Mobile crowd sensing (MCS) takes advantage of pervasive mobile devices that are equipped with multi-sensors to collect rich data of a certain geographic area. Because of the importance of incentivizing users to participate, auction-based open MCS markets have been proposed in past literature. Note that their focus is to achieve critical economic properties but fail to protect bid privacy. Although there are limited schemes dealing with this issue, they are designed only for single-side auctions and are unsuitable for double-side auctions whose properties are quite different. In this paper, inspired by uniform pricing and ex-ponential mechanism, we propose a differentially private double auction (DPDA) scheme for MCS to protect bid privacy for both auction sides. In addition, the traditional economic properties, such as γ-truthfulness, individual rationality and budget balance, are guaranteed as well. Besides, we derive closed forms over the computation complexity and the approximate optimal platform revenue achieved by the scheme. Extensive simulations have been conducted on real-world datasets to validate the efficiency and effectiveness of DPDA.
展开▼