Research On Security Protocol Of Smart Grids Based On Homomorphic Cryptosystem

The smart grid significantly improves the reliability,efficiency,security,and sustainability of electricity services.However,the development of smart grid brings convenience with a series of security and privacy issues.The smart grid collects more and more users' metering data with the continuous increase in the number of smart grids,which means the fine-grained metering data analysis may leak the personal privacy information.Therefore,applying privacy-preserving data aggregation mechanism to meterreport protocol for smart meters has great application value.Based on the actual requirements of the smart meter data aggregation protocol,this paper focuses on the analysis and research of the homomorphic cryptosystem and aims at solving the issues of privacy protection,secure transmission and limited resources.The main works are as follows: Firstly,after analyzing the characteristics and status of smart grid,it is pointed out that there are many isolated smart grid devices in the current smart grid.An authentication protocol based on the homomorphic cryptosystem is designed for isolated smart grids.The proposed protocol consists of an encryption scheme with additively homomorphic property and a linearly homomorphic signature scheme which can prevent the theft or manipulation of metering data and securely aggregate users' metering data.The proposed protocol is deployed in the environment of smart grid for analysis at the same time.Secondly,most power usage data aggregation protocols do not use a formal methodology for evaluating their security or privacy guarantees under the colluding attacks.To solve this problem,a security definition and model of data aggregation protocols under the colluding attacks is proposed.The security model is applied for security proof of two data aggregation protocols based on the addictively homomorphic encryption scheme and it is found that there exists security flaws in practical applications of the two protocols.The proposed security model can also be used for other data aggregation protocols,and it is beneficial for secure design of protocols as well.