Research On Malware Spread Laws And Patch Injection Control Strategies In Industrial Internet Of Things

Industrial Internet of Things(IIoT)is an innovative mode that combines Internet of Things technology with industrial production.By integrating sensors,communication technology,cloud computing,big data and other technologies together,intelligent,automated,interconnected and data sharing of industrial equipment,production,products,supply chain and other links can be realized.However,the large number of sensors,robotic arms and other simple devices in IIoT as well as insecure communication channels induce malware to spread increasingly rampant in IIoT,and bringing disaster to many enterprises.Now,part of the research on malware in IIoT focuses on the intrusion detection of malware,and the other part focuses on the discovery of the spread law and spread control strategy of malware.The latter is of great significance for understanding the spread characteristics and behavior patterns of IIoT malware and effectively curbing their spread in the network.This paper mainly studies the modeling of malware’ spread law and the implementation of spread control strategy in IIoT,in order to solve the key problem of effectively curbing the spread of malware in IIoT.This paper is mainly divided into two parts,the first is to find the spread law of malware by considering the characteristics of malware in IIoT,and the second is use the patch injection approach to curb the spread of malware in IIoT by developing an efficient control strategy.On the one hand,we propose a spread model describing the spread law of malware in IIoT.Considering the spread characteristics of malware in IIoT,such as point-like outward diffusion,limited energy and communication radius,and replacement of industrial equipment,a new malware spread model in IIoT was proposed,and the differential equations of spread dynamics were established.In addition,an individual-group game describing the interaction between malware and nodes from the micro level was proposed to construct the spread payoff matrix and automatically calculate the infection rate and recovery rate.In order to evaluate the stability of the model,firstly,in theory,we calculated the basic regeneration number of the model to analyze the asymptotic stability of the malware-free equilibrium point and the endemic equilibrium point of the model.Secondly,in practice,16 groups of experiments were carried out to effectively verify and deeply analyze the stability of the model and the threshold of malware proliferation or extinction.On the other hand,we proposed a malware-patch optimal control strategy scheme in IIoT.A hybrid patch injection strategy was selected as the active defense strategy of network nodes against malware attacks,and the spread model was extended to the malware-patch composite spread model under its influence.In particular,three activity parameters were introduced as the control parameters of the optimal control problem of malware and patches defined by the differential game,which were used to solve the optimal control strategy by using the minimum principle of Pontryagin.Finally,we improved the algorithm double deep Q–network to realize the malware spread control algorithm DDPV in IIoT.By comparing with the classical algorithm deep Q–leaning network,it’s found that the algorithm DDPV reduces 89.2% in the proportion of the peak of unpatched and infected nodes,saves 55% in the time cost of spread,and increases at least 6.4% in the maximum reward.