Privacy preserving data aggregation with fault tolerance in fog-enabled smart grids

Abstract Smart grid (SG) seamlessly integrates electrical and communication network elements to provide sustainable and affordable electricity in an efficient and stable manner. Smart meters (SMs) are installed at customer places, providing an efficient way of recording household electricity consumption data and communicating it back to electricity providers using Internet. In smart cities, SG is compute-intensive application, which demands strict latency-aware capabilities, and low delays. Through the usage of fog-enabled Internet of Things (IoT)-Aided SG architectures, localized computing and processing facilities can be deployed at the network edge offering: low latency, reduced communication costs, low congestion, and latency-sensitive analytics. Energy consumption data that the meters collect can disclose sensitive information of an individual. The two most important challenges faced by the future smart grid are security and privacy. Although, privacy-preserving fog-enabled aggregation for SG communications have been studied in recent years. However, there exists limited literature that addresses the requirement of fault tolerance, allowing data aggregation activity to continue effectively and efficiently for functional SMs, even in the presence of faulty meters. Therefore, in this paper, we propose a fog-enabled privacy-preserving secure data aggregation scheme with fault-tolerance. Metering data privacy is achieved through the Boneh-Goh-Nissam (BGN) cryptosystem and authentication of the source is provided through elliptic curve digital signature algorithm (ECDSA). The ECDSA generates signatures in quick manner, due to smaller key sizes and is suitable for constrained resources devices like Advanced Metering Infrastructure (AMI). Detailed security analysis demonstrate that our proposed scheme preserves the data privacy of metering data, ensures source authentication, supports fault-tolerance, and prevents false data injection (FDI) attacks. In addition, extensive performance evaluations also show that our proposed scheme is efficient in terms of encryption, aggregation, decryption and communication costs when compared with the three existing state-of-the-art schemes.