc manner. The routing protocols in VANET play a significant role in the delivery of data packets for such applications. However, the rapid changes in the topology of urban-VANET affect the relative stability among moving vehicles and consequently cause a high drop in the network performance. Several works considered urban metrics and game theory models for stable relay selection and cluster formation where the stability of a relay is measured compared to all the neighbor vehicles rather than selection based on relative stability between a vehicle and its relay. In addition, some works proposed matching theory that focuses on the individual preferences of players, to introduce stability. Alternatively, the Hedonic game considers preferences and enhances stability through joining coalitions to improve the overall utility of all participating members in each coalition. Therefore, in this paper, we propose a two-phase Hedonic game model that improves the coalitions' stability for QoS-OLSR in urban-VANET by considering the urban characteristics such as traffic lights, intersections, and lanes' directions. Players play the game in an ad-hoc manner by exchanging a set of messages and join coalitions that include relatively more stable members. Hence, the relative stability among coalition members is the utility of the coalition. A coalition utility depends on the urban characteristics which affect several metrics such as the time to leave the street and the probability of the direction to leave the nearest intersection. Simulations, that are conducted using SUMO and NS3, show high network performance in terms of packet delivery ratio, throughput, and end-to-end delay compared to benchmark protocols.