移动互联网病毒及免疫策略评价体系研究

The project is intended to establish a perfect system for evaluating the perniciousness of a mobile Internet virus as well as the effectiveness of an immunization strategy, by taking the expected profit of the virus under the strategy, which is closely related to the expected infected function (i.e., the function of expected numbers of infected nodes at all times), as the measure of both viral perniciousness and strategical effectiveness. First, introduce a set of propagation models of mobile Internet virus, and thereby propose a set of candidate models for calculating the expected infected function. Second, acquire a set of empirical expected infected functions by conducting experiments simulating the propagation of mobile Internet viruses. Third, establish a set of effective models for calculating the expected infected function by properly modifying their candidates. Next, understand the limit set of each expected infected function by applying the qualitative analysis theory of differential dynamical systems to the viral propagation models, and estimate the expected time needed for a node to be infected by applying the stochastic process theory to the viral propagation model. Finally, establish the desired evaluation system by integrating a set of effective models for calculating the viral expected profit. The resulting evaluation system can evaluate, in a comprehensive way, the viral perniciousness as well as the strategical effectiveness. Thereby the available resources can be allocated so that the most pernicious virus is first treated, with the cleverest immunization strategy. As a result, the loss incurred by mobile Internet viruses can be minimized.

本项目旨在以移动互联网病毒在一定免疫策略下的期望收益（与期望染毒函数密切相关）作为病毒危害性及免疫策略有效性的衡量标准，建立完善的评价体系。第一步，引入一系列移动互联网病毒传播模型，据此提出一系列初步的期望染毒函数计算模型。第二步，通过移动互联网病毒传播模拟实验，获取一批经验期望染毒函数。第三步，通过必要的修正，得到一系列有效的期望染毒函数计算模型。第四步，将动力系统定性分析理论应用于病毒传播模型，了解期望染毒函数的极限集；将随机过程理论应用于病毒传播理论，估计节点染毒平均用时。第五步，整合一系列有效的病毒期望收益计算模型，构建评价体系。所构建的评价体系能全面地评估移动互联网的危害性和免疫策略的有效性，据此合理地分配资源，采用最佳免疫策略优先处理最危险的病毒，将移动互联网病毒造成的损失降到最低。

项目组取得了如下四个方面的研究成果：（1）计算机病毒传播与控制理论，一是提出了一系列计算机病毒传播模型，揭示了模型的动力学性质，二是将计算机病毒控制问题转化为一系列最优化模型、最优控制模型或博弈模型，通过模型求解得到了最优或者较优的病毒控制策略，发表相关SCI论文17篇。（2）谣言传播与控制理论，一是提出了一系列谣言传播模型，揭示了模型的动力学性质，二是将谣言控制问题转化为一系列最优化模型或最优控制模型，通过模型求解得到了最优的谣言控制策略，发表相关SCI论文5篇。（3）病毒营销理论，一是提出了一系列病毒营销模型，揭示了模型的动力学性质，二是将病毒营销问题转化为一系列最优化模型或最优控制模型，通过模型求解得到了最优的病毒营销策略，发表相关SCI论文5篇。（4）高级持续性威胁（APT）理论，一是评估了APT的影响，二是将APT防御问题转化为一系列最优控制模型或博弈模型，通过模型求解得到了最优或者较优的APT防御策略，发表相关SCI论文6篇。