Quality-Driven Auction-Based Incentive Mechanism for Mobile Crowd Sensing
Authors:
- 17 October 2014
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 64 (9) , 4203-4214
Abstract
The recent paradigm of mobile crowd sensing (MCS) enables a broad range of mobile applications. A critical challenge for the paradigm is to incentivize phone users to be workers providing sensing services. While some theoretical incentive mechanisms for general-purpose crowdsourcing have been proposed, it is still an open issue as to how to incorporate the theoretical framework into the practical MCS system. In this paper, we propose an incentive mechanism based on a quality-driven auction (QDA). The mechanism is specifically for the MCS system, where the worker is paid off based on the quality of sensed data instead of working time, as adopted in the literature. We theoretically prove that the mechanism is truthful, individual rational, platform profitable, and social-welfare optimal. Moreover, we incorporate our incentive mechanism into a Wi-Fi fingerprint-based indoor localization system to incentivize the MCS-based fingerprint collection. We present a probabilistic model to evaluate the reliability of the submitted data, which resolves the issue that the ground truth for the data reliability is unavailable. We realize and deploy an indoor localization system to evaluate our proposed incentive mechanism and present extensive experimental results.Keywords
Funding Information
- National Natural Science Foundation of China (NSFC 61202373)
- Shanghai Basic Research Key Project (13510711300, 12JC1405200, 11JC1405100)
- Open Foundation of State key Laboratory of Networking and Switching Technology (Beijing University of Posts and Telecommunications) (SKLNST-2013-1-16)
- National Science Foundation (CNS-1320736)
This publication has 20 references indexed in Scilit:
- Communicating Is Crowdsourcing: Wi-Fi Indoor Localization with CSI-Based Speed EstimationJournal of Computer Science and Technology, 2014
- Toward optimal allocation of location dependent tasks in crowdsensing2014
- Efficient crowdsourcing for multi-class labeling2013
- Push the limit of WiFi based localization for smartphones2012
- Centaur2012
- Zee2012
- SpinLoc2012
- Crowds in two seconds2011
- MAP: Multiauctioneer Progressive Auction for Dynamic Spectrum AccessIEEE Transactions on Mobile Computing, 2010
- Revenue failures and collusion in combinatorial auctions and exchanges with vcg payments2004