Invisible, Yet Trusted: Blockchain-Empowered Privacy-Preserving Spatio-Temporal Task Matching in Crowdsourcing With Efficient Public Verification

Abstract

Task matching is a core component of crowdsourcing systems, enabling efficient assignment of tasks to appropriate workers. However, this process often requires both task requesters and workers to disclose sensitive contextual information, such as spatial locations and temporal availability, raising serious privacy concerns. Although various cryptographic techniques and blockchain-based solutions have been proposed to preserve privacy and enhance trust, existing schemes still face three critical limitations: reliance on centralized key authorities, lack of support for fine-grained spatio-temporal constraints, and inefficient public verification mechanisms. To address these challenges, we propose PVSTMatch, a blockchain-empowered privacy-preserving spatio-temporal task matching scheme in crowdsourcing. PVSTMatch eliminates the need for trusted third parties through an authority-free authorization mechanism, supports secure spatio-temporal matching by designing a compact secure comparison method, and enables efficient public verifiability via a succinct verification mechanism. Furthermore, we introduce GE-PVSTMatch, a gas-efficient variant that significantly reduces on-chain storage costs. Security analysis and performance evaluation demonstrate that our schemes achieve strong bilateral privacy, verification correctness, and practical efficiency, making them well-suited for real-world decentralized crowdsourcing platforms.