Research On Malicious Software Propagation Model And Defense Strategy In Mobile Wireless Sensor Networks

Wireless Sensor Networks (WSNs) are widely applied in many fields, such as Smart Home, Environment Monitoring, and Border Protecton like those commercial and military fields. However, since WSNs use wireless channels to transmit data and sensors have limited resources, such as battery energy and storage capacity, the security problem of WSNs is crucial. Compared with static WSNs, sensors in MWSNs not only have the ability of sensing, computing, storing and communicating, but aso have the mobile property. Due to the mobility of sensors in MWSNs, the topologies among sensors also changs as time passes. Thus, security in MWSNs faces with more challenges. Currently, the attack aiming at MWSNs is various. Attack inflicted by malware is a typical attacking behavior. When a contaminated sensor transmit data or control messages with its susceptible neighbors, malware are injected into those susceptible nodes, which may deplete battery energy of sensors, block regular communication channels, or even damage the integrity of regular data packets. Thus, how to mathematically model the propagation dynamics of malware in MWSNs become a crucial problem.Installing security patched is an effective method to prevent malware’s propagation. That is, through restoring the security flaws which may be employed by malware immunizes susceptible nodes, or through removing infected nodes by malware let nodes work normally. However, installing immunization patches continuously or removing malware at high intensity will badly occupy the limited communication channels; deplete limited battery energy, which definitely exacerbates communication delay and shorten the life-cycle of networks. Thus, proposing optimized counter-measures also is as critical problem. The goal of the proposed countermeasures is to not only prevent the further propagation of malware but also minimize the cost of countermeasures.In view of the above two problems, the research work of this thesis can be generized as the following two aspects:a) Mathmatically model the propagation dynamics of malware in MWSNs; b) Build optimized countermeasures to prevent the further propagation of malware in MWSNs.Compared with previous works in this field, the innovations and contributions can be summarized as follows:(1) Based on epidemic theory and Reaction-diffusion model, this thesis proposed a spatial-temporal dynamic model. Based on this model, we mainy discussed:a) the impact of network parameters on malware propagation, such as communication ridus, data sending rates and mobile velocity; b) Through analyzing the existence and stability of equilibrium points, we obtain the threshold of whether malware will continuously propagates or dies out as time passes. c) Based on the spatial distribution of malware predicted by the proposed model, we proposed target removing strategy. To the best of our knowledge, works in this thesis is unique in predicting spatial distribution of malware and corresponding target removing strategy.(2) Based on epidemic theory and pulse-differential equation model, this thesis proposed pulse-immunization strategy for preventing malware propagation. Based on this model, this thesis drives the maximum immunization interval. Compared with the continuous immunization strategy, the proposed pulse-immunization strategy let malware extinct in network with lower immunization frenquency so that which lower the cost of security operations. Moreover, the maximum immunization interval provides effective references to network governors.(3) In view of the mobile behaviours of sensors, this thesis proposed dyamic model to describe malware propagation in MWSNs. Based on this model, this thiesis proposed an optimized function which let at any final time, the network has least infected nodes and entire security cost. Selecting immunization rate of susceptible nodes and removing rate of infected nodes as optimized control variables, we obtain the optimized immunization rate and removing rate by using Pontryagin maximum principle, which is meaningful to build countermeasures to prevent malware propagation in MWSNs.