Research On Security Technology For Internet Of Things System

With the development of communication technology and hardware technology,the application of the Internet of Things has gradually combined with social needs,which has greatly enhanced the automation of social life and improved the productivity of human society.The combination of the Internet of Things with cloud computing and big data technology has changed the production mode of traditional industries,and at the same time inspired the development ideas of new industries.In the process of integrating the Internet of Things and cloud computing into social production,due to the large amount of data collection and migration,the security problems of Internet of Things applications have gradually emerged,and the form of security has become increasingly severe.This article studies the related issues of IoT security from the IoT architecture,namely the IoT perception layer,the IoT network layer,and the IoT application layer.At the same time,the cloud-sea-fog structure of the Internet of Things is considered,and corresponding research is performed on key management,communication security,intrusion detection,defect prediction,and so on.The main contents of the study are as follows:(1)The key management of the secure part in Wireless HART of Industrial Internet of Things was studied.In the original protocol,the entire network used centralized key management.Because the energy consumption of network equipment is considered,the key update frequency in the original protocol is extremely low,and the data is encrypted by AES with a 128-bit key.Low password strength,low update frequency,easy to be cracked by blasting.In view of this problem,a key update scheme using transmitted data is proposed.During the key update process,no additional communication resources are used,the network life cycle is extended,and the frequency of key update is greatly increased,which improves the network security.Using the simulation experimental environment,the original key update scheme in Wireless HART is compared with the proposed key update scheme,and it is proved that the proposed key update scheme greatly improves the security of the network and extends the life cycle of the network.(2)The defensive scheme of communication delay attack in wireless network is studied.Cryptography-based security policies can effectively enhance data confidentiality and prevent data leakage.But the vast majority of Internet of Things applications have high service quality and low latency application requirements.As long as the response time of normal services can be extended and the data transmission rate of legitimate devices can be reduced,the entire security system will collapse.Aiming at the delayed attack of the Internet of Things,a defense scheme based on the transmission signal-to-noise ratio and adjusting the transmission power is proposed.Through research on Shannon's theorem and attack model,the IoT device dynamically adjusts its own transmission power according to the transmission power in the application environment and the attacker's transmission power to achieve the best transmission quality for legitimate devices in the application environment.And in terms of energy consumption,the normal life cycle of the equipment can also be guaranteed.Using the simulation experiment environment,the relationship between the transmission efficiency and energy consumption between the use of this defense scheme and ordinary transmission,legal equipment and attacking nodes is compared.The results show that the proposed defense scheme can guarantee the transmission quality,reduce the transmission delay,and quickly exhaust the energy of the attacker within a certain range of interference transmission power.(3)The intrusion detection technology using clustering algorithm is studied.The correspondence between the E-CARGO model and the characteristics of the Internet of Things was studied in depth,and the characteristics of the E-CARGO data for the intrusion detection of the Internet of Things were designed.On the edge of the Internet of Things,perform data clustering calculations and complete the Internet of Things intrusion detection function.Use edge computing technology to improve the intrusion detection efficiency of the Internet of Things and the response processing speed to the corresponding invading nodes.(4)The data cleaning method of the defect data uploaded by the perception layer was researched in the Internet of Things.The proposed data cleaning method retains the originaldata cleaning rules in the cloud system.Using the hypergraph model,the defect rules for repairing defect data are sorted,and the defect data is cleaned according to this rule order,which greatly improves the efficiency of data cleaning and reduces the calculation amount of repeated cleaning in the cloud.(5)The method of classifying the defect programs of the IoT device firmware in the IoT system is studied.Through the code extraction and irrelevant information cleaning of the firmware of different devices,the IoT firmware data is obtained.The data is organized into a Product Tree data structure,and codes with structural characteristics are clustered to obtain different firmware code classes.The proposed fine-grained firmware classification greatly reduces the total work of code analysts and defect repairers in the later stages.(6)A multi-source cross-project software defect prediction method for IoT software applications in an IoT system is studied.Using the source code data,project developer data,and project communication data of the IoT application project,the data characteristics applicable to the IoT application project are designed.Data is extracted according to these features,and a training data set is formed after preprocessing.Using semi-supervised learning algorithms in machine learning to complete software defect prediction for IoT software projects from multiple sources and cross-projects.