Key points are not available for this paper at this time.
Trust management has been widely employed to determine a user's trustworthiness based on evaluations from other entities, and trustless users are those with low trustworthiness due to dishonest or malicious behaviors. To overcome the defects of traditional centralized trust management, decentralized trust management has been proposed, leveraging blockchain to store trust data, e.g., user interaction evaluations, securely. However, blockchain-based decentralized trust management usually suffers from throughput limitation, hindering timely record users' trust data, delaying expose trustless users. As a result, trustless users may still interact with others in the system with the outdated trustworthiness, undermining the reliability and fairness of the system. Furthermore, decentralized pseudonymous networks suffer from a trust cold-start problem due to lacking users' prior interaction history or endorsements from trusted third parties, making it hard for newly joined users to assess the trustworthiness. To address these issues, we propose TUES in this paper, an efficient Trustless User Exposure Scheme. TUES stores trust data in a trust blockchain collectively maintained by all users. To efficiently expose trustless users, we design a dynamic consensus mechanism for TUES. This dynamic consensus mechanism integrates three novel consensus algorithms, efficiently utilizing network throughput to record trust data of trustless users and ensure the consistency of the blockchain. Additionally, TUES includes a multi-signature-based scheme to allocate initial trust values to users, thus resolving the trust cold-start problem in decentralized pseudonymous networks. Analysis and experiments show that TUES improves the efficiency of exposing trustless users while maintaining the consistency of the trust blockchain. It also increases the cost for adversaries conducting Sybil, whitewashing and Byzantine attacks.
Yu et al. (Thu,) studied this question.