Date of Award
Doctor of Philosophy (PhD)
Leveraging multiple wireless technologies and radio access networks, vehicles on the move have the potential to get ubiquitous broadband Internet connectivity. Many studies have put lots of efforts on vehicle-to-vehicle networks for relaying strategies, popular content distribution, etc. However, in dominant infrastructure-based vehicular networks, supporting continuous and fast data transfer for today's prevalent services, e.g. video streaming, for vehicles anytime and anywhere is still a difficult research problem. By looking into such problem, impacts such as intermittent connectivity, dynamic network topology, fluctuating signal coverage, and inefficient transmissions, all result from two root causes in vehicular network- mobility and limited infrastructure resources. This dissertation investigates three core functions of infrastructure-based vehicular networks in the presence of the two causes: 1) network selection; 2) data forwarding; 3) data retrieval. By leveraging a compute cloud's abundant computing and data storage resources, three cloud-based strategies are proposed to achieve robust connectivity and efficient transmission for vehicular networks. First, network selection is essential to maintain robust connectivity for vehicles on the move. Serving a large number of vehicles, today's centralized and distributed network selection solutions require sophisticatedly designed utility functions and optimization with complex computing, reducing flexibility and efficiency. A fast, game-based network selection scheme is proposed in this dissertation. Vehicles can select best access networks through a coalition formation game approach by leveraging wider scope network information for decision making. A one-iteration fast convergence algorithm is proposed to achieve the final state of coalition structure in the game. Through extensive simulation, the proposed network selection scheme was shown to balance system throughput and fairness without the need of an explicit fairness metric in the utility function. The algorithm efficiency showed eight-fold enhancement over a conventional coalition formation algorithm. Second, efficient data forwarding strategies are important for transferring data to vehicles. Previous studies have developed numerous methods to disseminate common content among vehicles using broadcasting or multicasting, however, these methods cannot automatically be applied to multiple unicasts to vehicles on the road. An inter-session network coding scheme for Internet-to-vehicles unicasts is proposed. The proposed scheme makes efficient utilization of network coding by applying the proposed dynamic, optimal grouping algorithm. Leveraging cloud assistance, the grouping algorithm can compute optimal flow routing using real-time network topology information. Extensive simulation showed that the proposed scheme achieved less delay by up to 6+ times than conventional routing for UDP traffic and zero end-to-end packet loss rate. Within a general range in terms of number of vehicles (10~40) in a 2 km*2 km area and average speed (10~50 mph), the proposed scheme maintained an apparent advantage over conventional routing. Third, reliable and efficient content retrieval on demand from vehicles on the move becomes more and more important. Today's transport mechanism in vehicular networks still inherits traditional TCP/IP's client-server manner, working with opportunistic scheduling. However, opportunistic scheduling cannot fundamentally get rid of the vulnerability of client-server architecture using TCP/IP in the presence of fast speed. Content-centric vehicular networking (CCVN) is proposed as a potential framework to achieve efficient data retrieval in vehicular networks instead of patching over TCP/IP. The connection-less, in-network caching and distributed characteristics of CCVN are able to adapt to vehicular environments easily. CCVN uses a cloud-based face management mechanism for enhancement. Emulation results showed CCVN has promising results compared to three selected opportunistic scheduling schemes using TCP/IP- MV-MAX, load-reducing scheduling, and pre-caching scheduling. The promising performance of the proposed cloud-based strategies validates the usefulness and importance of the cloud-based system. This dissertation is expected to provide guidance for implementing cloud-based applications in vehicular networks in the real world, such as telematics and location-based services.
Xu, Ke, "Cloud-based Strategies for Robust Connectivity and Efficient Transmission in Vehicular Networks" (2015). All Dissertations. 1507.