Resource management of fog computing in future networks

Publication Type:
Thesis
Issue Date:
2019
Full metadata record
By enabling task offloading among edge network nodes (such as base stations, switches, and routers), fog computing is able to process the computationally demanding tasks at the point of capture. It has the potential to provide low-latency services, increase network capacity, and relieve network congestions. In this thesis, we focus on three exemplary scenarios of fog computing, including (1) single-cell multiuser fog computing to jointly optimize task offloading and resource allocation for different applications with heterogeneous quality of service requirements; (2) fog computing among selfish devices to design incentive mechanism and efficient management of task offloading, processing, and result retrieving; and (3) fog computing across large-scale edge cloud for scalable and distributed resource management in the presence of a large number of geo-distributed edge servers. We present five new approaches to address the challenges for efficient and scalable fog computing in the three scenarios. The first three approaches are for the first scenario, i.e., single-cell multiuser fog computing, for three different types of applications, including delay-tolerant, delay-sensitive, and data-partition tasks. The fourth approach is for the second scenario, where distributed tit-for-tat mechanism is proposed to incentivize the cooperation of selfish devices. The fifth approach is for the last scenario, where collaborative regions are created for preventing tasks being offloaded beyond the vicinity of the point of capture in large-scale networks.
Please use this identifier to cite or link to this item: