Abstract

During the past decade, consumers all over the world have been showing an incremental interest in vehicular technology. The world’s leading vehicle manufacturers have been and are still engaged in continuous competitions to present for today’s sophisticated drivers, vehicles that gratify their demands. This has lead to an outstanding advancement and development of the vehicular manufacturing industry and has primarily contributed to the augmentation of the twenty first century’s vehicle with an appealing and intelligent personality. Particularly, the marriage of information technology to the transport infrastructure gave birth to a novel communication paradigm known as Vehicular Networking. More precisely, being equipped with computerized modules and wireless communication devices, the majority of today’s vehicles qualify to act as typical mobile network nodes that are able to communicate with each other. In addition, these vehicles can as well communicate with other wireless units such as routers, access points, base stations and data posts that are arbitrarily deployed at fixed locations along roadways. These fixed units are referred to as Stationary Roadside Units (SRUs). As a result, ephemeral and self-organized networks can be formed. Such networks are known as Vehicular Networks and constitute the core of the latitudinarian Intelligent Transportation System (ITS) that embraces a wide variety of applications including but not limited to: traffic management, passenger and road safety, environment monitoring and road surveillance, hot-spot guidance, on the fly Internet access, remote region connectivity, information sharing and dissemination, peer-to-peer services and so forth. This thesis presents an in-depth investigation on the possibility of exploiting mobile vehicles to establish connectivity between isolated SRUs. A network of intercommunicating SRUs is referred to as an Intermittently Connected Roadside Communication Network (ICRCN). While inter-vehicular communication as well as vehicle-to-SRU communication has been widely studied in the open literature, the inter-SRU communication has received very little attention. In this thesis, not only do we focus on inter-SRU connectivity establishment through the transport infrastructure but also on the objective of achieving delay-minimal data delivery from a source SRU to a destination SRU in. This delivery process is highly dependent on the vehicular traffic behaviour and more precisely on the arrival times of vehicles to the source SRU as well as these vehicles’ speeds. Vehicle arrival times and speeds are, in turn, highly random and are not available a priori. Under such conditions, the realization of the delay-minimal data delivery objective becomes remarkably challenging. This is especially true since, upon the arrival of vehicles, the source SRU acts on the spur of the moment and evaluates the suitability of the arriving vehicles. Data bundles are only released to those vehicles that contribute the most to the minimization of the average bundle end-to-end delivery delays. Throughout this thesis, several schemes are developed for this purpose. These schemes differ in their enclosed vehicle selection criterion as well as the adopted bundle release mechanism. Queueing models are developed for the purpose of capturing and describing the source SRU’s behaviour as well as the contents of its buffer and the experienced average bundle queueing delay under each of theses schemes. In addition, several mathematical frameworks are established for the purpose of evaluating the average bundle transit delay. Extensive simulations are conducted to validate the developed models and mathematical analyses.