Efficient Privacy-Preserving Data Aggregation and Replication for Fog-Enabled IoT

Abstract

With the increasing popularity of Fog computing to provide computation, analysis and storage of data at the edge of IoT networks, the fulfilment of data privacy requirements over fog networks can be seen as one of the biggest security challenges. Data aggregation is considered an essential privacy requirement as it combines data from different IoT devices to protect the data leakage of an individual IoT device. It also reduces data redundancy while improving data analysis speed in Fog-enabled IoT networks. For preserving the privacy of data aggregation, the heavyweight cryptosystems are considered, which faces issues related to performance overhead and single point of failure risks due to data aggregation at a single fog node. In addition, no secure data replicas exist for data recovery and reliability in case of a data breach in Fog-enabled IoT applications. This thesis proposes an efficient privacy-preserving scheme for data aggregation to overcome the limitations of fog-enabled IoT applications. This thesis also proposes an efficient privacy-preserving data replication scheme for data reliability and recovery. The proposed data aggregation scheme is based on lightweight data encryption and data division method. This method effectively divides data according to Level of Privacy (LoP). It distributes the data among participating fog nodes for aggregation and storage processing and reduces computational and memory overhead in the processing simultaneously. The proposed data aggregation scheme is further extended to optimize the time and energy consumption of the data division method. The multi-objective optimization method in defined in this thesis, which is based on the NSGA-III (non-dominated sorting genetic Algorithm III) to find optimal solutions concerning time consumption and energy consumption. A data replica creation scheme and a data replica placement scheme are proposed to preserve the privacy of data replicas. The data replica creation scheme is based on a Level of Privacy (LoP) defined by data-owners and the service capacity of fog nodes. The proposed data replica placement scheme is based on the priority level of fog nodes. Moreover, this thesis conducts comprehensive simulations and systematic experiments to demonstrate and evaluate the effectiveness and efficiency of our proposed schemes compared with the state-of-the-art schemes. The results demonstrate that the proposed scheme can efficiently achieve data privacy in the fog computing paradigm and outperform other schemes in terms of performance efficiency.