Fig 8 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Source publication
![Fig. 1. Growth of computing over time [1].](publication/364336162/figure/fig1/AS:11431281095739314@1668008974297/Growth-of-computing-over-time-1_Q320.jpg)
Artificial intelligence (AI) and 5G system have been two hot technical areas that are changing the world. On the deep convergence of computing and communication, networking systems of AI (NSAI) is presenting a paradigm shift, where distributed AI becomes immersive in all elements of the network, i.e., cloud, edge, and terminal devices, which make A...
Context in source publication
Context 1
... Micro and Nano Electronics: As shown in Fig. 8, stateof-the-art AI electronics are mainly divided into general chips and special chips in terms of technical architecture. The general-purpose chip design refers to the traditional chip architecture, which supports deep learning and complex neural network algorithm through software programming, mainly including CPU, GPU, DSP, FPGA, ...
Citations
... Another new direction in promising research is the combination of computing and communication for the development of AI-based networking systems, functional at different levels of the network. Of course, this is a strategy in which AI should be further embedded into the communication network infrastructure, drawing itself closer to the realization of ubiquitous brain networks (UBNs), which may likely redefine the management and processing of data across networks [23]. ...
... These developments have spurred the emergence of Intelligent Decision Support Systems (IDSS), becoming increasingly crucial for managing and understanding the vast data volume produced by IoE devices. IDSS, integrating a variety of technologies such as deep learning, machine learning, evolutionary algorithms, artificial neural networks, and intelligent agents, are revolutionizing wireless communication and network efficiency within IoE platforms [2]. Their applications span across diverse domains including business, healthcare, and smart transportation, facilitating advanced data analysis and decision-making support. ...
In an era where Intelligent Decision Support Systems (IDSS) are integral to managing the vast data from Internet of Everything (IoE) systems, this study introduces IDSDeep-CCD, a novel IDSS approach for detecting concrete cracks, a critical issue in civil infrastructure maintenance. Traditional visual inspection methods for crack detection are time-consuming and error-prone. To address this, IDSDeep-CCD employs deep learning, utilizing an open-source dataset of concrete crack images for enhanced accuracy. The system's two-stage approach, featuring feature extraction via pre-trained models and a fully connected network for classification, significantly outperforms existing methods. Our results demonstrate a remarkable 99.9% accuracy rate in crack detection without data augmentation, marking a notable advancement in automatic detection technology. This breakthrough offers practical benefits for the industry by enabling more precise and efficient identification of concrete cracks, paving the way for timely and accurate remedial actions.
... The future of cloud-based accounting lies at the intersection of these trends, fostering a holistic and integrated approach to efficiency, scalability, and data security. AI, blockchain, edge computing, and other emerging technologies will converge to create a dynamic and robust ecosystem for accounting in the digital age (Song et al., 2022). ...
Cloud computing has emerged as a transformative force, revolutionizing the landscape for accounting firms. This comprehensive review delves into the profound impact of cloud computing on accounting firms, focusing on key dimensions such as efficiency, scalability, and data security. Examining the shift from traditional infrastructure to cloud-based solutions, the review navigates through the tangible benefits and potential challenges that cloud adoption brings to the accounting domain. Efficiency stands out as a cornerstone of cloud computing's influence on accounting firms. The agility and accessibility offered by cloud-based platforms streamline routine tasks, facilitating seamless collaboration among accounting professionals. The scalability afforded by cloud services empowers firms to dynamically adjust their computing resources, adapting to fluctuations in workload and business demands. This ensures that accounting firms can efficiently handle diverse workloads without being constrained by rigid infrastructure limitations. Scalability further intersects with efficiency, enabling accounting firms to optimize resource allocation and enhance overall productivity. The scalability of cloud solutions aligns with the dynamic nature of the accounting profession, allowing firms to scale up during peak seasons and scale down during lulls, ultimately fostering cost-effectiveness and operational agility. However, the review also critically evaluates the nuances of data security in the cloud computing paradigm. Addressing concerns related to data privacy, confidentiality, and compliance, the review navigates the intricate landscape of securing financial data in a cloud-based environment. It probes into the robustness of encryption protocols, authentication mechanisms, and compliance frameworks, ensuring a comprehensive understanding of the security implications inherent in cloud adoption by accounting firms. In conclusion, this review encapsulates the multifaceted impact of cloud computing on accounting firms. Efficiency gains and scalability advantages are juxtaposed against the imperative of fortifying data security. This examination provides a roadmap for accounting professionals, offering insights into harnessing the full potential of cloud technologies while ensuring the integrity and security of sensitive financial data. As accounting firms increasingly pivot towards cloud adoption, this review serves as a strategic guide, equipping practitioners with the knowledge to navigate the evolving landscape of cloud computing in the realm of accounting.
... In contrast, a mesh topology could distribute data across multiple nodes, providing a larger and more diverse data set for training. We refer to [277] for a comprehensive survey on the convergence, robustness and privacy of ML algorithms with respect to network architecture and implementation in the context of 5G networks. ...
... The work of [309] considers itself as an update to [306], covering more recent developments and discussing recent IDS datasets. Additionally, several surveys consider ML approaches for a subset of networked systems, such as vehicular networks in [315], [316], Software-Defined Networks (SDN) in [305], mobile/wireless/ubiquitous networks in [4], [277], edge computing [303], [304] or network traffic monitoring and analysis [313]. The work of [307] takes a unique stance and provides the joint application of recent ML and Blockchain technologies for networking problems. ...
Machine learning has found many applications in network contexts. These include solving optimisation problems and managing network operations. Conversely, networks are essential for facilitating machine learning training and inference, whether performed centrally or in a distributed fashion. To conduct rigorous research in this area, researchers must have a comprehensive understanding of fundamental techniques, specific frameworks, and access to relevant datasets. Additionally, access to training data can serve as a benchmark or a springboard for further investigation. All these techniques are summarized in this article; serving as a primer paper and hopefully providing an efficient start for anybody doing research regarding machine learning for networks or using networks for machine learning.
... This layer consists of applications centered on network capabilities and under network operator control to provide enriched services to vertical applications. In the context of 6G systems, network-centric applications emerge to increase automation and distributed intelligence characteristics of the network services ( [46], [47]). Where applicable, they take advantage of increasing in-network capabilities such as Data Management and AI techniques [46] [47] [48], or compute [49], etc, with further integration (e.g. ...
As the fifth generation (5G) mobile communication systems are commercially deployed, they bring new services, enhance user experiences, and offer various opportunities to different industries. Despite its advancements, 5G encounters several challenges. To tackle these issues, global industrial, academic, and standards organizations are actively researching on sixth generation (6G) wireless communication systems. 6G networks are envisioned as a transformative shift in the interactions between the physical, digital, and human realms, paving the way for a pervasive human-centered cyber-physical world. 6G is positioned to be a platform that offers communication and beyond communication services considering both performance and value centric technological development approaches. In this paper, we present the way forward towards the design of 6G end-to-end (E2E) system as a consolidated view of leading industry stakeholders and academia in the European level 6G flagship project Hexa-X-II. We discuss the key challenges with 5G and how 6G is expected to tackle those by introducing new technological innovations and supporting novel use case requirements. We provide a comprehensive methodology for the design of a 6G E2E system including ten principles, a blueprint, and a structured design process. The architecture design principles prioritize environmental sustainability, digital inclusiveness, and trustworthiness, considering their impact on the 6G E2E system. The blueprint is described corresponding to the infrastructure, network centric application, and application layers, as well as the pervasive functionalities and the relevant technological innovations. Following the design principles and the system blueprint, the design process is demonstrated as two-way approaches (i.e., 1) key performance and value indicators based design process. 2) top-down versus bottom-up alignment process) through the lens of a collaborative robot use case. Through this use case, a special attention is given to the technological enablers that cover management and orchestration functionalities and their 6G enhancement to go beyond the limitations characterizing the previous network generation. In addition, virtual modelling aspects related to digital twining and simulations for 6G E2E system design are also discussed.
... A evolução das redes celulares trouxe inovações para atenderàs diversas demandas da indústria, destacando-se as redes de quinta geração (5G), capazes de suportar uma ampla gama de serviços [Liu et al. 2022]. Nesse contexto, as aplicações de Realidade Estendida (XR) se destacam, combinando elementos do mundo real e virtual para proporcionar experiências imersivas ] [Song et al. 2022]. ...
... A arquitetura de renderização dividida proposta pelo 3rd Generation Partnership Project (3GPP), na qual o tráfego XRé enviado intermitentemente para servidores de borda, pode resultar em atrasos e problemas de transmissão, afetando a experiência do usuário [Song et al. 2022]. Entretanto, com a chegada das redes de sexta geração (6G), espera-se atenderàs crescentes demandas das aplicações imersivas, oferecendo comunicações hiperconectadas e inteligentes [Liu et al. 2022]. ...
Esta revisão sistemática tem como objetivo principal identificar estudos que abordem questões relacionadas às aplicações imersivas com base nas tecnologias habilitadoras B5G/6G, MEC e IA. A motivação é compreender os avanços realizados nesse campo e identificar soluções emergentes e lacunas de conhecimento. Os resultados obtidos indicam a predominância do uso da técnica de Aprendizado por Reforço Profundo para abordar soluções de rede baseadas na computação de borda de múltiplo acesso, a fim de viabilizar as aplicações imersivas. Espera-se que esta revisão sistemática contribua para uma melhor compreensão do estado atual das aplicações imersivas e para pesquisas futuras relacionadas ao tema.
... Additionally, to support its attractive applications, varied networking architecture, and various resources, 6G will become increasingly complicated. As a result, traditional network management approaches involved with human-controlled ideologies will become insufficient, necessitating the use of advanced artificial intelligence (AI)-based solutions [445] inside the RAN for achieving self-organizing, autonomous operation, and capital expenditure or operational expenditure savings [446]- [448]. ...
With advancements of cloud technologies Multi-Access Edge Computing (MEC) emerged as a remarkable edge-cloud technology to provide computing facilities to resource-restrained edge user devices. Utilizing the features of MEC user devices can obtain computational services from the network edge which drastically reduces the transmission latency of evolving low-latency applications such as video analytics, e-healthcare, etc. The objective of the work is to perform a thorough survey of the recent advances relative to the MEC paradigm. In this context, the work overviewed the fundamentals, architecture, state-of-the-art enabling technologies, evolving supporting/assistive technologies, deployment scenarios, security issues, and solutions relative to the MEC technology. The work, moreover, stated the relative challenges and future directions to further improve the features of MEC.
... Video anomaly detection (VAD) aims to automatically analyze the spatial-temporal patterns and contactlessly detect anomalous events of concern (e.g., traffic accidents, violent acts, and illegal operations) from surveillance videos [1], [2], which has promising applications in emerging areas such as traffic management [3], [4], security protection [5], [6], and intelligent manufacturing [7], [8]. However, anomaly is a vague concept, making anomalous events unbounded and difficult to predefine. ...
Video anomaly detection is an essential yet challenging task in the multimedia community, with promising applications in smart cities and secure communities. Existing methods attempt to learn abstract representations of regular events with statistical dependence to model the endogenous normality, which discriminates anomalies by measuring the deviations to the learned distribution. However, conventional representation learning is only a crude description of video normality and lacks an exploration of its underlying causality. The learned statistical dependence is unreliable for diverse regular events in the real world and may cause high false alarms due to overgeneralization. Inspired by causal representation learning, we think that there exists a causal variable capable of adequately representing the general patterns of regular events in which anomalies will present significant variations. Therefore, we design a causality-inspired representation consistency (CRC) framework to implicitly learn the unobservable causal variables of normality directly from available normal videos and detect abnormal events with the learned representation consistency. Extensive experiments show that the causality-inspired normality is robust to regular events with label-independent shifts, and the proposed CRC framework can quickly and accurately detect various complicated anomalies from real-world surveillance videos.
... Song et al. [118] AI, Networkng systems AI (NSAI) ...
... Song et al. [118] Networking systems of AI (NSAI) This paper's contribution entails (1) offering an all-encompassing structure for the deep convergence of computing and communications, where the network and implementation can be collectively improved as one cohesive unit, and (2) suggesting the overall strategy and addressing open investigations in acknowledging the online-evolutionary convergence of the digital world, the natural environment, and society as a whole, towards the ubiquitous neurological systems (UBNs), which require cooperation in the functioning of the internet, the physical world, and society at large. ...
Mobile edge computing (MEC) supported by non-orthogonal multiple access (NOMA) has recently gained a lot of interest due to its improved ability to lessen power consumption and MEC offload delay. In recent decades, the need for wireless communications has increased tremendously. Fifth-generation (5G) communications will soon be widely used and offer much more functionality than a fourth generation (4G). Between 2027 and 2030, an innovative wireless communication paradigm is known as the sixth generation (6G) system is projected to be introduced with the full help of artificial intelligence (AI). Advanced system capacity, higher data rate, lower latency, advanced security, and improved quality of service (QoS) than 5G systems are a few of the main challenges to resolve with 5G. The growing need for data rates in the sixth generation (6G) communication networks are being met by extraordinary technologies such as NOMA, Soft Computing (SC), and MEC. Owing to the massive attention to the NOMA-enabled MEC, there has been a significant spike in the number of papers published in this area, while more comprehensive studies and classifications are still needed. Using the preferred reporting items for systematic reviews and meta-analysis (PRISMA) guidelines, the investigation reports a systematic literature review (SLR) of NOMA-enabled MEC. This survey also evaluates numerous pieces of literature prudently chosen over a multi-step procedure and meets the selection criteria described in the paper summarizing our review.
... With the rapid development of smart devices and the internet of things, various smart applications based on lightweight terminals are becoming popular. For example, today's smartphones are capable of higher computing and communication capabilities while consuming less power, making it possible to detect driving styles by portable devices [40]. In recent years, driving style detection (DSD) has been widely investigated in real-life applications, such as traffic management, car insurance and fuel consumption optimization [7,13]. ...
Driving style detection is an essential real-world requirement in diverse contexts, such as traffic safety, car insurance and fuel consumption optimization. However, the existing methods either rely on handcrafted features or fail to explore deep spatial-temporal features from multi-modal sensing signals. In this paper, we propose a novel attention-based hybrid convolutional neural network (CNN) and long short-term memory (LSTM) framework named DSDCLA to address these problems. Specifically, DSDCLA first introduces CNN and self-attention for extracting local spatial features from multi-modal driving sequences. Then, we utilize LSTM and multi-head attention to explore the long-term temporal relationships between timesteps. Therefore, DSDCLA can identify driving style by short- and long-term spatial-temporal features. Furthermore, we design three variants with different levels of fusion, which shows the advantage of selecting components and improves the interpretability. We extensively evaluated the proposed DSDCLA on two public real-world datasets, and the experimental results show that DSDCLA outperforms the current state-of-the-art methods, achieving the F1-scores of 97.03% and 97.65%. Numerous ablation studies and visualizations indicate the effectiveness of the model and the importance of multi-level attention fusion for identifying driving style between timesteps.
