.png){kind=icon}
.png){kind=icon}
Return
SecDS: A security-aware DAG task scheduling strategy for edge computing
.png){kind=icon}
PRE
AI
DOI：10.1016/j.future.2024.107627.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Edge computing typically handles applications as multiple dependent sub-tasks presented with Directed Acyclic Graphs (DAGs) and offloads them to multiple edge servers to reduce network bandwidth pressure thereby improving task execution efficiency. However, security risk maybe increased due to the involvement of many edge servers in a DAG-based task execution. For instance, execution may fail if any involved edge server is attacked. Nonetheless, security management mechanisms have not been adequately explored. To address the issue, this paper proposes a security-aware DAG task scheduling strategy, termed SecDS, which improves task execution efficiency while ensuring task execution security. The core idea is to adaptively select from various security algorithms for task offloading in light of security estimation. Specifically, a security- trust model is first proposed to define the security risk level between user devices and edge servers. Then, a novel security mechanism is designed to indicate a suitable security algorithm for the task offloading at each security risk level. Furthermore, to offset the cost of the security mechanism, a novel task scheduler is proposed to minimize completion time while satisfying security requirements. Finally, extensive experiments with heterogeneous DAG task flows (generated by an open-source generator, DAGGEN) have been conducted to demonstrate the effectiveness of SecDS.
Keywords:
DAG task
Edge computing
Security risk
Task scheduling
Journal
F
IF:
6.1
Papers:
6.8K
Citations:
2.3W