A Fog-Assisted Privacy-Preserving Task Allocation in Crowdsourcing

Abstract

As a people-centric sensing paradigm, crowdsourcing has attracted considerable attention since it can solve a complicated task by gathering the wisdom of a crowd of workers. To increase the efficiency of task allocating and recruit suitable workers, the crowdsourcing server needs to obtain information related to task content and worker profiles, which poses a threat to privacy leaks of both tasks and workers. Although several privacy-preserving crowdsourcing mechanisms with simultaneously achieving task privacy and worker privacy have been proposed, there are two problems for these schemes, either the ability of workers is not taken into account or they are computationally intensive. To address the above problems simultaneously, fog-assisted privacy-preserving task allocation in crowdsourcing is proposed by means of the advantages of fog computing, which can not only achieve privacy protection of both the task and the worker but also alleviate the workers’ computational burden by offloading partial computation to the fog node. By applying threshold secret sharing technology, the proposed scheme enables that only workers satisfying task requirements can decrypt the task content, which achieves the verification of the workers’ ability and resists the attacks of the greedy workers. Then, rigorous security proofs about privacy properties are given in the proposed scheme. Finally, the proposed scheme is estimated through theoretical analysis and experiments. The experimental results show that compared with the current state-of-the-art scheme, the proposed scheme has more advantages in terms of computational cost and storage overhead. Especially for a worker, it only requires one pairing and two multiplication operations in the decrypting phase.