Supporting artificial intelligence (AI) models training is one of the visions for future 6th generation (6G) networks. An extensive quantum of data and computational capabilities are necessitated for the training of A...
详细信息
ISBN:
(纸本)9798350304060;9798350304053
Supporting artificial intelligence (AI) models training is one of the visions for future 6th generation (6G) networks. An extensive quantum of data and computational capabilities are necessitated for the training of AI models. However, withthe development of AI models, it is evident that the existing edge computing network architectures are inadequate to meet the massive computing power and communication demands of distributed training for models with a growing number of parameters. In this paper, we propose a distributed training framework based on the edge-network-cloud architecture. Considering the architecture of the network and the computing capabilities of network nodes, the framework actively adapts the functional partitioning and allocation of data of the network nodes during the process of distributed training. Specifically, aggregation nodes are responsible for parameter aggregation and updating, while training nodes execute training tasks and transmit model gradients to the aggregation nodes asynchronously. To improve training efficiency and reduce communication time, we introduce a solution based on Deep Reinforcement Learning (DRL). the algorithm intelligently allocates suitable data to nodes and selects node types by task-related information, thus accelerating distributed training across network nodes. Experimental results demonstrate that the proposed algorithm effectively accelerates large-scale model training tasks.
this work proposes a real-time sentiment analysis pipeline on customer feedback using Yelp and addresses the high-volume dynamic user-generated contents processing problem. the proposal integrates state-of-the-art mac...
详细信息
the rapid development of industrial automation technology has made intelligent maintenance of electrical equipment the key to ensuring production and safety. the purpose of this article is to conduct an in-depth study...
详细信息
Due to the complex internal structure of capacitors and frequent switching operations, failures occur frequently. the damage to capacitors is closely related to their manufacturing and installation quality, as well as...
详细信息
Obstacle avoidance is a critical aspect of robotics that plays a vital role in ensuring safe and efficient operations. Detecting obstacles in complex real-Time environments poses several challenges in robotics. For ex...
详细信息
the proceedings contain 15 papers. the special focus in this conference is on Cloud computing and Artificial Intelligence: Technologies and Applications. the topics include: CAReNet: A Promising AI Architecture for...
ISBN:
(纸本)9783031786976
the proceedings contain 15 papers. the special focus in this conference is on Cloud computing and Artificial Intelligence: Technologies and Applications. the topics include: CAReNet: A Promising AI Architecture for Low Data Regime Mixing Convolutions and Attention;Designing Converged Middleware for HPC, AI, and Big Data: Challenges and Opportunities;key Mechanisms and Emerging Issues in Cloud Identity systems;Power Consumption in HPC-AI systems;integrated Architecture for Cloud and IoT with Logical sensors and Actuators - Logical IoT Cloud;the Need for HPC in AI Solutions;facts and Issues of Neural Networks for Numerical Simulation;scalable Deep Learning for Industry 4.0: Speedup withdistributed Deep Learning and Environmental Sustainability Considerations;multi-domain Dataset for Moroccan Arabic Dialect Sentiment Analysis in Social Networks;on the Challenges of Migrating to Microservices Architectures for Better Cloud Solutions;missing Data Imputation Approach for IoT Using Machine Learning;holistic Approach for Enhanced Object Recognition in Complex Environments;analyzing Sentiment in Arabic Tweets: A Study Using Machine Learning and Deep Learning Techniques.
the proceedings contain 18 papers. the topics discussed include: investigating root causes of authentication failures using a SAML and OIDC observatory;dependable wildlife DTN: wearable animal resource optimization fo...
ISBN:
(纸本)9781728176512
the proceedings contain 18 papers. the topics discussed include: investigating root causes of authentication failures using a SAML and OIDC observatory;dependable wildlife DTN: wearable animal resource optimization for sustainable long-term monitoring;physical fingerprinting of ultrasonic sensors and applications to sensor security;an ensemble feature selection method for IoT IDS;an attack tree template based on feature diagram hierarchy;modeling and analysis of malware propagation for cluster-based wireless sensor networks;a scalable blockchain based system for super resolution images manipulation;and a secret sharing scheme based on game theory and bp neural network.
this study presents a novel approach to enhancing touch sensor technology by utilizing machine learning techniques and neuromorphic models. Tactile sensing has struggled to accurately understand complex physical inter...
详细信息
the massive ultra-reliable and low-latency communications (mURLLC) services are emerging as a new traffic type to support massive numbers of mobile users (MUs) demanding the stringent delay and error-rate bounded qual...
详细信息
ISBN:
(纸本)9798350354720;9798350354713
the massive ultra-reliable and low-latency communications (mURLLC) services are emerging as a new traffic type to support massive numbers of mobile users (MUs) demanding the stringent delay and error-rate bounded quality-of-services (QoS) requirements over 6G. Among multiple 6G mURLLC services, digital twins (DT) has been widely envisioned as a major intelligent application to support efficient interactions between physical and virtual objects. Moreover, multi-tier caching, which is one of the key distributedcomputing techniques, stores the frequently-demanded data items at different wireless network tiers to efficiently reduce mURLLC streaming delay and data move. However, how to efficiently cache mURLLC-based DT data items at different caching tiers of wireless networks and how to statistically upper-bound both delay and error-rate for DT communications remain challenging problems. To overcome these difficulties, in this paper we propose a multi-tier caching mechanism to support DT communications over 6G mobile networks. First, we propose the DT data adaptive collection scheme applying finite blocklength coding (FBC) to dynamically encode a physical object into its virtual representation according to the current network and wireless channel statuses. Second, we develop inter-tier and intra-tier collaborative caching mechanisms, where DT data items are selectively cached at different wireless network caching tiers according to their popularities including: router tier, massive-multiple-input-multiple-output (MIMO) basestation tier, and mobile device tier. third, our proposed intertier collaborative caching mechanisms maximize the aggregate epsilon-effective capacity across all three caching tiers, and our proposed intra-tier collaborative caching mechanisms minimize the sum of data transmission delay for all DT data items cached in each caching tier. Finally, we numerically validate and evaluate our developed multi-tier hierarchical caching schemes over 6G DT-enab
the distributed in-wheel motor drive system features a simple mechanical structure, high transmission efficiency, flexible control, and ease of integrated design, making it an inevitable trend in the technological dev...
详细信息
暂无评论