Distributed and cooperative information dissemination in vehicular networks

Abstract

Emerging vehicular applications need an adaptive communication framework. However, with unique network characteristics, these networks command challenges in developing an adaptive framework. Inter-vehicular communication projects and data aggregation techniques have significantly contributed towards successful contribution in these networks. However, they congest the network with increased network density and fail to address network fragmentation. Asynchronous communication is famous for distributed information routing and decoupling nature. However, it becomes necessary to address routing from the network layer perspective. This thesis combines the advantages of middleware and network layer routing to design an efficient framework for dissemination in vehicular networks during network fragmentation also reducing network overhead. The proposed framework combines the advantages of middleware and network layer routing. The application domain is designed using XML. Content based subscription model is used to convert XML based events to subscriptions. At the network layer, SMRP constructs a dynamic dissemination mesh overlay to forward the notification filter across the network to reach the interested subscribers. On the other hand, for efficient implementation of applications, cooperation plays a vital role in these networks. However, cooperation in such networks is highly ubiquitous due to the self-organizing structure and minimum network resource such as bandwidth. Reputation and Incentive methods are not applicable as following the past behavior of nodes becomes impossible due to high mobility. Traditional game theoretic approaches have extended a considerable contribution. However continuous learning over time is required. Hence, mandatory cooperation is not guaranteed. Evolutionary Game Theory is a theoretical framework to model group interactions in such complex networks to model the behavior of individuals. This thesis further contributes to model group interactions using Public Goods Game. The group interaction model using PGG is used to study the evolution of cooperation in vehicular networks under different networking conditions and compatible benefits.