A walrus optimization algorithm for sustainable internet of robotic things based on Q-Learning

The Internet of Robotic Things (IoRT) integrates cloud robotics, artificial intelligence, and the Internet of Things to work collaboratively and is popularly employed in various autonomous systems. However, these devices often struggle to meet real-time applicability due to limited battery, low comp...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:The Journal of supercomputing Ročník 81; číslo 15; s. 1464
Hlavní autori: Varshney, Hirdesh, Singh, Avtar
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: New York Springer US 14.10.2025
Springer Nature B.V
Predmet:
Algorithms
Artificial intelligence
Compilers
Computation offloading
Computer Science
Cost control
Data processing
Edge computing
Energy consumption
Energy efficiency
Energy requirements
Internet of Things
Interpreters
Machine learning
Multiple objective analysis
Neural networks
Optimization
Performance enhancement
Processor Architectures
Programming Languages
Real time
Robot learning
Robotics
Internet of robotic things
Multi-objective optimization
Load balancing
Cloud robotics
Task offloading
Energy optimization
ISSN:1573-0484, 0920-8542, 1573-0484
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:The Internet of Robotic Things (IoRT) integrates cloud robotics, artificial intelligence, and the Internet of Things to work collaboratively and is popularly employed in various autonomous systems. However, these devices often struggle to meet real-time applicability due to limited battery, low computational capability, and high latency, which necessitate high-performance computing and distributed architectures. Therefore, the present work develops an efficient task offloading mechanism by considering a multi-objective optimization approach to reduce energy consumption based on sampling rate, transmission interval, and data processing while achieving the deadline constraints and load balancing. The methodology is executed via fog computing to lower the communication overhead among edge devices and the cloud. Further, the Q -learning approach is integrated with the walrus optimization algorithm to develop Q WaOA. This integration helps improve the balance between exploration–exploitation by incorporating the intelligence of reinforcement learning. The effectiveness of the proposed work is confirmed by simulations, which show that the proposed strategy reduces the energy requirements by at least 27.14% compared to existing methods. The experimental findings validate the proposed schema compared to other existing approaches in enhancing the performance of IoRT devices.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1573-0484
0920-8542
1573-0484
DOI:10.1007/s11227-025-07933-0