Follow Me, If You Can: A Framework for Seamless Migration in Mobile Edge Cloud
Research output: Contribution to book/Conference proceedings/Anthology/Report › Conference contribution › Contributed › peer-review
Contributors
Abstract
Mobile Edge Cloud (MEC) enables cloud services in the proximity of user equipments. MEC will be greatly beneficial to various latency-sensitive applications such as autonomous cars and augmented reality/virtual reality applications. The challenge is to maintain an uninterrupted service in the presence of user's mobility. In maintaining a consistent state of MEC application, we advocate the decoupling of the state storage component from the computing one. We leverage a key-value storage layer to synchronize states between MEC servers. Subsequently, we propose a distributed key-value store framework, which decouples MEC application design into processing and state, to ensure service continuity. Evaluation results show that our solution reduces downtime by half in most of the cases, even under high load of state update. Furthermore, under moderate load of state updates, the framework can eliminate downtime completely.
Details
Original language | English |
---|---|
Title of host publication | IEEE INFOCOM 2020 - IEEE Conference on Computer Communications Workshops (INFOCOM WKSHPS) |
Publisher | IEEE Communications Society, New York |
Pages | 1178-1183 |
Number of pages | 6 |
ISBN (print) | 978-1-7281-8696-2 |
Publication status | Published - 9 Jul 2020 |
Peer-reviewed | Yes |
Conference
Title | IEEE INFOCOM 2020 - IEEE Conference on Computer Communications Workshops |
---|---|
Abbreviated title | IEEE INFOCOM WKSHPS 2020 |
Duration | 6 - 9 July 2020 |
Location | online |
City | Toronto |
Country | Canada |
External IDs
Ieee | 10.1109/INFOCOMWKSHPS50562.2020.9162992 |
---|---|
Scopus | 85091477146 |
ORCID | /0000-0001-7008-1537/work/142248634 |
ORCID | /0000-0001-8469-9573/work/161891005 |
Keywords
Keywords
- Servers, Cloud computing, Indexes, Process control, Synchronization, Operating systems, Data structures