Rapid Migration at the Edge of the Network
Principal Supervisor: Dr Blesson Varghese
Second Supervisor: Dr Hans Vandierendonck
+ Project Description
The vision of edge computing is to facilitate general purpose computing on nodes located at the edge of the network. Edge nodes may take the form of traffic routing nodes (for example, routers, mobile base stations and switches) on which spare computing resources are used for general purpose computing or dedicate nodes (for example, a cluster of low power processors co-located with traffic routing nodes). Edge nodes are anticipated to execute workloads that are offloaded from user devices or the data center to improve the Quality-of-Service of applications that service user requests.
One key challenge in executing offloaded workloads on edge nodes is that these workloads will need to be rapidly migrated from one edge node to another. This is essential for maintaining seamless service as the traffic on edge nodes increase and they do not have spare computing resources or users move away from the coverage of an edge node that was servicing requests. To enable migration in any computing environment the state of execution, dependencies of the workload and associated data will need to be accounted for during relocation. These tasks however need to be done rapidly in edge environments that are typically resource constrained.
The aim of this project is to develop transformative technologies at the system software or application level to facilitate rapid migration. Research literature will be iteratively surveyed during the project to determine the state-of-the-art in migration across networks. Then a test-bed using a collection of low power processors, such as Odroid boards or Raspberry Pis that can host offloaded workloads and a set of relevant applications will be used to investigate requirements for ‘when to migrate’ in edge environments. This will set the basis for determining policies for migration. Then the primary goal of the project ‘how to migrate’ will be pursued by designing, developing and deploying novel mechanisms for rapidly migrating workloads at the edge. The project is flexible – the candidate may choose to focus on mechanisms at the system software and/or application level.
+ How to Apply
Applicants should apply electronically through the Queen’s online application portal at: https://dap.qub.ac.uk/portal/
+ Contact Details
|Supervisor Name:||Dr. Blesson Varghese|
+44 (0)28 9097 5431