Bi-Tier Differential Privacy for Precise Auction-Based People-Centric IoT Service

With the fast proliferation of device sensing and computing, crowed sensing has become the building block of the Internet of things. Consequently, various data collection and incentive mechanisms are investigated for people-centric services. In this paper, we have investigated the problem of privacy-aware people-centric IoT service based on a tailored auction approach. We applied a bi-tier differential privacy methodology on the data collected from crowdsensing IoT devices. A corresponding pricing scheme is also proposed to ensure the property of incentive compatibility, precise service data, and anonymized query results. Comparing to traditional privacy-aware auction schemes which only focus on the cost, our corresponding precise privacy-aware auction scheme provides a tailored IoT service based on the customers' request. The proposed trial query technique is able to provide a precise assessment of service quality, thus improves the efficiency of the people-centric IoT service. The customer could enjoy the convenience of service evaluation before making a bid, while the actual service data is anonymized to guarantee the service providers' interests. We evaluate the proposed bi-tier differential privacy schema for auction-based service by conducting extensive simulations. The experimental results show that our proposed method yields higher data utility and accuracy for the IoT service customers with privacy concerns.