.png){kind=icon}
.png){kind=icon}
Return
Task offloading method of edge computing in IoT based on Deep Q Learning Algorithm
.png){kind=icon}
OA
AI
DOI：10.1080/24751839.2025.2578892.png)
.png){kind=icon}
.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To improve the efficiency of task offloading in edge computing of the Internet of Things, a multi-task offloading optimization model combining software definition network and dual depth Q network is proposed. First of all, the edge computing framework of the Internet of Things is readjusted by using the software definition network, and a new edge computing architecture is built. Secondly, the decision layer of dual depth Q network is used to optimize the task offloading strategy to address the overestimation of Q value, and a multi-task offloading model is built by combining the new edge computing architecture and dual depth Q network. The research outcomes denote that the success rate of task offloading in the edge industry IoT security dataset and mobile edge computing dataset of the offloading model is 95.7% and 97.2% respectively, and the bandwidth utilization rate is 98.6% and 99.1% respectively. In practical applications, the average task processing delay of this model in drone collaborative task scenarios is as low as 8.5 ms, and the average task offloading success rate is as high as 98.5%. This shows that the raised model can provide an effective solution for task offloading optimization in edge computing of the Internet of Things.
Keywords:
DDQN
IoT
edge computing
SDN
architecture design
task offloading
AI Summary
Key information extracted from the uploaded paper, including a brief overview, abstract, background, key highlights, visual analysis, and future outlook.
Journal
IF:
1.7
Papers:
68
Citations:
419