On Modeling Hybrid IoT Malware Propagation Considering Size Distributions

• Main Authors: Jiong-Lun Li, 李炅倫
• Other Authors: Shin-Ming Cheng
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/7au86f

碩士 === 國立臺灣科技大學 === 資訊工程系 === 106 === With the increasing use of Interent of Things (IoT) in current network, more and more malware can launch attack through IoT network, and poses a critical threat to network security. Accurate malware propagation modeling in IoT network rep- resents a fundamental research issue which show how malware dynamic infected. Inspired from epidemiology, many paper analyze the mixed behaviors of delocalized infection and ripple-based propagation for the hybrid malware in IoT network by SI model. Howerver, this kind of model will overestimate the malware infected rate. Because the node mobility is random and node infected is not always ripple-based propagation. In this paper. We consider how malware propagates in network and infected node size distrubution. We formula a modify propage model based on SI model to analyze infection rate to t real network trace accuratly. Since we consider di erent size distrubution in di erent time stage, we need to give a transition condi- tion to make distrubution conform the network propagation rate. Our contribution is presente a stage transition condition to determine suitable size distrubution to formula model. Otherwise, based on our formulation, we give a judgment to quickly determine if the network will be completely paralyzed. Finaly, we compare other proposed model in real dataset and the result shows that experiments have been more closer to real data.