With the dramatic growth of vehicles around the world, Vehicular Ad-hoc Networks (VANETs) have been proposed as a solution to advance road safety, improve transportation efficiency, and satisfy road users. In the VANET environment, vehicles communicate with each other and with road infrastructure in an ad-hoc manner. This communication may be safety-related or non-safety-related and may often include vehicle information (e.g., location, direction, speed, and control), road conditions, and events. A key component in assessing the veracity of the information is the trustworthiness of the information source. Thus, trust evaluation is one of the main requirements of VANET design. In this work, we investigate performance improvements in the trust evaluation framework of VANETs.
First, we propose a risk-based trust evaluation model (RTEAM) to estimate the risk of taking action or refraining from action regarding a reported event (in case of receiving conflicting messages about the event's existence). Some trust metrics such as direct trust, hop-based trust values, proximity to the event, and consequences of acting on a wrong decision are used to estimate the risk of the vehicle’s actions. Vehicles make individual decisions by seeking the action with the lowest risk.
Second, we propose a fog-based reputation evaluation model (FREM) to support trust management framework. We promote fog computing as a new paradigm since it can provide several services to users in the edge layer. In our work, Fog supports the decision-making process in the reputation evaluation framework. Fog nodes play a key role in collecting vehicles' reputation records and cooperating with the roadside units (RSUs) to update these records. We propose the use of Digital Trustworthiness Cards (DTC), where the latest reputation evaluation of a vehicle automatically appears on its card. The benefits of the DTC are twofold: 1) the communication load on vehicles is reduced, and 2) historical trust records are established for each vehicle. We also take advantage of fog’s familiarity and greater knowledge of the vehicles that frequently visit its zones; with more intimate knowledge, fog can smartly employ vehicles to perform specific tasks based on their experiences. Further, we implement a strategy for establishing trust based on specific task categories. This permits a nuanced evaluation of the vehicle best suited for the task at hand and has the further benefit of preventing malicious vehicles from being naively trusted based on successful completion of unimportant or non-safety-related tasks.
Finally, we expand the role of the fog in the decision-making process when vehicles need to ensure the existence of serious events. We propose a fog-based event validation model (FEVM) to validate the event’s existence through cooperation between vehicles and fog nodes. The vehicles are used as mobile fog nodes, which compute their confidence in events based on the available information. Fog nodes then validate the event after combining vehicles’ confidence values by applying the Extended Dempster-Shafer (EDS) theory of evidence. To test our proposed models, we conduct many experiments to investigate their performance and compare them with other existing models.
| Identifer | oai:union.ndltd.org:uottawa.ca/oai:ruor.uottawa.ca:10393/43167 |
| Date | 19 January 2022 |
| Creators | Atwah, Rasha |
| Contributors | Flocchini, Paola, Nayak, Amiya |
| Publisher | Université d'Ottawa / University of Ottawa |
| Source Sets | Université d’Ottawa |
| Language | English |
| Detected Language | English |
| Type | Thesis |
| Format | application/pdf |
| Rights | Attribution-NonCommercial-ShareAlike 4.0 International, http://creativecommons.org/licenses/by-nc-sa/4.0/ |
Page generated in 0.002 seconds