Research On Fog Computing-enabled Secure Hierarchical Multi-dimensional Data Aggregation

Industrial Internet of Things(IIoT)uses Internet of Things technology to collect data in industrial systems and detect equipment failures and system weaknesses in real time.However,with the rapid deployment of IIoT devices,massive real-time data are frequently transmitted and processed,so that the traditional network architecture is no longer suitable for IIoT in a big data scenario.At the same time,public data communication networks usually hide a variety of malicious attacks,leaving IIoT with a serious security and privacy crisis.Therefore,how to efficiently collect and process data under the premise of protecting the security and privacy of big data in IIoT has become a key issue to be solved urgently.Focusing on the current security and privacy issues in IIoT,as well as the efficiency and resource consumption issues in data aggregation schemes,the main innovative contributions of this dissertation include the following three aspects:1)Fog-assisted layered aggregation mechanism with efficient privacy preservation: Aiming at the problem that current IIoT data aggregation research schemes require a large amount of computing and bandwidth resources in the process of achieving secure and efficient data collection and application,a fog-assisted layered aggregation mechanism with efficient privacy preservation is proposed.In this mechanism,a layered aggregation architecture based on fog computing is first designed,which can effectively reduce the amount of data transmission.At the same time,the modified homomorphic Paillier algorithm is used to ensure the privacy and confidentiality of data.Then,based on lightweight hash chain technology,an efficient authentication mechanism is implemented to ensure data integrity.Furthermore,according to the Chinese remainder theorem,the cloud is able to obtain aggregated results of subregions and global regions to provide fine-grained data services,which enables it can provide fine-grained data services.Particularly,the fault-tolerance is supported in this mechanism to enable the cloud to decrypt incomplete aggregated ciphertexts.Finally,simulation experiments show that compared with other aggregation schemes,this mechanism can meet the security requirements in IIoT scenarios and reduce computation and communication overheads.2)Consortium blockchain based secure distributed aggregation and access mechanism: In view of the problem that most current data aggregation solutions are designed based on a centralized model,which make them facing single point of failure and tampering threats,a secure distributed aggregation and access mechanism based on consortium blockchain is designed to improve security and performance.This mechanism combines fog computing and consortium blockchain to design a distributed aggregation framework that solves single point of failure and tampering crises,and provides strong support for secure data collection,communication and storage.At the same time,a threshold version of the homomorphic Paillier algorithm is incorporated to design a secure and fault-tolerant access control mechanism,which can ensure the confidentiality and privacy of data on the blockchain.Meanwhile,an efficient anonymous authentication method is proposed to guarantee identity privacy and data integrity.Finally,simulation results show that this mechanism can resist malicious attacks and ensure system security,and it has lower computation cost and higher efficiency than other aggregation schemes.3)Fog-assisted privacy preservation multi-dimensional hierarchical data aggregation mechanism: Aiming at the problem that most privacy protection data aggregation schemes do not support heterogeneous data aggregation,a fog-assisted privacy preservation multi-dimensional hierarchical data aggregation mechanism is proposed.This mechanism designs a layered aggregation framework without the third-party organizations,and combines Horner's rule and homomorphic encryption algorithms to achieve privacy protection aggregation of multi-dimensional data,which significantly reduces the computation and communication overheads.Meanwhile,based on the Horner's rule,different aggregated results can be parsed from the total aggregated ciphertext,which can meet the needs of fine-grained analysis.Furthermore,lightweight elliptic curve algorithm and batch verification are combined to design a signature authentication mechanism,which can effectively guarantee identity validity and data integrity.Finally,simulation results prove that this mechanism has high security,flexibility and low energy consumption.