AHSWN Home · Issue Contents · Forthcoming Papers

Dynamic Scheduling Strategy for Cloud Computing Services Supported by 6G Network Slicing in Ad Hoc Networks
Bingxin Si

In the realm of cloud computing (CC) services, task scheduling (TS) and resource allocation (RA) are key components that have a considerable impact on system performance, particularly with the emergence of next-generation networks like 6G. While current approaches to scheduling and resource management in cloud environments rely mainly on conventional optimization techniques, they frequently fail to meet the dynamic and complicated needs of 6G networks, such as ultra-low delay, high flexibility, and network slicing. The research presents a new dynamic scheduling technique for CC services that uses 6G network slicing in ad hoc networks, which are highly dynamic and decentralized environments. The goal is to improve TS and RA using Intelligent Coyote Optimised Deep Reinforcement Learning (ICO-DRL) techniques to address the constraints of existing systems. In particular, the research employs a hybrid approach for RA and TS in a 6G-enabled cloud environment and ad hoc network scenarios that combines deep reinforcement learning (DRL) with an Intelligent Coyote Optimisation (ICO) algorithm. The DRL component offers flexibility in response to changing network circumstances, while the ICO offers effective RA. The findings demonstrate that the suggested approach outperforms conventional techniques, resulting in reduced latency, optimal cloud resource utilization, and increased efficiency in TS and RA. Task scheduling efficiency is 90.6% and resource allocation efficiency is 92.46%. The proposed model offers real-time, adaptive decision-making, which greatly enhances the quality of service in cloud computing systems based on 6G. The ICO-DRL algorithm improves task scheduling and resource allocation by employing real-time feedback, rather than the “static” approaches taken by traditional optimization models. Therefore, ICO-DRL is more amenable to deployment in a cloud context and ad hoc network environments, where applications, processing, and scalability need to occur instantaneously. Finally, the findings showed that combining sophisticated optimization approaches with deep learning (DL) can help handle the issues of TS and RA in 6G-enabled CC services.

Keywords: cloud computing (CC), task scheduling (TS), Intelligent Coyote Optimised Deep Reinforcement Learning (ICO-DRL), resource allocation (RA), 6G network slicing, Ad hoc networks.