Research On Privacy-preserving Of Edge Computing In Ubiquitous Power Internet Of Things

In order to meet the challenge of energy transformation and electric power reform,the state grid proposes to construct ubiquitous power iot,and introduces edge computing into it,forming a cloud-pipe-edge-end architecture.In this framework,smart meter as one of the user-side acquisition equipment,directly connected with the user,power consumption data in real-time report,processing,analysis process is easy to cause the user’s privacy leakage.Therefore,it is of great significance to study the privacy protection of Ubiquitous Computing at the edge of power physical network.To solve the problem of system structure and privacy in the edge computing of Power Internet of things,an improved multi-secret key homomorphic algorithm is proposed to protect the privacy of power data Then,a privacy protection scheme based on differential privacy and Gan is proposed to solve the problem of privacy leakage during data reporting.The work of this dissertation is as follows:1.Analyzed the hierarchical structure of the ubiquitous power Internet of Things,and studied the possible types of attacks and the risk of privacy leakage for the reporting and processing scenarios of electricity consumption data,and constructed the ubiquitous power Internet of Things edge computing privacy protection system structure.2.In order to solve the problem of limited computing power of power terminal equipment,this dissertation focuses on the single secret key and high risk of leakage in smart meters,improves the multi-key homomorphism algorithm,and introduces the idea of re-encryption,in the decryption phase,the encrypted message of multi-key encryption is converted into decrypted message of single key,which realizes the independent decryption of cloud node,improves the privacy of secret key and reduces the communication and computing cost,the simulation results show that the scheme has better effect.3.Aiming at the problem of fault tolerance of privacy protection scheme in edge computing environment,this dissertation proposes a privacy protection scheme based on differential privacy and Gan,adding noise to each node to protect privacy,and training Gan to generate aggregated encrypted data in order to improve the data accuracy under fault tolerance,taking into account the problem that some nodes fail to report data in time,the Bloom filter and batch verification are combined to improve the efficiency of verification,and the efficient verification of mass data is realized.Finally,an experimental environment is set up and a simulation experiment is carried out.The experimental results prove the effectiveness of the scheme.