Incentive Mechanism Design For Interdependent Security Investment Game

In recent years,with the rapid development of network science,network security issues for complex networks have received widespread attention.Since relying on firewalls,intrusion detection and other technical means can not solve many problems which faced by current network security,more and more researches have begun to focus on improving network security by improving network users' security policies.Since cyber security is non-exclusive as a public good,and users' security risks are interdependent,selfish.A large number of network users may take free-riding behaviors in order to maximize their own revenue.In the security investment game,the user's free-riding behavior has become the main factor affecting network performance,so the design of incentive mechanism to suppress the selfish behavior of complex network users has received more and more attention.In this thesis,the research work of existing incentive mechanism is summarized,then,the incentive mechanisms are designed for the security strategy of random network and square network users,and the effectiveness of the mechanism is deeply studied.Tax-based security investment incentive mechanism and safety information sharing incentive mechanism based on cross-report are proposed,the main research work includes the following two aspects:Propose a tax-based security investment incentive mechanism which is called quantitative externality.The quantification of externality mechanisms is inspired by the principles of the VCG mechanism,where each network user pays for the loss of its immediate adjoining network user revenue.In order to meet the principle of voluntary participation,this article allows network users to voluntarily choose whether they are willing to join the mechanism according to their own status.Network users with lower revenues will consider whether the incentive mechanism can bring them higher returns,if possible,join the mechanism,and vice versa.At the same time,the thesis analyzes the equilibrium state of the incentive mechanism by using evolutionary game theory,and verifies the performance of the incentive mechanism by experiment.The mechanism can improve the network security investment while satisfying the voluntary participation,thus effectively improving the network security level.Propose a security information sharing incentive mechanism based on cross-report.In order to encourage network users to share real security information to improve the network security level,the user's immediate neighboring users are required to make predicted cross-reports to other users according to the security status similarity.The difference between these reported values indirectly affects user revenue.If the gap between the two is too large,it will lead to a rapid decline in revenue,thus encouraging users to report real information.This thesis introduces the cross-reporting incentive mechanism by simulating network virus propagation,using the degree of virus infection at steady state to initialize the user's real shared information.The simulation results show that the mechanism can effectively improve the authenticity of users sharing security information without additional verification mechanism,thus improving the network security level.