Mobile crowd sensing as an enabler for people as a service mobile computing

Mobile Crowd Sensing (MCS) is a new sensing paradigm exploiting the capabilities of smart devices (smartphones, wearables, etc.) to gather large volume of data. Gathering contextual information is a very expensive activity in terms of mobile device resource consumption, so limiting this consumption is essential for user satisfaction. The architectural style applied to the MCS platform largely affects the consumption of these resources. A server-centric MCS is more efficient when there are many entities interested on the gathered information, whilst a mobile-centric architecture has lower consumption when real-time information is required. In this paper, we propose a platform combining both architectural styles. This allows us to reduce the resource consumption of mobile devices, since it is easier to take advantage of the benefits of each style, and to better facilitate user aggregation, being able to group users both at the server and at the client-side depending on the freshness of the required information and the sensing task to be assigned. Finally, we have evaluated this platform for two different case studies, obtaining very promising results.