Data Quality Guarantee for Credible Caching Device Selection in Mobile Crowdsensing Systems


Mobile crowdsensing systems (MCSs) present a flexible and economical alternative to traditional infrastructure-based large-scale sensing through the recruitment of personal mobile devices as data sources. As this becomes a popular sensing approach, it will impact the capacity of typical centralized cellular communication infrastructures widely adopted by MCS applications and any costs accrued. Following the trend toward edge processing, mobile edge caching offloads data and services from the system core to reduce service latency and bandwidth occupation. However, in the MCS case the edge devices are owned by the general public and are therefore more vulnerable to data or calculation manipulation by the user. We now better understand sensor data and user trustworthiness but have no way to determine which devices can also be trusted (i.e., act as a credible caching device). In this article, we treat the quality of sensing data reported by each user as an indication of their possibility of providing credible caching services. Specifically, we conduct a comprehensive study of the data quality problem with regard to cache-enabled MCSs, and develop an incentivization method to encourage users to actively provide high-quality data. That is, quality-aware behavior evaluation is core to the credible caching device selection process. Results of extensive simulations based on realworld data verify the effectiveness of our design. We also highlight several promising research directions that remain open for further elaboration.

Citation information

Cong Zhao, Shusen Yang, Ping Yan, Qing Yang, Xinyu Yang, Julie McCann, ‘Data Quality Guarantee for Credible Caching Device Selection in Mobile Crowdsensing Systems’, in IEEE Wireless Communications, 25(3): 58-64, June 2018.

Turing affiliated authors