Research And Implementation Of IoT Firmware Security Situation Assessment System Based On DS Evidence Theory

At present,the Internet of Things security situation assessment system has been widely studied and developed.Based on this system,it can reflect the security situation information about the Internet of Things system in real time,which is of great significance for preventing major security risks and avoiding property losses.The current security situation assessment system often adopts the neural network model to analyze the data of situation elements.These models have complex structure,high training difficulty,excessive reliance on expert knowledge,low generality,and it is difficult for the system to dynamically adapt to the actual scene.With regard to the above problems,the IoT firmware security situation assessment with DS evidence theory is studied.Firstly,the evidence distribution matrix is calculated based on firmware vulnerability information.Secondly,the improved Topsis method is used to measure the credibility of evidence,and the weighted average iterative method is used to obtain the situation assessment results.Finally,according to the situation assessment results of multiple time,dynamic discounting evidence theory is used to carry out situation awareness.The main work of this thesis is as follows:1.A method of IoT firmware security situation assessment based on the improved Topsis theory is proposed.Three methods of measuring evidence credibility are put forward based on the Topsis method,in the first method the credibility of conflicting evidence is holistically measured by the evidence distribution matrix,in the second method the credibility of conflicting evidence is measured by aggregating the conflict between the two evidence,adding and normalizing,and in the third method according to the scene of situation assessment to improve expected positive solution vector.In the credibility experiment of different numbers of evidence,the third method fully limits the credibility of conflicting evidence,improves the credibility of mutual supporting evidence,and achieves significant effect in identifying conflicting evidence as soon as possible.The situation assessment work is carried out in the laboratory environment,and the result is reasonable and effective.2.A method of IoT firmware security situation awareness based on the dynamic discounting evidence theory is proposed.A security situation awareness method flow of IoT firmware is proposed.The evidence of situation assessment at different time is discounted with time sacle.The closer a evidence is to the current time,the smaller the discount rate of this evidence is,and otherwise the greater the discount rate of this evidence is.The proportion information of highly risk vulnerability is used for adaptive dynamic weighting,and the highly risk information at different time is discounted into the recognition framework,and finally the situation awareness result is fused.The situation awareness work is carried out in the laboratory environment,and the result is superior to the traditional DS evidence theory in the perception of high risk and critical risk.3.The IoT firmware security situation assessment system is designed and implemented.The crawler system is used to download firmware programs produced by 13 IoT manufacturers,and 369 firmware programs are analyzed by security analysis technology.The Vue framework is used to design the front-end page and the Django framework is used to design back-end functions.The system displays statistics data based on vulnerability information,situation assessment results in the recent 6 years and situation awareness result on the Dashboard page,which make this theory have engineering value.