Mobile Edge Distributed Intelligence
Lead Research Organisation:
University of Surrey
Department Name: Institute of Communications Systems
Abstract
Mobile Edge Distributed Intelligence research is combining three technology subject areas; Distributed Computing, Mobile Communication Networks and Artificial Intelligence. In contrast to the industry standard of centralised cloud computing, Mobile Edge is a new computing paradigm where compute and storage nodes are located at the edge of the mobile network and therefore, closer to the user. Mobile Edge is often associated with Fog computing, a similar distributed computing paradigm, the main differences being Mobile Edge is exclusively applied to mobile telecom networks whereas Fog is more overarching and can be applied to a variety of networks such as corporate networks, also in contrast to Mobile Edge, Fog is not necessarily at the edge of the network and could be more akin to a distributed cloud within a network. Both Mobile Edge and Fog architectures utilise the compute at the edge as an intermediary between the user and a centralised cloud datacentre, depending on the application user requests can be handled at the edge and then when required passed through to the cloud. Mobile Edge (also known as Multi-access Edge Computing) is being standardized by ETSI and the OpenFog consortium is a group of industry and academic institutions aiming to standardize and promote Fog computing. The ETSI standard discusses how Mobile Edge will generate another revenue stream for mobile operators and enable third-party applications to be hosted at the edge.
New 5G Mobile Networks features combined with Mobile Edge provides new opportunities and applications that would not have been possible in the past. Mobile User Equipment (UEs) are devices that connect to the mobile network, typically these are mobile telephones however with 5G other Internet of Things (IoT) devices from sensors to connected cars could use the mobile network. UEs are constrained by factors such as battery life, processing power, storage, connectivity and data caps, by using Mobile Edge and techniques such as computation offloading, some of these limitations can be alleviated. Mobile Edge with 5G network also enables other features such as ultra-low latency, high availability and reduced data transfer to the cloud. Compared to centralised systems, the distribution aspect of Mobile Edge throws up new challenges especially when considering the scale and geographic nature of a mobile network. Mobile Edge faces the following key challenges; Networking and Quality of Service, Computational Offloading and Distribution, Provisioning and Resource Management, Security and Privacy, Interface and Programming Model and Accounting and Billing.
There are a wide range of applications for Mobile Edge Distributed Intelligence which fall into three categories; Content Delivery Networks (CDN's) improving the speed and reliability users web requests are returned, Cyber-Physical Systems such as Internet of Things (IoT) sensors and connected cars which is mostly computational offloading, data processing and aggregation and low latency connection to near-by devices and Software Based Networking where the network operator aims to optimise both the local and wider network from network data at the edge, faster fault detection and distribute network systems such as authentication. Mobile Edge Distributed Intelligence takes the Mobile Edge concept further and sees how the intelligence can be used, learnt and distributed at the edge. There have been trials of a few Mobile Edge and Fog applications but as of now, there are no examples of a fully operational Mobile Edge systems.
From background reading, this is the first research to investigate how artificial intelligence methods can be integrated, distributed and utilised in Mobile Edge on a 5G Network. The research focuses on how intelligence can be integrated, distributed, learnt and shared efficiently in a distributed mobile edge system such as the aforementioned applications and how local optimisation can lead to global optimisation.
New 5G Mobile Networks features combined with Mobile Edge provides new opportunities and applications that would not have been possible in the past. Mobile User Equipment (UEs) are devices that connect to the mobile network, typically these are mobile telephones however with 5G other Internet of Things (IoT) devices from sensors to connected cars could use the mobile network. UEs are constrained by factors such as battery life, processing power, storage, connectivity and data caps, by using Mobile Edge and techniques such as computation offloading, some of these limitations can be alleviated. Mobile Edge with 5G network also enables other features such as ultra-low latency, high availability and reduced data transfer to the cloud. Compared to centralised systems, the distribution aspect of Mobile Edge throws up new challenges especially when considering the scale and geographic nature of a mobile network. Mobile Edge faces the following key challenges; Networking and Quality of Service, Computational Offloading and Distribution, Provisioning and Resource Management, Security and Privacy, Interface and Programming Model and Accounting and Billing.
There are a wide range of applications for Mobile Edge Distributed Intelligence which fall into three categories; Content Delivery Networks (CDN's) improving the speed and reliability users web requests are returned, Cyber-Physical Systems such as Internet of Things (IoT) sensors and connected cars which is mostly computational offloading, data processing and aggregation and low latency connection to near-by devices and Software Based Networking where the network operator aims to optimise both the local and wider network from network data at the edge, faster fault detection and distribute network systems such as authentication. Mobile Edge Distributed Intelligence takes the Mobile Edge concept further and sees how the intelligence can be used, learnt and distributed at the edge. There have been trials of a few Mobile Edge and Fog applications but as of now, there are no examples of a fully operational Mobile Edge systems.
From background reading, this is the first research to investigate how artificial intelligence methods can be integrated, distributed and utilised in Mobile Edge on a 5G Network. The research focuses on how intelligence can be integrated, distributed, learnt and shared efficiently in a distributed mobile edge system such as the aforementioned applications and how local optimisation can lead to global optimisation.
Organisations
People |
ORCID iD |
Klaus Moessner (Primary Supervisor) | |
Alexander Grace (Student) |
Studentship Projects
Project Reference | Relationship | Related To | Start | End | Student Name |
---|---|---|---|---|---|
EP/N509383/1 | 30/09/2015 | 30/03/2021 | |||
1945816 | Studentship | EP/N509383/1 | 30/09/2017 | 29/09/2020 | Alexander Grace |