An Efficient Blockchain-Based Bidirectional Friends Matching Scheme in Social Networks

Abstract

In social networks, the personal attributes or hobbies of the users are exposed to the server to establish the relationships. Service providers may store these information for commercial purpose or statistical analysis. Furthermore, the server may expose to external attacks, which may disclose users' privacy information. In this paper, we present a hierarchical blockchain-based attribute matching scheme, which realizes privacy-preserving attribute matching under multiple semi-trusted servers. The scheme employs CP-ABE and bloom filter to satisfy the requirements of the users to make friend discovery, and reduces the computation cost of users by outsourcing decryption of CP-ABE. Besides, the hierarchical blockchain only implements the consensus and storage of matching results on the blockchain, while the complex calculations and a large amount of data storage are off-chain, which reduces the consumption of the blockchain and improves the operation efficiency. Finally, we prove the scheme can resist single point failure, collusion attack, internal attack and external attack, the experimental results demonstrate the proposed scheme is feasibility and efficiency.