Cooperative heterogeneous vehicular clustering for road traffic management / Iftikhar Ahmad

Iftikhar , Ahmad (2019) Cooperative heterogeneous vehicular clustering for road traffic management / Iftikhar Ahmad. PhD thesis, Universiti Malaya.

	PDF (The Candidate's Agreement) Restricted to Repository staff only Download (198Kb)
Preview	PDF (Thesis PhD) Download (3827Kb) | Preview

Abstract

Vehicular ad hoc networks (VANETs) are being incorporated with new wireless and telecommunication technologies for various applications. The incorporation of telecommunication technologies, such as long-term evaluation (LTE), within VANETs offers new application opportunities. The use of LTE for road traffic management, especially categorized under driving efficiency class, is gradually increasing with the increase in vehicles on the road. This increasing LTE network usage incurs added cost. In most cases, multiple vehicles in proximity require the same information from traffic information systems (TISs) while traveling in the same direction. This situation results in redundancy in acquired TIS information and poor LTE spectrum utilization. Vehicular clustering is a solution to minimize the use of LTE and the added cost. The first problem is that vehicular clustering faces traditional VANET constraints of high mobility and rapid topology changes, which increase cluster instability. A clustering solution is ineffective if instability is not addressed properly. The second problem is that, given that a cost factor is involved, a level of cooperation is required to determine who and why one should pay the cost and share accessed information to other cluster members (CMs). The third problem is that unstable cluster has a short lifetime, which is not useful for driving assistance and route planning applications that require temporal location information. Thus, developing a solution for traffic efficiency applications to address the aforementioned problems well is a challenging task. Existing solutions fail to address these problems comprehensively and are not designed specifically for traffic efficiency applications. Therefore, an emerging solution that incorporates new information communication technologies (ICTs) is advocated to reduce LTE usage, instability, and non-cooperation among CMs. This research focuses on the development of a novel heterogeneous network infrastructure-based vehicular clustering framework called destination and interest-aware clustering (DIAC). DIAC assumes that vehicles in proximity that travel toward the same destination or a milestone ahead form a cluster and considers the common interest of vehicles to access the same information. In addition, a clustering criterion is defined to avoid instability. A cooperative mechanism based on strategic game theory (SGT) is also developed to motivate vehicles to participate in cluster formation and to enforce vehicles to share accessed information and cost of use of the Internet. A control mechanism is defined as well to control the non-cooperative behavior and fair-use policy among CMs. Furthermore, a self-location calculation algorithm is developed to enable vehicles to calculate their location in the absence of global positioning system (GPS) signals. This implementation increases the level of synchronization of CMs with a cluster head, thereby increasing cluster stability. The proposed framework is simulated, and benchmarking is conducted with existing state-of-the-art approaches. The system model of the proposed solution is also developed and evaluated at macro level through formal verification. This research redirects existing efforts in acquiring TIS information toward a novel perception that is, designing a stable heterogeneous communication framework for cooperative data access to remote servers to minimize network usage and cost.

Actions (For repository staff only : Login required)