Jun 2026· Al-Farooq Journal of Sciences· 0 citations· 14 references
TL;DR
A new dynamic energy-efficient Deep Learning-Based Predictive Resource Allocation Framework (DLPRAF) is introduced for timely allocation of resources while upholding SLA adherence and is able to achieve 98.6% SLA compliance for this framework while providing cloud infrastructure with meaningful sustainability benefits.
Abstract
Energy efficiency and service provisioning are the two major challenges in current days cloud computing paradigm. In this article, a new dynamic energy-efficient Deep Learning-Based Predictive Resource Allocation Framework (DLPRAF) is introduced for timely allocation of resources while upholding SLA adherence. This framework incorporates several deep learning architectures—namely Long Short-Term Memory (LSTM), Gated Recurrent Unit (GRU) and Convolutional Neural Networks (CNN)—to accurately forecast cloud workload trends. We incorporate temporal and spatial feature extraction ability to capture complex nonlinear dependencies in cloud workloads. By allowing for proactive resource provisioning instead of reactive approaches, the recommended system in better use of resources, lower energy consumption and improved QoS. We experimentally evaluate the effectiveness of DLPRAF and show on real-world cloud datasets (Google Cluster Data and Alibaba traces), that DLPRAF are 32.5% more resource utilization efficient, 43.3% timely and incur 26.6% lower operational costs than threshold-based approaches2. We are able to achieve 98.6% SLA compliance for our framework while providing cloud infrastructure with meaningful sustainability benefits.
Task scheduling plays a crucial role in optimizing performance, reducing costs, and enhancing system reliability by efficiently allocating resources to workloads. Traditional task scheduling methods lack the ability to efficiently manage workloads and resource distribution, leading to potential inefficiencies in performance and energy consumption. To address these limitations, advanced techniques leveraging deep learning and reinforcement learning are explored. This study proposes a deep learning-based model for task scheduling in cloud computing. The model employs a convolutional neural network (CNN) to predict the optimal machines for task allocation. Additionally, Q-learning is integrated with CNN to facilitate load shifting between machines, ensuring efficient utilization of resources. The dataset used in this work consists of task attributes, such as execution time, resource requirements, which were loaded from a CSV file. Comparative analysis with existing models shows that the proposed approach achieves approximately 94% accuracy and consumes less energy than other models, demonstrating its effectiveness in cloud task scheduling.
Kavita Rani, O. Sangwan, R. Garg· IAES International Journal o...· 0 citations
The rapid growth of IoT devices demands efficient resource allocation to satisfy low-latency and high-performance requirements in fog computing. However, most existing approaches predominantly rely on CPU-based resource scheduling, failing to leverage GPU resource potential for heterogeneous workloads. This paper introduces ARAMS (Adaptive Resource Allocation with Machine Learning and Shortest Job First), a novel hybrid framework that integrates machine learning–based CPU/GPU resource prediction with a greedy SJF scheduler to achieve real-time, adaptive task distribution. The technical novelty of ARAMS lies in its predictive allocation model that dynamically maps IoT workloads to heterogeneous fog resources based (CPU/GPU) on computational intensity and latency constraints. Experimental evaluations on real-world multi-domain IoT traffic datasets demonstrate that ARAMS achieves a 99.4% allocation ratio at 900-task capacity, reduces allocation time to 0.07 seconds, maintains latency below 0.12 seconds, and improves Strict Prediction–Constrained Allocation Accuracy (Strict PCAA) to 0.90, outperforming Hybrid Architecture (HA) and FCFS baselines. These results confirm ARAMS’s scalability, adaptivity, and computational efficiency for heterogeneous fog environments.
Luthfan Hadi Pramono, Shan-Hsiang Shen· Cluster Computing· 0 citations
Efficient resource allocation in Cloud computing necessitates the capability to accurately predict and adapt to Cloud workloads with both precision and flexibility. Nevertheless, current approaches struggle to predict Cloud workloads with high dimensionality. This leads to inefficient resource utilization and the failure to meet service level agreements (SLAs). For these important challenges to be addressed, a dynamic recurrent neural network is proposed to accurately predict workloads. The proposed algorithm integrates an auto-encoder to effectively extract representations from the original workload data with high dimensionality to enable adaptive and accurate predictions for highly variable workloads. Experiments are conducted using Google Cloud data centers to validate the effectiveness and adaptability of the proposed scheme across various workload types and prediction lengths. The proposed scheme achieves 81.48% lower MSE and maintains an MSE of 0.005 to 0.015 compared to traditional Deep learning-based methods with up to 0.027 for Cloud workloads. This is achieved by successfully predicting continuous resource metrics (CPU/Memory/GPU utilization) rather than categorical labels, demonstrating strong adaptability across various prediction lengths.
Okore Kalu, Chijioke Okafor, P. Asuquo et al.· E3S Web of Conferences· 0 citations
Sensitivity and ablation studies confirm stable learning and controllable latency-cost trade-offs, demonstrating that lightweight RL can effectively deliver cost-efficient, adaptive autoscaling in hybrid cloud environments.
Bekzat Kobei, N. Seilova, Zarina A. Kashaganova· AI@DTESI· 0 citations
Energy efficiency in cloud computing has become a critical concern due to the growing energy demands of large-scale datacenters, operational costs and environmental impacts. This research proposes Deep Learning-Enhanced Energy-Aware Task Scheduling (DL-EATS), a novel solution that combines LSTM-based workload prediction with a hybrid multi-objective optimization model integrating Genetic Algorithm and Particle Swarm refinement. DL-EATS intelligently schedules tasks to virtual machines or containers by anticipating workload variations, thereby minimizing energy consumption, ensuring SLA compliance and maximizing system throughput. The method was evaluated against state-of-the-art algorithms, including Reinforcement Learning-based Virtual Machine Placement (RLVMP), Enhanced Adaptive Moth-Flame Optimization (EA-MFO), Q-Learning-based Resource Dynamic Optimization (Q-RDO) and Task Scheduling using Grey Wolf Optimizer (TS-GWO), under identical cloud workload scenarios. Experimental results demonstrate that DL-EATS achieves the lowest energy consumption (480 kWh), shortest makespan (38 s), minimal SLA violation rate (1.2%) and highest resource utilization (92%), representing an 18.5% improvement in energy efficiency over the next best method and substantial gains across all performance metrics. These findings confirm that integrating predictive deep learning with hybrid heuristic optimization provides a scalable, reliable and energy-efficient solution for modern cloud datacenter management.
Abdulmumini Adamu, A. A. Abdulwasiu· Journal of Science Research...· 0 citations
Cloud computing provides scalable and cost-effective resources for modern digital enterprises, but increasing workload diversity, changing user demands, and complex infrastructures make resource management challenging. Traditional resource allocation methods often fail to adapt to dynamic cloud environments, resulting in inefficient resource usage, SLA violations, and higher operational costs. This study proposes a Context-Aware AI framework for dynamic cloud resource management that incorporates workload patterns, user behavior, network conditions, infrastructure health, and business objectives. The framework combines context acquisition, real-time analytics, Long Short-Term Memory (LSTM) workload prediction, Deep Reinforcement Learning (DRL)-based optimization, and adaptive orchestration. Experimental results show improved resource utilization, response time, energy efficiency, cost reduction, and service reliability. The framework supports autonomous cloud management and provides a foundation for future technologies such as edge computing, IoT, 6G networks, and intelligent enterprise applications.
Richard Evans, Karen Lewis· International Journal of App...· 0 citations