Edge Of Network Research Articles

Programmable packet-optical network security and monitoring using DPUs with embedded GPUs [Invited

Data processing units (DPUs) with embedded graphics processing units (GPUs) have the potential to revolutionize optical network functionalities at the edge. These advanced units can significantly enhance the performance and capabilities of optical networks by integrating powerful processing capabilities directly at the network edge, where data is generated and consumed. We explore the use cases for DPUs in optical data monitoring with local artificial intelligence (AI) processing and embedded security. This paradigm shift aims to enable more efficient data handling, reduced latency, and improved overall network performance by leveraging local AI processing capabilities embedded within DPUs. In this paper, we show how DPUs can analyze vast amounts of optical data in real-time, implementing advanced data analysis algorithms and security protocols directly on the DPUs to provide robust monitoring and protection for the optical networks. Results indicate that DPUs with embedded GPUs can significantly improve the detection and response times to network anomalies, performance issues, and security threats.

TinyML and edge intelligence applications in cardiovascular disease: A survey.

Tiny machine learning (TinyML) and edge intelligence have emerged as pivotal paradigms for enabling machine learning on resource-constrained devices situated at the extreme edge of networks. In this paper, we explore the transformative potential of TinyML in facilitating pervasive, low-power cardiovascular monitoring and real-time analytics for patients with cardiac anomalies, leveraging wearable devices as the primary interface. To begin with, we provide an overview of TinyML software and hardware enablers, accompanied by an examination of networking solutions such as Low-power Wide area network (LPWAN) that facilitate the seamless deployment of TinyML frameworks. Following this, we delve into the methodologies of knowledge distillation, quantization, and pruning, which represent the cornerstone strategies for optimizing machine learning models to operate efficiently within resource-constrained environments. Furthermore, our discussion extends to the role of efficient deep neural networks tailored specifically for cardiovascular monitoring on wearable devices with limited computational resources. Through a comprehensive review, we analyze the applications of prominent artificial neural network architectures including Convolutional Neural Networks (CNNs), Autoencoders, Deep Belief Networks (DBNs), and Transformers in the domain of Electrocardiogram (ECG) analytics, shedding light on their efficacy and potential in advancing healthcare technology.

Advancing Cybersecurity: A Machine Learning Framework for Detecting Distributed Denial of Service (DDoS) Attacks

This paper explores the nature of threats posed by Distributed Denial of Service (DDoS) attacks on large networks, such as the Internet, emphasizing the need for effective detection and response mechanisms. These mechanisms must be implemented not only at the network edge but also within its core. The paper introduces methods to detect DDoS attacks by analyzing entropy and frequency-sorted distributions of specific packet attributes. Anomalies in these attributes' characteristics serve as indicators of potential DDoS attacks. The proposed methods are evaluated for detection accuracy and performance using live traffic traces collected from diverse network environments, including core Internet nodes and edge networks. Results demonstrate the effectiveness of these methods against current DDoS attacks and provide insights into improving detection capabilities for more sophisticated, stealthier threats. Additionally, the paper describes a detection-response prototype and discusses how the detection system can be extended to support effective response decision-making.

Quantitative analyses of single mitochondrial components reveal an early role of small-mitochondrial-networks in priming neoplasticity.

The challenges of understanding the complex role of the multifaceted and heterogenous cellular organelles, mitochondria, can be potentially overcome with their quantitative analyses. We use a combinatorial approach of quantitative analyses of single-mitochondrial-components and scRNA-seq to elucidate a mechanism of mitochondrial priming of cancer initiation by a model carcinogen. We propose that conversion of large Hyperfused-Mitochondrial-Networks (HMNs) to Small-Mitochondrial-Networks (SMNs) primes non-transformed keratinocytes towards their neoplastic transformation. Mechanistically, the SMNs, enriched by modulation of the physical nodes and edges of mitochondrial networks, tunes mitochondrial redox balance to establish transcriptomic interactions towards specifying a state of stemness. Further probing of our fundamental findings in the light of cancer heterogeneity may facilitate refinement of the various proposed mitochondria based targeted cancer therapies.

RAM-MEN: Robust authentication mechanism for IoT-enabled edge networks

The rapid expansion of Mobile Edge Computing (MEC) and the Internet of Things (IoT) has revolutionized technology by enabling real-time data processing at the network edge, which is essential for applications such as autonomous vehicles and smart cities. With the advent of 6G networks, which promise ultra-fast speeds, vast connectivity, and low-latency communication, MEC-IoT systems are becoming more powerful but also face significant security challenges. Existing authentication mechanisms (AMs) are often vulnerable to attacks like impersonation and insider threats. This paper introduces a novel lightweight AM, called RAM-MEN that employs cryptography and physically unclonable functions (PUFs) to secure IoT-enabled MEC environments in the 6G era. It protects against insider threats and fake MEC access points while ensuring efficiency and scalability. Additionally, the proposed RAM-MEN establishes a secure communication channel (session key) between IoT devices and the MEC server, enabling secure offloading of computationally intensive tasks. The security of the session is rigorously evaluated using formal methods, including Scyther and the random or real model, alongside informal approaches. Comparative performance evaluations show that the proposed RAM-MEN reduces communication costs by 21.54% to 45.53% and computational costs by 17.09% to 83.72%, while providing enhanced security features.

Self-Learning Based Dependable Offloading Optimization in Semi-Trusted Vehicular Edge Computing and Networks

Self-Learning Based Dependable Offloading Optimization in Semi-Trusted Vehicular Edge Computing and Networks

A New English Education Model Based on 6G and Sliced Network Virtual Reality Platform

ABSTRACTThe information society has led to a shift in traditional English education methods, with the evolution of technology, particularly internet and communication network technologies, reshaping the teaching landscape. This facilitated innovative instructional approaches and enhanced the learning experience. This research introduces a novel virtual learn net architecture (VLNA) within the 6G network layers, which processes the performance of the virtual reality‐based English education system (VR‐EES) model to provide a seamless, personalized learning experience for online learners. This architecture is structured into several layers: The user equipment (UE) layer connects VR headsets to the network with ultrareliable, low‐latency links; the radio access network (RAN) layer, employing massive MIMO and beam forming, enhances connection speed, capacity, and coverage. Edge computing handles latency‐sensitive tasks like speech recognition and adaptive content delivery, reducing the load on the core network. The core network layer (CLN) manages network slices for specific learning tasks such as real‐time interaction, high‐definition multimedia, and computation‐intensive processes, with control plane and user plane separation (CUPS) optimizing network management and security through end‐to‐end encryption. Software‐defined networking (SDN) and network function virtualization (NFV) provide centralized, dynamic control, allowing real‐time resource allocation based on demand. Cloud‐edge integration supports Artificial intelligence (AI)‐driven adaptive learning, optimizing educational content delivery based on individual progress. The study results demonstrate that stimulation of VLNA achieved significant improvements in latency reduction, bandwidth utilization, throughput, packet loss rate, jitter, user engagement, learning efficiency, and user satisfaction. The integration of edge computing and network slicing led to a significant reduction in latency, while the enhanced throughput enabled seamless VR experiences. In this study, latency reduction, bandwidth utilization, and user satisfaction emerge as the most significant factors, with user satisfaction standing out as the top performer due to its substantial impact on enhancing the overall learning experience. The packet loss rate is maintained to a certain level, ensuring reliable data transmission. The VR‐EES model's experimental results also enhanced visual learning, multimedia quality, user pleasure, learning effectiveness, and user engagement.

Extreme soil salinity reduces N and P metabolism and related microbial network complexity and community immigration rate.

Soil microbiomes are well known to suffer from the effects of rising salinity. There are, however, no current understandings regarding its specific effects on microbial metabolic functions associated with nitrogen (N) and phosphorus (P) cycling, particularly in the Yellow River Delta (YRD), one of the largest estuaries in the world. This research examined soil microbiomes at 50 sites in the YRD region to analyze their co-occurrence networks and their relationship with N (nitrification, denitrification, dissimilatory, assimilatory, fixation, and mineralization) and P (solubilization, mineralization, transportation, and regulation) metabolism processes. Our findings indicate a notable reduction in soil multifunctionality as salinity levels increase, with Halofilum-ochraceum playing a significant role in nitrification, whereas Bacteroidetes-SB0662-bin-6 helps solubilize inorganic P in highly saline areas. High soil salinity negatively affected the amoA gene involved in nitrification and increased the nosZ gene involved in denitrification in extreme salinity soil with 8.2g/kg salt content. Extreme salinity significantly reduced the expression of genes involved in inorganic P solubilization, such as ppa and ppx. Additionally, the alkaline P gene phoD exhibited significant decreases in extremely saline soils, thereby impeding the mineralization of organic P. The neutral community models indicated that microbial community immigration rate showed a linear negative relationship with soil EC in the six N and four P processes. Salinization, however, displayed a nonlinear pattern with clearly defined thresholds on the community of microbes involved in N and P cycling. Reduced microbial diversity and interactions are causing a decline in soil multifunctionality, and the soil multifunctionality and network edges jointly limited the microbial community immigration rate involved in N and P cycling. It is crucial to preserve soil microbial functions to support nutrient cycling and predict the ecological effects of soil salinization.

Average AoI Minimization With Directional Charging for Wireless-Powered Network Edge

Average AoI Minimization With Directional Charging for Wireless-Powered Network Edge

In Vitro Model of Vascular Remodeling Under Microfluidic Perfusion

We developed a portable microfluidic system that combines spontaneous lumen formation from human umbilical endothelial cells (HUVECs) in fibrin–collagen hydrogels with active perfusion controlled by a braille actuator. Adaptive interstitial flow and feedthrough perfusion switching enabled the successful culture of spontaneously formed naturally branched lumens for more than one month. We obtained many large-area (2 mm × 3 mm) long-term (more than 30 days per run) time-lapse image datasets of the in vitro luminal network using this microfluidic system. We also developed an automatic image analysis pipeline to extract the morphology of the lumen network and node–edge network structure weighted with segmentwise flow parameters. The automatic lumen area measurements revealed that almost all lumens were successfully cultured in this system for approximately 50 days, following the meshwork, sprouting, remodeling, stability, and erosion stages. We found that the optimization of the lumen network during the remodeling stage can be explained by the decrease in the betweenness centrality of the WSS-weighted network and the increase in the strength centrality of the flow-rate-weighted network.

Prototype‐Based Collaborative Learning in UAV‐Assisted Edge Computing Networks

ABSTRACTContextThe rise of artificial intelligence of things (AloT) has enabled smart cities and industries, and UAV‐assisted edge computing networks are an important technology to support the above scenarios. UAV‐assisted refers to leveraging UAVs as a dynamic, flexible infrastructure to assist edge network data processing and communication tasks. Multiple UAVs can use their own resources, and collaborate edge servers to train artificial intelligence (Al) models.ObjectiveCompared with cloud‐based collaborative computing scenarios, UAV‐assisted edge collaborative learning can reduce training and inference delays and improve user satisfaction. However, UAV‐assisted edge networks scenario brings new challenges in terms of transmission burden and energy consumption.MethodThis paper proposes a prototype‐based joint optimization and training software system. The system consists of an optimization module and a training module. The optimization module first models an optimization problem including energy consumption and prototype error. Then it solves the optimization problem by problem transformation and plans the location of each UAV given the objects' position. After UAVs fly to the designated area and complete data collection, UAVs and the edge server train a model according to the proposed prototype‐based collaborative training module. Our training module enables multiple UAVs and an edge server to collaboratively train a model by lightweight prototype transmission and prototype aggregation. We also prove the convergence of the proposed collaborative training method.ResultsResults show our method reduces prototype error and energy consumption by at least 12.31% and improves model accuracy by 3.62% with a little communication burden.ConclusionFinally, we verify system performance through experiments.

Human-AI Enabled Edge Computing for Data Processing: A Comprehensive Analysis

The exponential growth in data generation and processing requirements has driven the need for more efficient and intelligent computational approaches at the edge of networks. This comprehensive article investigates the integration of human expertise with AI-enabled edge computing systems, focusing on optimization strategies, implementation frameworks, and real-world applications. Through extensive analysis of implementation cases and performance metrics, the research demonstrates significant improvements in processing efficiency, with systems achieving a reduction in latency and improvement in decision accuracy through human-AI collaboration. The article presents a detailed framework for implementing these hybrid systems, addressing critical aspects including technical architecture requirements, integration guidelines, and risk mitigation strategies. Results indicate substantial benefits in operational efficiency, resource utilization, and decision-making capabilities across various industrial applications. While identifying implementation challenges and limitations, the article provides strategic recommendations for successful deployment and highlights opportunities for future advancement in the field. The article contributes significantly to understanding how organizations can effectively leverage human-AI collaboration in edge computing environments to address the growing demands of data processing in modern digital infrastructure.

A Survey on Securing Image-Centric Edge Intelligence

Facing enormous data generated at the network edge, Edge Intelligence (EI) emerges as the fusion of Edge Computing and Artificial Intelligence, revolutionizing edge data processing and intelligent decision-making. Nonetheless, this emergent mode presents a complex array of security challenges, particularly prominent in image-centric applications due to the sheer volume of visual data and its direct connection to user privacy. These challenges include safeguarding model/image privacy and ensuring model integrity against various security threats, such as model poisoning. Essentially, those threats originate from data attacks, suggesting data protection as a promising solution. Although data protection measures are well-established in other domains, image-centric EI necessitates focused research. This survey examines the security issues inherent to image-centric EI and outlines the protection efforts, providing a comprehensive overview of the landscape. We begin by introducing EI, detailing its operational mechanics and associated security issues. We then explore the technologies facilitating security enhancement (e.g., differential privacy) and edge intelligence (e.g., compact networks and distributed learning frameworks). Next, we categorize security strategies by their application in data preparation, training, and inference, with a focus on image-based contexts. Despite these efforts on security, our investigation identifies research gaps. We also outline promising research directions to bridge these gaps, bolstering security frameworks in image-centric EI applications.

Real -Time Network Packet Classification Exploiting Computer Vision Architectures

The upcoming 6G and NextG networks underscore the necessity of sophisticated security methods based on Artificial Intelligence (AI) in order to detect malicious activity and adjust to new threats. Because computer vision techniques may be used to recognize complex patterns, their incorporation into the cybersecurity industry is a promising development. In this work, we present a computationally effective categorization technique that enforces the real-time conversion of packets into pictures by directly acting upon the raw packets gathered at base stations. The suggested solution's novel features include its lightweight implementation, which well satisfies the requirements of upcoming 6G networks, and its network edge operation, which permits early threat detection as near to the packet origin as feasible. We examine the efficacy of this methodology in terms of F1-score and prediction time by employing cutting-edge computer vision architectures and a customized Convolutional Neural Network (CNN) to tackle an intrusion detection task utilizing a substantial 5G dataset. The CNN design is superior than complicated models, as demonstrated by the results of experiments. The CNN consistently beats the other cutting-edge computer vision models over several packet window sizes N (i.e., 10, 50, and 100 packets), reaching very high F1-scores (0.99593, 0.99860, and 0.99895). A scalability investigation reveals a trade- off between the performance and scalability of CNN, with higher N values resulting in longer prediction times. However, the scalability of the other computer vision models is superior, allowing for an ideal model selection free of compromises

Epihiper—A high performance computational modeling framework to support epidemic science

Abstract This paper describes Epihiper, a state-of-the-art, high performance computational modeling framework for epidemic science. The Epihiper modeling framework supports custom disease models, and can simulate epidemics over dynamic, large-scale networks while supporting modulation of the epidemic evolution through a set of user-programmable interventions. The nodes and edges of the social-contact network have customizable sets of static and dynamic attributes which allow the user to specify intervention target sets at a very fine-grained level; these also permit the network to be updated in response to nonpharmaceutical interventions, such as school closures. The execution of interventions is governed by trigger conditions, which are Boolean expressions formed using any of Epihiper’s primitives (e.g. the current time, transmissibility) and user-defined sets (e.g. people with work activities). Rich expressiveness, extensibility, and high-performance computing responsiveness were central design goals to ensure that the framework could effectively target realistic scenarios at the scale and detail required to support the large computational designs needed by state and federal public health policymakers in their efforts to plan and respond in the event of epidemics. The modeling framework has been used to support the CDC Scenario Modeling Hub for COVID-19 response, and was a part of a hybrid high-performance cloud system that was nominated as a finalist for the 2021 ACM Gordon Bell Special Prize for high performance computing-based COVID-19 Research.

Using Network Analysis of Adolescent Self-Ratings of ADHD Symptoms to Identify Central Symptoms and Their Associations With Each Other and Global Functioning.

This research examined the network properties (network graph, centrality, and edge weights) of the 18 ADHD symptoms, based on the self-ratings of 300 adolescents. The findings indicated the three symptoms with the highest centrality values were "inattention", "wait", and "interrupt". For edge weights, there were positive large effect size associations for "lose" with "forgetful", "fidget with "run", "blurt" with "wait", and "wait" with "interrupt"; and positive moderate effect size associations for "careless" with "instruction", and "avoid" with "listen". Five IA symptoms ("careless", "instruction", "avoid", "distracted", and "forgetful") and one HI symptom ("quiet") were associated negatively and significantly with global functioning. Overall, these associations and relations should be prioritized when planning treatment for ADHD. This is the first study to examine the network properties of ADHD symptoms for adolescent self-ratings. The implications of the findings for theory include a better understanding of the relationships and interrelations between ADHD symptoms, especially in terms of the clustering of IA and HI symptoms and their associations with global functioning. In practice, the findings indicate there are different symptoms that could be the focus for assessment and treatment according to the ADHD presentation type.

A Survey on Privacy-Preserving Caching at Network Edge: Classification, Solutions, and Challenges

Caching content at the edge network is a popular and effective technique widely deployed to alleviate the burden of network backhaul, shorten service delay and improve service quality. However, there has been some controversy over privacy violations in caching content at the edge network. On the one hand, the multi-access open edge network provides an ideal entrance or interface for external attackers to obtain private data from edge caches by extracting sensitive information. On the other hand, privacy can be infringed on by curious edge caching providers through caching trace analysis targeting the achievement of better caching performance or higher profits. Therefore, an in-depth understanding of privacy issues in edge caching networks is vital and indispensable for creating a privacy-preserving caching service at the edge network. In this article, we are among the first to fill this gap by examining privacy-preserving techniques for caching content at the edge network. Firstly, we provide an introduction to the background of privacy-preserving edge caching (PPEC). Next, we summarize the key privacy issues and present a taxonomy for caching at the edge network from the perspective of private information. Additionally, we conduct a retrospective review of the state-of-the-art countermeasures against privacy leakage from content caching at the edge network. Finally, we conclude the survey and envision challenges for future research.

Research on Distributed Secure Storage Framework of Industrial Internet of Things Data Based on Blockchain

The conventional centralized Industrial Internet of Things (IIoT) framework is plagued by issues like subpar security performance and challenges related to storage expansion. This paper proposes a two-tier distributed secure storage framework based on blockchain for IIoT data. The authors first introduce the two-layer framework, which includes the edge network layer and the blockchain storage layer. The nodes in the edge network layer are classified into administrator nodes and ordinary nodes. It provides a lower latency network environment compared to cloud computing to preprocess raw industrial data. The blockchain storage layer provides storage space to keep data secure and traceable. Secondly, the authors propose a differentiated storage solution. Based on the timestamps of industrial data and the specific media access control (MAC) address, the Universally Unique Identifier (UUID) of the raw data is generated and uploaded to the blockchain for secure storage. Encrypt the corresponding raw data using the elliptic curve cryptography algorithm, and then upload it to InterPlanetary File System (IPFS) to expand the storage capacity of the blockchain. Deploy a smart contract on the blockchain to compare UUIDs for consistency in an automated, lightweight method to determine data integrity. Finally, we analyze the advantages brought by the integration of blockchain and IIoT. Additionally, the authors design comparative tests on different storage methods. The results prove that the security of this paper’s scheme is improved, and the storage performance is extended. Noteworthy enhancements include heightened throughput of data uploaded to the blockchain and minimized delay overhead.

PopFL: A scalable Federated Learning model in serverless edge computing integrating with dynamic pop-up network

With the rapid increase in the number of Internet-of-Things (IoT) devices, the massive volume of data creates significant challenges for traditional cloud-based solutions. These solutions often lead to high latency, increased operational costs, and limited scalability, making them unsuitable for real-time applications and resource-constrained environments. As a result, edge, and fog computing have emerged as viable alternatives, reducing latency and costs by processing data closer to its source. However, managing the flow of such vast and distributed data streams requires well-structured data pipelines to control the complete lifecycle—from data acquisition at the source to processing at the edge and fog layers, and finally storage and analytics in the cloud. To dynamically handle data analytics at varying distances from the source, often on heterogeneous hardware devices with collaborative learning techniques such as Federated Learning (FL). FL enables decentralized model training by leveraging the local data on Edge Devices (EDs), thereby preserving data privacy and reducing communication overhead with the cloud. However, FL faces critical challenges, including data heterogeneity, where the non-independent and identically distributed (non-IID) nature of data degrades model performance, and resource limitations on EDs, which lead to inefficiencies in training and biases in the aggregated models.To address these issues, we propose a novel FL solution, called Pop-Up Federated Learning (PopFL) in edge networks. This solution introduces hierarchical aggregation to reduce network congestion by distributing the aggregation tasks across multiple Fog Servers (FSs), rather than relying solely on centralized cloud aggregation. To further enhance participation and resource utilization at the edge, we incorporate the Stackelberg game model, which incentivizes EDs based on their contribution and resource availability. Additionally, PopFL employs a pop-up ad-hoc network for scalable and efficient communication between EDs and FSs, ensuring robust data transmission in dynamic network conditions. Extensive experiments conducted on three diverse datasets highlight the superior performance of PopFL compared to state-of-the-art FL techniques. The results show significant improvements in model accuracy, robustness, and fairness across various scenarios, effectively addressing the challenges of data heterogeneity and resource limitations. Through these innovations, PopFL paves the way for more reliable and efficient distributed learning systems, unlocking the full potential of FL in real-world applications where low latency and scalable solutions are critical.

Joint AI Inference and Target Tracking at Network Edge: A Hybrid Offline-Online Design for UAV-Enabled Network

Joint AI Inference and Target Tracking at Network Edge: A Hybrid Offline-Online Design for UAV-Enabled Network