Security-Aware Incentive Mechanism Design for Cooperative Networks
Abstract
Society is organized largely in various kinds of networks (Internet, mobile networks, power grids, transportation systems, online social networks etc.). The well-functioning of these networks often relies on various levels of coordinated behavior and cooperation among the autonomous network users. For instance, in mobile ad hoc networks (MANETs), users forward traffic unrelated to its own use for other users. In all these cooperative networks, undertaking the cooperative actions is often costly (e.g. energy- and/or time-consuming) yet creates no immediate benefit for users who provide the cooperative service. Absent incentives for cooperation, these networks will fail to work effectively. A huge literature was dedicated to designing incentive mechanisms to promote cooperation in order to achieve specific network-level objectives. None of these works considers the potential security risks that individual users face when they are interacting with other users and taking cooperative actions. Security cost is completely different in nature from conventional types of costs ( e.g. energy consumption) since security risk is often interdependent among users and a result of the collective behavior of all users in the network. Without fully understanding how interdependent security risks emerge in cooperative networks, how they shape user incentives to cooperate, and how they affect the incentive mechanism design, simply applying existing solutions may cause significant performance loss or even catastrophic damage to both individual users and the overall network. The objective of this project is to develop a fundamental analytical framework for understanding user incentives in cooperative networks under interdependent security risks, and designing security-aware incentive mechanisms. To this end, this project pursues four specific tasks: ( 1) modeling cooperative networks under interdependent security risks; (2) understanding user incentives and the resulting interdependent security risks in random-mixing-based models; (3) understanding user Incentives and the resulting interdependent security risks in graph-based models; (4) incentive and protection mechanisms co-design for performance optimization. Successfully completing this project will answer a series of fundamental questions central to enhancing cooperative network performance with security assurance, such as: How do interdependent security risks influence the cooperative behavior of users and vice versa? What cooperative networks are formed by strategic agents under interdependent security risks? Does more cooperation always enhance the performance of the network? How should incentive mechanisms be adapted to mitigate the interdependent security risks? With cooperative networks becoming increasingly prevalent and cybersecurity remaining a key challenge, the proposed research will 3ddress the extremely compelling open problem of security-aware incentive mechanism design that current science has little knowledge of, thereby promoting a new research area of paramount importance. In particular, the proposed research will lay a scientific basis for understanding the interplay between cooperation incentives and the interdependent security risks emerging due to cooperation, and push the knowledge frontier of incentive mechanism design in novel network scenarios. The approach adopted in the proposed research represents a significant departure from existing works, which integrates epidemic modelling and game theory in innovative ways to derive critical knowledge for designing security-aware incentive mechanisms. Outcome of this research will potentially transform the way cooperative networks are designed and the associated security technologies are developed for protecting the Nation s various critical infrastructures.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Feb 14, 2019
- Source ID
- W911NF1810343
Entities
People
- Jie Xu
Organizations
- Army Contracting Command
- United States Army
- University of Miami