Final published version
Research output: Contribution to Journal/Magazine › Journal article › peer-review
Research output: Contribution to Journal/Magazine › Journal article › peer-review
}
TY - JOUR
T1 - Effective truth discovery and fair reward distribution for mobile crowdsensing
AU - Shi, Fengrui
AU - Qin, Zhijin
AU - Wu, Di
AU - McCann, Julie A.
PY - 2018/12
Y1 - 2018/12
N2 - By leveraging the sensing capabilities of consumer mobile devices, mobile crowdsensing (MCS) systems enable a number of new applications for Internet of Things (IoT), such as traffic management, environmental monitoring, and localisation. However, the sensing data collected from the crowd workers are of various qualities, making it difficult to discover the ground truth and maintain the fairness of incentivisation schemes. In this paper, we propose a truth discovery algorithm based on a two-stage Maximum Likelihood Estimator (MLE), which explicitly characterises the heterogeneous sensing capabilities of the crowd and is able to estimate ground truth accurately using only a small amount of data from IoT infrastructures. Moreover, based on the truth discovery algorithm, two reward distribution schemes, LRDS and MRDS, are proposed to ensure fairness of rewarding the crowd according to their effort levels. We evaluate the estimation accuracy of the truth discovery algorithm and the fairness of the reward distribution schemes using both simulations and real-world MCS campaigns. The evaluation results indicate that the proposed methods achieve superior performance compared with state-of-the-art methods in terms of estimation accuracy and fairness of reward distribution. © 2018 Elsevier B.V.
AB - By leveraging the sensing capabilities of consumer mobile devices, mobile crowdsensing (MCS) systems enable a number of new applications for Internet of Things (IoT), such as traffic management, environmental monitoring, and localisation. However, the sensing data collected from the crowd workers are of various qualities, making it difficult to discover the ground truth and maintain the fairness of incentivisation schemes. In this paper, we propose a truth discovery algorithm based on a two-stage Maximum Likelihood Estimator (MLE), which explicitly characterises the heterogeneous sensing capabilities of the crowd and is able to estimate ground truth accurately using only a small amount of data from IoT infrastructures. Moreover, based on the truth discovery algorithm, two reward distribution schemes, LRDS and MRDS, are proposed to ensure fairness of rewarding the crowd according to their effort levels. We evaluate the estimation accuracy of the truth discovery algorithm and the fairness of the reward distribution schemes using both simulations and real-world MCS campaigns. The evaluation results indicate that the proposed methods achieve superior performance compared with state-of-the-art methods in terms of estimation accuracy and fairness of reward distribution. © 2018 Elsevier B.V.
KW - Incentivisation
KW - Maximum Likelihood Estimator
KW - Mobile crowdsensing
KW - Truth discovery
KW - Environmental management
KW - Internet of things
KW - Crowd sensing
KW - Environmental Monitoring
KW - Internet of Things (IOT)
KW - Maximum likelihood estimator
KW - Maximum likelihood estimators (MLE)
KW - State-of-the-art methods
KW - Maximum likelihood estimation
U2 - 10.1016/j.pmcj.2018.09.007
DO - 10.1016/j.pmcj.2018.09.007
M3 - Journal article
VL - 51
SP - 88
EP - 103
JO - Pervasive and Mobile Computing
JF - Pervasive and Mobile Computing
SN - 1574-1192
ER -