| Autor: |
Kumar, Ashok, Kumar, Ch. Mohan Sai, Salve, Archana Ravindra, Bhutani, Priyanka, Radhakrishnan, S., Turukmane, Anil V. |
| Předmět: |
|
| Zdroj: |
Library of Progress-Library Science, Information Technology & Computer; Jul-Dec2024, Vol. 44 Issue 3, p14846-14854, 307p |
| Abstrakt: |
To enhance the scheduling of time-sensitive jobs in collaborative edge computing environments for Internet of Things applications. Deep Reinforcement Learning (DRL) will be utilized to achieve this. The rising complexity and scale of Internet of Things (IoT) systems make efficient and effective task scheduling more and more important. This is due to the requirement to minimize latency and resource contention while also meeting strict deadlines. We provide a DRL-based approach that uses dynamic learning and modification to dynamically learn and modify scheduling strategies based on job deadlines, real-time system conditions, and resource availability. Furthermore, we recommend that this approach be implemented. This adaptive solution greatly improves scheduling decisions, ensuring that Internet of Things tasks that are time-sensitive are finished within the allocated time frames. Furthermore, we employ CPLEX optimization approaches to handle resource allocation-related mixed-integer linear programming (MILP) constraints. This strong architecture can assist in resolving the intricate scheduling problems that occur in edge computing. Our simulation findings clearly show that our DRL-based scheduler outperforms the standard scheduling methods. It is a great choice for Internet of Things applications that have time constraints because it cuts down on work latency and improves resource efficiency. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
