Collaboration and Competition in Interest-Driven Cyber Ecosystems
Type of DegreePhD Dissertation
Computer Science and Software Engineering
MetadataShow full item record
The next-generation cyber ecosystems are characterized by novel distributed paradigms composed of various network entities, such as individuals, devices, and software. The efficiency, scalability, and security of the systems highly rely on the interconnection of these autonomous and intelligent entities. These coexisting entities can be cooperative, selfish, or malicious, and make local-information-based decisions on their own behalf. Conventional optimization methods or security approaches dealing with pure engineering issues are either inefficient or impractical when applied to decentralized architectures in practice. Therefore, integrated approaches combining multiple scientific and engineering disciplines, including engineering, economics, and operations research, are required to transform the vision of modern cyber system paradigms into reality. This research addresses the practical concerns of the cyber ecosystem entities, analyzes their complex interactions and develops multi-agent models using mathematical tools including optimization and game theory. The research goal is to support interdisciplinary decision-making in real-world cyber systems. First, I propose a flexible network infrastructure sharing framework to incentivize cooperation between operators in a volatile and competitive market. Second, I explore the attackers' profit-driven behaviors and study the economic value of traded target information in security games. Finally, I design an incentive mechanism for crowdsourcing to facilitate the cooperation between the service providers and the service users.