基于微分博弈的在线社交网络恶意程序传播优化控制方法

针对在线社交网络（OSN）易传播恶意程序的现状，通过扩展传统的传染病理论，在考虑防御者和恶意程序主观努力度的基础上，提出了能确切描述OSN恶意程序的微分方程模型。利用微分博弈，建立了能反映防御者和恶意程序交互过程的OSN“恶意程序防御微分博弈”模型，当恶意程序动态改变其最优控制策略时，为防御者给出最优动态控制策略。实验结果表明，提出的方法能明显地抑制OSN恶意程序的传播，为防御OSN恶意程序提供了新途径。

Differential Game-Based Optimal Control Method for Preventing Malware Propagation in Online Social Network

Online social network（OSN）is prone to propagating malware because of their special characteristics.By developing traditional epidemic theory and considering effort intensities of the defender and malware，a differential-equation model to suitably describe characteristics of OSN malware was proposed.Using the differential game，a malware-defense game was constructed to reflect interactions between the defender and malware.Thus，optimal dynamic control strategies for the defender were given when the malware dynamically changed its optimal control strategies.Numerical experiments show that the proposed method is obviously able to suppress the malware propagation.Therefore，a novel way to defend OSN malware was provided.