A Context Aware Framework for Mobile Crowd-Sensing

Abstract

Context awareness plays ever increasing role in Mobile Crowd-Sensing (MCS), which relies on sensing capabilities of mobile devices to collect real-time user data and related context. The paper proposes a MCS framework for valuable data collection in order to enable smart applications. The paper also addresses a key challenge in MCS on how to reduce energy consumption in order to encourage user participation. The paper argues that to optimize task allocation costs, it is important for a given query to select the most appropriate participants according to the context of the device, the participant, and the sensing task. Context awareness can significantly reduce the sensing and communication costs. Yet to incorporate context awareness into MCS, there is a need for a standard and overarching context model. This paper proposes a multi-dimensional context model to capture related contextual information in the MCS domain, and incorporate it into a context-aware MCS framework to improve energy efficiency and support task allocation. The paper concludes with discussing implementation and evaluation of the proposed approach.