Computing at the Edge of the Internet

Lead Research Organisation: University College London
Department Name: Electronic and Electrical Engineering

Abstract

Cloud computing (or computing within distant data centres) has contributed much to the development of the Internet over the past twenty years or so. However, new applications in the area of IoT, but not only, are pushing the limits of cloud computing to the point where computing at far-away data-centres cannot cope with application requirements. Emerging trends in computing, communications and networking move towards mobile edge- or fog-computing. According to these trends computation moves to the edges of the network (hence from cloud to fog) to serve newly emerging applications with strict response deadlines. The challenges pertaining to this move defy some of the most fundamental operations taking place in the Internet infrastructure today.
We will investigate the changes that the Internet needs to undergo from the ground up. Starting from network layer protocols, we will explore issues related to routing and resolution of services to micro-data centres, or computation spots. Resource allocation in terms of which services run where through learning algorithms or market-driven approaches will need to be integrated into routing and transport layer protocols. Last, but not least, privacy and security issues and business models to accommodate such a move will be studied as an overarching framework. Distributed Ledger Technologies (DLTs) will be investigated to deal with transactions through trust-less IoT and edge-computing nodes.
Relevance to EPSRC Thematic Areas: ICT, sub-themes: ICT Networks and Distributed Systems, Pervasive and Ubiquitous Computing, Architectures and Operating Systems

Publications

10 25 50

Studentship Projects

Project Reference Relationship Related To Start End Student Name
EP/N509577/1 01/10/2016 24/03/2022
1959129 Studentship EP/N509577/1 01/11/2017 31/01/2022 Adrian-Christian Nicolaescu
 
Description The start towards an Edge storage system was made. The first workshop paper shows promising results and potential for collaborations, considering the positive reactions and feedback that was given afterwards. Steps are currently being taken, towards developing the storage system and its management, to improve performance and conceptual design further, while discovering new paradigms.
Exploitation Route A management system will be developed, for storage placement. This shall be used for the improvement of both system performance and user and application/content provider Quality of Experience (QoE) and Quality of Service (QoS). Further, the collaborations proposed are various. However, these collaborations are mostly related to further Edge computing and storage environments, and management intelligence and performance development. These approaches will, most probably, be the way to go forward, once the next paper (on computational resource management and content placement in storage) is published.

Current outcomes, like the simulator and the previous demonstration of function execution, using ICN, could also be further used in the development of other ICN- and Edge-storage-based solutions.
Sectors Digital/Communication/Information Technologies (including Software),Electronics