A survey on resource scheduling approaches in multi-access edge computing environment: a deep reinforcement learning study

Cluster Computing, Год журнала: 2025, Номер 28(3)

Опубликована: Янв. 21, 2025

Язык: Английский

---

SCOF: Security-Aware Computation Offloading Using Federated Reinforcement Learning in Industrial Internet of Things With Edge Computing

IEEE Transactions on Services Computing, Год журнала: 2024, Номер 17(4), С. 1780 - 1792

Опубликована: Март 19, 2024

Industry 5.0 facilitates the intelligent upgrade of smart factories in Industrial Internet Things (IIoT), and also introduces a plethora data processing challenges. Mobile edge computing offloads to servers for processing, easing pressure reducing system cost. However, contain numerous sensitive information, offloading them directly may pose risk leakage. To address these challenges, we investigate localedge collaborative factory system. Specifically, firstly model tasks as directed acyclic graph, formulate problem Markov decision process, considering optimization latency, energy consumption number overtime tasks. Then, propose security-aware computation method using federated reinforcement learning IIoT, named SCOF. SCOF employs learning, keeping local uploading parameters aggregation. The transmission passes through an artificial noise channel protect against eavesdropping. Meanwhile, utilizes differential privacy security deep selecting near-optimal decisions. Finally, abundant experiments are conducted under real dataset. results show that has better perfomance than state-of-the-art baseline algorithms.

Язык: Английский

---

Towards sustainable smart cities: Workflow scheduling in cloud of health things (CoHT) using deep reinforcement learning and moth flame optimization for edge-cloud systems

Future Generation Computer Systems, Год журнала: 2025, Номер unknown, С. 107821 - 107821

Опубликована: Март 1, 2025

Язык: Английский

---

Artificial Intelligence-Defined Wireless Networking for Computational Offloading and Resource Allocation in Edge Computing Networks

IEEE Open Journal of the Communications Society, Год журнала: 2024, Номер 5, С. 2039 - 2057

Опубликована: Янв. 1, 2024

The advent of the Internet Everything and new Ultra-Reliable Low-Latency Communication (URLLC) services has resulted in an exponential growth data demands at network's edge. To meet stringent performance requirements evolving 5G (and beyond) applications, deploying dedicated resources closer to mobile users is essential. Multi-Access Edge Computing (MEC) a promising technology for bringing computational users. However, distributed limited MEC must be effectively optimized maximize number benefiting from low-latency each time slot highly congested, large-scale, dynamic wireless network scenarios. In this research, we propose evaluate novel Artificial Intelligence-Defined Wireless Networking (AIDWN) approach that builds on conventional Software-Defined (SDN), implementing AI-defined application plane offloading resource allocation MEC-enabled networks. AIDWN implements deep reinforcement learning framework neural networks dynamically adapt optimal decisions while considering handover, mobility, coordinated challenges multi-MEC server environments. Compared recent state-of-the-art proposals, proposed demonstrates substantial improvement, utilizing more than 90% per across all servers. It also accommodates significantly congested We identified various future research directions highlighting potential simplifying management next-generation

Язык: Английский

---

Towards Failure-Aware Inference in Harsh Operating Conditions: Robust Mobile Offloading of Pre-Trained Neural Networks

Electronics, Год журнала: 2025, Номер 14(2), С. 381 - 381

Опубликована: Янв. 19, 2025

Pre-trained neural networks like GPT-4 and Llama2 have revolutionized intelligent information processing, but their deployment in industrial applications faces challenges, particularly harsh environments. To address these related issues, model offloading, which involves distributing the computational load of pre-trained models across edge devices, has emerged as a promising solution. While this approach enables utilization more powerful models, it significant challenges environments, where reliability, connectivity, resilience are critical. This paper introduces failure-resilient inference mobile (FRIM), framework that ensures robust offloading without need for retraining or reconstruction. FRIM leverages graph theory to optimize partition redundancy incorporates an adaptive failure detection mechanism with efficient fault tolerance. Experimental results on DNN (AlexNet, ResNet, VGG-16) show improves performance resilience, enabling reliable operating

Язык: Английский

---

Optimizing Workflow Offloading and Migration under Timed Constraints in Fog and Cloud Computing

Journal of Grid Computing, Год журнала: 2025, Номер 23(1)

Опубликована: Фев. 26, 2025

Язык: Английский

---

Resource-aware and computation offloading based on space–air–ground–sea integrated network

The Journal of Supercomputing, Год журнала: 2025, Номер 81(5)

Опубликована: Март 18, 2025

Язык: Английский

---

AI-driven job scheduling in cloud computing: a comprehensive review

Artificial Intelligence Review, Год журнала: 2025, Номер 58(7)

Опубликована: Апрель 11, 2025

Язык: Английский

---

Peer-driven task scheduling and resource allocation for enhanced performance in industrial IoT systems

Scientific Reports, Год журнала: 2025, Номер 15(1)

Опубликована: Апрель 25, 2025

Язык: Английский

---

Edge Computing: Virtualization Techniques for Performance Optimization and Management Analysis

Lecture notes in computer science, Год журнала: 2025, Номер unknown, С. 88 - 98

Опубликована: Янв. 1, 2025

Язык: Английский

---