Efficient sharing of privacy-preserving sensing data on consortium blockchain via group key agreement

Abstract

Due to the self-organizing and distributed nature of Wireless Sensor Networks (WSN), valuable data in sensor networks faces security and privacy risks. A blockchain-based approach enables secure and convenient sharing of sensor information among different users. Compared to public and private blockchains, consortium blockchain is widely used across different industries and use cases in WSN due to its auditability and high transaction rate. However, sensing data sharing via consortium blockchain raises the privacy issue. Therefore, the data from the sensor node is encrypted by the key (named sensorkey) shared by the sensor node and the sink node and then sent to the blockchain network to not reveal the privacy of the sensing data. Since the infrastructure in large-scale WSN is usually owned and managed by multiple organizations, encrypted sensing data needs to be authorized by these multiple organizations for computation. Organizations requesting privacy-preserving data are referred to as data sharers. Distributing the sensorkey to each data sharer requires separate encryption of the key using the data sharer’s public key. The sink node needs to be online when each data sharer asks for the sensorkey, and one encryption of the sensorkey for each data sharer consumes precious resources. This work proposes GSChain for efficient privacy-preserving sensing data sharing on consortium blockchain. Multiple data sharers resort to asymmetric group key agreement protocol to maintain a shared group encryption key and their respective group decryption keys, enabling efficient sensorkey retrieval from the consortium blockchain. The sensorkey is encrypted only once by the group encryption key and stored on the consortium blockchain along with the privacy-preserving sensing data. Our scheme improves the efficiency of privacy-preserving data sharing among multiple data sharers while reducing the online demand for sink nodes. Although the data-sharing group should remain stable for a long time, we design the group key update scheme. We also discuss how old and new data sharers access different ranges of privacy-preserving sensing data as the data-sharing group changes. We build a complete implementation of GSChain based on the Hyperledger Fabric framework and conduct a comprehensive set of experimental studies. Our experimental results demonstrate that GSChain improves the privacy-preserving sensing data sharing efficiency with tolerable time and storage overhead. In addition, the time overhead caused by the recovery of the GSChain system is tolerable when the membership of the data-sharing group changes.