Abstract :
[en] This thesis proposes energy-efficient mobile crowdsensing (MCS) solutions for smart cities. Specifically, it focuses on sensing and communications processes in distributed computing paradigms and complex urban dynamics in city-wide scenarios. MCS is a data collection paradigm that has gained significant attention in recent years and has become appealing for urban sensing. MCS systems rely on contributions from mobile devices of a large number of participants or a crowd. Smartphones, tablets, and wearable devices are deployed widely and already equipped with a rich set of sensors, making them an excellent source of information. Mobility and intelligence of humans guarantee higher coverage and better context awareness if compared to traditional sensor networks. At the same time, individuals may be reluctant to share data for devices’ battery drain and privacy concerns. For this reason, MCS frameworks are specifically designed to include incentive mechanisms and address privacy concerns. Despite the growing interest in the research community, MCS solutions still need a more in-depth investigation and categorization on many aspects that span from sensing and communication to system management and data storage. This Ph.D. thesis focuses not only on sustainable MCS solutions to challenging problems in urban environments but also on a comprehensive study aiming to clarify concepts, aspects, and inconsistencies in existing literature from a global perspective. Specifically, this manuscript proposes the following contributions:
• Present the MCS paradigm as a four-layered architecture divided into application, data, communication, and sensing layers, proposing novel taxonomies related to each layer. The detailed taxonomy aims to shed light on the current landscape, covering all MCS aspects and allowing
for a simple and clear classification of applications, methodologies, and architectures.
• A significant improvement of the previously developed simulation environment CrowdSenSim by implementing a set of novel features. The novelties include easy-to-use city-wide street networks, more realistic pedestrian mobility models, and real battery drain measurements over several other features.
• An analysis of energy efficiency that poses the basis for sustainable MCS data collection frameworks (DCFs). It includes both a theoretical methodology to assess different DCFs and real energy measurements conducted in a laboratory, simulated in large scale urban environments.
• A study that exploits crowdsensed data for a learning-driven estimation of local businesses’ attractiveness in cities to show how MCS systems can support urban planning.
• A novel efficient edge data centers deployment in real urban environments based on human mobility and traffic generated from mobile devices. The citizens’ mobility is developed by feeding CrowdSenSim with crowdsensed data.
