Search

Mobile crowdsensing (MCS) has now become an effective paradigm to collect massive data for various sensing applications. However, the interactions between mobile users and the platform, and the data release to third parties, pose severe challenges of privacy leakage for MCS systems, such as the leakage of users' identities and locations. Although several works on MCS have explored the privacy issues in task allocation, incentive, and data reporting, there is still a lack of a comprehensive privacy preserving framework for MCS to protect the privacy of users throughout users' involvement in crowdsensing tasks. In this article, we divide the life cycle of each crowdsensing task in MCS into four phases: task allocation, incentive, data collection, and data publishing, and design a privacy-preserving framework for MCS to protect users' privacy in the whole life cycle of MCS. [ABSTRACT FROM AUTHOR]

Recently radio frequency identification (RFID) is promoting the rapid development of Internet of Things and has been widely applied in numerous industrial applications such as inventory management, object tracking and smart logistics, etc. Tags in some RFID applications can be classified into multiple categories according to the types of objects they are attached to. One of the important functionalities for the multi-category RFID systems is the missing tag cardinality estimation of each category. This paper focuses on solving the missing tag iceberg query problem for multi-category RFID systems, i.e., to determine a set of categories, whose missing tags are more than a threshold, with a required reliability. We firstly propose two basic missing tag i ceberg query schemes called MAC-SZE and MAC-HZE, which adopt the singleton-zero estimator and homogeneous-zero estimator, respectively. To further improve the query efficiency, we then propose two segmented enhanced schemes called MAC-SSZE and MAC-SHZE, which eliminate the unnecessary slots in each frame. We conduct theoretical analysis to guarantee the required reliability is satisfied. The simulations are finally conducted and the results illustrate that the proposed missing tag iceberg query schemes greatly outperform other existing one. [ABSTRACT FROM AUTHOR]