| Summary: | 碩士 === 國立臺灣科技大學 === 資訊工程系 === 105 === The development of communication technology has provided malicious users formidable means to launch attacks through Internet of Things (IoT). The features of IoT devices which include constrained resources, heterogeneous links, and vulnerable usability facilitate the malware propagation, thereby raising new challenges on handling IoT-empowered malware for cyber security. The malware propagation control scheme in traditional wireless networks where nodes can be directly repaired and secured is not suitable for IoT since IoT devices are hard to be patched. To efficiently and effectively suppress the spreading of harmful information, blocking malware via patching the intermediate nodes (e.g., base stations, access point) instead of the infected mobile devices becomes our better choice. On the other hand, we analyze this network by exploiting the well-known epidemic model and the concept of spectral clustering. The clustering algorithm can avoid that all the patching resources being given to the area with the highest average traffic volume and neglect the intermediate node in other areas which also need to be patched. This article proposes a novel clustered traffic-aware patching scheme to select important infrastructures to patch, which is suitable for the IoT system with limited patching resources and response time constraint. We conduct experiments on real-world trace datasets by using Opportunistic Network Environment (ONE) simulator to show the advantage of clustered traffic-aware patching scheme in mitigating malware propagation, and clustered traffic-aware patching has better performance than intuitive degree-based patching.
|
|---|
