Multi-hop collaborative edge computing involving constrained IoT devices at the far edge
As network intelligence is moved to the edge, the cloud continuum is expanded and new challenges arise as network nodes become more heterogeneous and many times constrained in memory and/or battery. In this scenario, collaborative edge computing mechanisms are essential, particularly to implement so...
Saved in:
| Published in: | International Conference on the Network of the Future (Online) pp. 22 - 24 |
|---|---|
| Main Authors: | , , , , |
| Format: | Conference Proceeding |
| Language: | English |
| Published: |
IEEE
02-10-2024
|
| Subjects: | |
| Online Access: | Get full text |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | As network intelligence is moved to the edge, the cloud continuum is expanded and new challenges arise as network nodes become more heterogeneous and many times constrained in memory and/or battery. In this scenario, collaborative edge computing mechanisms are essential, particularly to implement some of the use cases foreseen by the future sixth generation of cellular networks (6G). Furthermore enabling multi-hop cooperation, which does not rely on a central gateway or offloading node, might enhance even more these use cases. In this demonstration, we implement and showcase a multi-hop collaborative edge computing protocol, capable of discovering neighbor nodes and efficiently distributing a computational load among them. The protocol is tailored for Low-power and Lossy Network (LLN) environments, requiring a small amount of control messages, no need of a routing table, and fast convergence time (scale of seconds). We prove this behavior with a testbed composed of four hardware LLN boards, the Contiki-NG operating system, and a graphical user interface to illustrate the execution of the protocol via a post-mortem log analysis. |
|---|---|
| ISSN: | 2833-0072 |
| DOI: | 10.1109/NoF62948.2024.10741440 |