TR-MCN: light weight task recommendation for mobile crowdsourcing networks

Abstract

To provide privacy protection, task recommendation protocols for mobile crowdsourcing networks typically encrypt tasks before publishing them to the service provider. However, current task recommendation protocols are mainly focusing the privacy of user data and lacking the protection for users’ real identities, resulting in a lot of security issues. Moreover, current privacy-preserving protocols for mobile crowdsourcing networks are typically built on bilinear pairing, leading to high computation costs. To address the above issues, we propose a novel task recommendation protocol with privacy-preserving called TR-MCN. Similar to protocols of this field, TR-MCN can provide privacy-preserving features for mobile crowdsourcing networks. However, different from other well-known approaches, TR-MCN uses pseudonyms instead of real identities, which can provide privacy protection for users’ real identities. Moreover, to simplify the management of pseudonyms and reduce the computation cost of bilinear pairing, we introduce the Bloom filter technique to TR-MCN and design a novel signcryption algorithm, which is much more efficient than current protocols. By doing so, TR-MCN can achieve high efficiency while still satisfying required security requirements. Experimential results show that TR-MCN is feasible for real world applications.