Deep Reinforcement Learning for QoS-Aware IoT Service Composition: The PD3QND Approach

Chen, Yi; Cheng, Lianglun; Wang, Tao

doi:10.1109/icsess58500.2023.10293060

Cited by 2 publications

(1 citation statement)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The paper in [15] have proposed an optimization algorithm called PD3QND, which is based on deep reinforcement learning. PD3QND incorporates various techniques, including Deep Q-Network (DQN), noise networks, prioritized experience replay, double dueling architecture, and demonstration learning.…”

Section: Literature Reviewmentioning

confidence: 99%

Toward Optimal Service Composition in the Internet of Things via Cloud-Fog Integration and Improved Artificial Bee Colony Algorithm

Xiao

2024

ijacsa

View full text Add to dashboard Cite

In the quest to delve deeper into the burgeoning realm of the service-oriented Internet of Things (IoT), the pressing challenge of smoothly integrating functionalities within smart objects emerges prominently. IoT devices, notorious for their resource constraints, often lean heavily on cloud infrastructures to function effectively. However, the emergence of fog computing offers a promising alternative, allowing the processing of IoT applications closer to the sensors and thereby slashing delays. This research develops a novel method for IoT service composition that leverages both fog and cloud computing, utilizing an enhanced version of the Artificial Bee Colony (ABC) algorithm to refine its convergence rate. The approach introduces a Dynamic Reduction (DR) mechanism designed to perturb dimensions innovatively. Traditionally, the ABC algorithm generates new solutions that closely mimic their parent solutions, which unfortunately slows down convergence. By initiating the process with significant dimension disparities among solutions and gradually reducing these disparities over successive iterations, this method strikes an optimal balance between exploration and exploitation through dynamic adjustment of dimension perturbation counts. Comparative analyses against contemporary methodologies reveal significant improvements: a 17% decrease in average energy consumption, a 10% boost in availability, an 8% enhancement in reliability, and a remarkable 23% reduction in average cost. Combining the strengths of fog and cloud computing with the refined ABC algorithm through the Dynamic Reduction mechanism significantly advances the efficiency and effectiveness of IoT service compositions.

show abstract

Section: Literature Reviewmentioning

confidence: 99%

Toward Optimal Service Composition in the Internet of Things via Cloud-Fog Integration and Improved Artificial Bee Colony Algorithm

Xiao

2024

ijacsa

View full text Add to dashboard Cite

show abstract

Leveraging Deep Learning-Based Approach for IoT Service Composition Through Local Service Selection

Awanyo,

Guermouche

2024

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Deep Reinforcement Learning for QoS-Aware IoT Service Composition: The PD3QND Approach

Cited by 2 publications

References 8 publications

Toward Optimal Service Composition in the Internet of Things via Cloud-Fog Integration and Improved Artificial Bee Colony Algorithm

Toward Optimal Service Composition in the Internet of Things via Cloud-Fog Integration and Improved Artificial Bee Colony Algorithm

Leveraging Deep Learning-Based Approach for IoT Service Composition Through Local Service Selection

Contact Info

Product

Resources

About