Skip to main content
eScholarship
Open Access Publications from the University of California

UC Santa Barbara

UC Santa Barbara Electronic Theses and Dissertations bannerUC Santa Barbara

Autonomous and Predictive Systems to Enhance the Performance of Resilient Networks

Abstract

Over half the planet's population currently has access to the Internet in some form. Consequently, Internet usage has increased exponentially on mobile and fixed networks. In particular, video traffic now accounts for 80% of all Internet traffic, largely thanks to the rise of streaming platforms and video content from social media. With this tremendous growth, it becomes challenging to maintain consistent and high-quality service to all connected users. This is particularly true in the case of video sessions, which demand relatively consistent, high volume bandwidth to maintain high-quality sessions for end-users.

However, the presence of connectivity does not necessarily equate to usable service. Even in well-provisioned deployments, network resiliency can be an issue. Our own assessment of multiple networks across the US indicates that there is a wide disparity between reported coverage and actual usability. Numerous studies, including ours, indicate that LTE penetration in rural areas is far less than what is typically reported by providers and much less prevalent and lower quality than that in urban areas. Similarly, for fixed broadband, studies have observed highly variable median speeds across the globe, including South America, the Middle East, and Africa. Within the US itself, broadband speeds vary considerably as well. Finally, networks have to support multitudes of end-user devices (e.g., mobile phones, laptops, desktops) and a wide variety of application types, many with challenging and strict delivery requirements. As a result, ensuring high-quality, low-variability service quality under variable conditions is exceptionally challenging.

This dissertation is aimed at tackling this challenge. We assess how existing technologies and standards function in complex environments: lack of deterministic pathways for Internet connectivity, presence of congestion and other resource constraints, and temporal or spatial variability in network performance. Based on our extensive analysis of real-world conditions, we produce cellular and broadband systems designed to bridge the technological gap for end-users in challenging network conditions.

Main Content
For improved accessibility of PDF content, download the file to your device.
Current View