Time-Triggered Task Offloading Scheduling in TSN-Based Edge Computing Power Networks
Computing power networks (CPN), connecting distributed computing resources over networks, have been proposed as potential next-generation novel network systems to enable optimal resource allocation among multiple computing nodes. Computing power scheduling is crucial in addressing the conflict betwe...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11000111/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Computing power networks (CPN), connecting distributed computing resources over networks, have been proposed as potential next-generation novel network systems to enable optimal resource allocation among multiple computing nodes. Computing power scheduling is crucial in addressing the conflict between computing power supply and demand, where the deterministic scheduling problem within the CPN needs to be urgently addressed. To break through the performance limitations of standalone computing power and enable time-sensitive services, the integration of edge computing with time-sensitive networking (TSN) emerges as a promising technology, thereby we propose a joint communication-offloading and scheduling (JCOS) scheme that enables computing power sharing to guarantee task determinacy. The abstract modeling of computing power node measure and computing power demand is established to enable the unified quantification of resources and services. Subsequently, we establish the correlation between tasks using Directed Acyclic Graph (DAG) and design a matching and load balancing-based task offloading algorithm (MLB-TOA). Moreover, we propose an efficient routing algorithm for task generation streams (TGS-RA) and an incremental task generation stream injection time planning algorithm (TGS-ITP) in time-sensitive networking. Experiments demonstrate that the proposed JCOS scheme significantly improves scheduling success rate and achieves low end-to-end latency compared to existing offloading approaches. Additionally, the proposed JCOS scheme reduces time overhead and achieves high scheduling success rate compared to existing scheduling approaches. The proposed JCOS scheme ensures the deterministic completion of diverse tasks and provides precise timing control. |
|---|---|
| ISSN: | 2169-3536 |