Bilateral Secure and Decentralized Crowdsourcing Task Matching Atop Consortium Blockchain

Task matching is an indispensable crowdsourcing procedure but encounters severe security threats against user privacy and matching integrity. Existing privacy-aware task-matching solutions are mostly specific to the centralized crowd-sourcing architecture with an honest- but-curious platform. De-spite a few attempts made to construct decentralized crowd-sourcing task-matching schemes, they are still constrained to the unitary matching model, which further makes the privacy concerns not comprehensive. In this paper, we present a novel secure and decentralized crowdsourcing task-matching resolution built atop consortium blockchain, in which the flexible matching and privacy demands of both task requesters and workers are taken into account. Particularly, we design a decentralized ciphertext-policy attribute-based keyword search scheme with hidden policy, simultaneously supporting fine-grained worker selection and worker-side keyword search of interested tasks with bilateral privacy. The matching procedure is conducted on a consortium blockchain with guaranteed matching integrity. Our scheme achieves the desired security goals in the random oracle model. Extensive experiments implemented over Hyperledger Fabric demonstrate better off-chain and on-chain performance.