Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article

Degree

Master of Engineering Science

Program

Electrical and Computer Engineering

Collaborative Specialization

Artificial Intelligence

Supervisor

Shami, Abdallah

2nd Supervisor

Haque, Anwar

Co-Supervisor

Abstract

This thesis covers two topics that optimize a network-related problem subject to environment-specific constraints; placing vehicular applications and executing network traffic assignment changes. The first topic introduces an optimization model, Resource and Delay-aware V2X service Placement (RDP), and a baseline approach that only considers the resource requirements of vehicular services. Both are responsible for placing vehicular services used by vehicular applications in an edge computing environment. Under different simulation scenarios, the results obtained by RDP satisfy the delay requirements of vehicular applications as opposed to the baseline approach. The second topic examines the efficient execution of inter-domain traffic changes under bandwidth, monetary, and infrastructural constraints. An oracle algorithm and two heuristics are formulated, and evaluation criteria are devised to reflect the constraints. These algorithms are evaluated on different networks, and the results reported show that OrderSteps (OSS) heuristic satisfies the constraints and outperforms the oracle implementation in terms of run-time.

Summary for Lay Audience

This thesis examines two topics that are related to autonomous vehicles and Internet data traffic routing. Autonomous vehicles are envisioned to transfer the authority in the decision-making process from the human driving it to the computer operating the vehicle. To achieve this vision, the vehicle must be equipped with sensors that are gathering information from their surrounding environment and applications, resembling the mobile applications, responsible for extracting useful insights that can be valuable in risky and dangerous situations. Such situations require fast processing and response in the space of 10ms from these applications. Unfortunately, all these applications can not be installed on a single vehicle due to its limited capacity. Therefore, the first topic investigates this issue by first breaking down these applications into smaller independent components for capacity purposes. After that, these components are installed on infrastructure of varying capacity close to the vehicles, so that the components’ requirements in terms of capacity and the vehicles’ requirements in terms of response time are satisfied. Currently, people are extensively using their mobile phones and their smart TVs to surf the Internet and to watch movies. Any activity related to the Internet requires, on the network level, to route the data from source to destination as a response for a specific request. To better manage the routing process, network operators have divided the network into a set of routers whereby each set is defined as an Autonomous System (AS). Each of these Autonomous Systems (ASes) forms relations with other ASes to route the traffic exiting them to its destination. The main objective of each AS is to maximize the profit and optimize its resource usage which is achieved by mandating the routing of traffic in specific ways in the infrastructure they manage. Each network operator is concerned in routing traffic generating mass data such as YouTube and Netflix traffic. Due to the traffic’s dynamicity, the network operators’ respective routing strategies change frequently, so that they keep adhering to their objective. The second topic covers transitioning between two routing strategies in a way that satisfies the network operators objectives.

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Available for download on Tuesday, August 31, 2021

Share

COinS