Incorporating Computer Networks and Machine Learning for AI-Driven Network Operators

This paper proposes an integrated approach combining computer networks and artificial neural networks to construct an intelligent network operator, functioning as an AI model. State information from computer networks is transformed into embedded vectors, enabling the operator to efficiently recognize different pieces of information and accurately output appropriate operations for the computer network at each step. The operator has undergone comprehensive testing, achieving a 100% accuracy rate, thus eliminating operational risks. Additionally, a simple computer network simulator is created and encapsulated into training and testing environment components, enabling automation of the data collection, training, and testing processes. This abstract outline the core contributions of the paper while highlighting the innovative methodology employed in the development and validation of the AI-based network operator.

Stochastic two-stage reliability-based Security Constrained Unit Commitment in smart grid environment

Urban rail transit operations have changed from a single line to a multiline network. The network operations have undergone quantitative and qualitative changes, and operations management is facing rapid internal and external changes. Using the Guangzhou Metro network operation practices as a case study, this paper first systematically analyzes the features of the operation scale, the proportion of urban mass transit, the surge in public demand, the security of the operational service capacity, reforms to the operation governance structure, the high-speed expansion of staff, and the development of knowledge and skills in urban mass transit networks. The paper then proposes several responses to the challenges that such networks face; for example, this paper proposes creating an innovative network operations management system, strengthening the management foundation, creating plans to promote operation capacity, enhancing security risk management and equipment quality management, developing a crisis public relations response, and applying information technology. In addition, this paper systematically describes countermeasures for multiline network operations, such as developing a management mechanism for network operations, actively cultivating staff skills, creating innovative transport organization models to enhance operational capacity, establishing a production service assessment system to continuously improve the level of transportation service, establishing a quantifiable safety assessment system and equipment quality index model, strengthening quality controls for security and equipment, extensively using information technology to ensure the health of the urban mass transit network operation, and implementing sustainable development measures.

Large scale IoT deployments for industrial, smart-city and many other applications require efficient mechanisms to monitor node health, and various performance influencing parameters of the underlying resource constrained network. Efficient and timely collection of such Network State Information (NSI) and its analytics help network operators ensure continuous operation, maintain end application QoS, and optimize the overall network performance through efficient network resource management. Considering the power and bandwidth constraints in these networks, it is imperative that the overheads inherently associated with NSI collection are minimized in transporting the monitored data comprising of different levels of demands and time criticality. We present CONETSI, a distributed algorithm and protocol for collecting NSI by forming node chains. Briefly, a node having NSI to be sent to a remote monitoring station opportunistically advertises for other potential sources of NSI. One of the neighboring nodes, chosen based on a distributed contention resolution logic running on the nodes, joins the chain. The exploration is then repeated by the joined nodes forming the NSI chain rooted at the originating node. The farthest node then collects NSI serially along the chain towards the originating node. CONETSI provides differential QoS to the NSI transfers by considering demand level and time criticality of the NSI data possessed by the network nodes. We evaluate theoretical bounds, backed by experimental results showing the performance gains that amortize the overheads associated with the exploration process. We also describe our software implementation in Contiki-NG operating system running the 6LoWPAN/RPL stack.

Recently, network operators around the world have encountered a paradox. On the one hand, they have made enormous investments to handle the tremendous growth of Internet media services triggered by the ubiquitous penetration of high-speed wireless networks and the increasing popularity of mobile devices. On the other hand, these investments have not delivered the expected revenue, leading to declining profitability. In response to this paradox, network function virtualization (NFV) has been proposed as a revolutionary technology to transform network architecture and operations. This emerging technology opens up significant opportunities to reduce the cost of operating media services. In this article, the authors present a survey on this emerging topic. Specifically, they analyze the forces driving the evolution of the network architecture, then describe a novel NFV-enabled media cloud system from two alternative perspectives: an end-to-end view and a layered view. They then illustrate the architectural changes' technical challenges, which range from optimal resource provision and request routing to automated optimization. Finally, they substantiate the NFV-enabled media cloud platform with proof-of-concept case studies. By demonstrating NFV's flexibility and effectiveness, they show that the cost of providing Internet media services by NFV-enabled systems could be substantially reduced in practice.

Purpose – This paper examines the influence of network operational digitalization and network operational automation on sustainable operations, with operational efficiency as a mediating variable. The study was conducted at PT XYZ, a telecommunications company in Indonesia, employing a quantitative approach through a survey of 161 respondents working in network operational units. Methodology/approach – The research instrument was developed based on indicators of digitalization, automation, operational efficiency, and sustainable operations. Data collected were analyzed using Structural Equation Modeling with Partial Least Square (SEM-PLS) through SmartPLS software. Findings – The results reveal that both digitalization and automation of network operations significantly influence operational efficiency. Furthermore, operational efficiency plays a mediating role in linking digitalization and automation to sustainable operations. The study also identifies a significant direct effect of network operational digitalization on sustainable operations. However, network operational automation does not yet demonstrate a direct influence on sustainability outcomes. These findings suggest that while automation contributes to efficiency, its role in sustainability is still evolving, reflecting the staged maturity of technological adoption within the telecommunications industry. Originality/Value – Theoretically, this study contributes to the development of an integrative model connecting digital transformation and sustainable operations in the telecommunications sector. Empirically, it highlights the unique asymmetry between digitalization and automation in driving sustainability, which has been underexplored in prior research. Practically, the results provide valuable insights for company management in formulating effective strategies to enhance digitalization and automation-based network operation systems as a foundation for long-term sustainable operations.

Analyzing network traffic is one of the fundamental tasks in both network operations and security incident analysis. Despite the immense efforts in workflow automation, an ample portion of the work still relies on manual data exploration and analytical insights by domain specialists. Current state-of-the-art network analysis tools provide high flexibility at the expense of usability and have a steep learning curve. Recent—often web-based—analytical tools emphasize interactive visualizations and provide simple user interfaces but only limited analytical support. This paper describes the tool that supports the analytical work of network and security operators. We introduce typical user tasks and requirements. We also present the filtering funnel metaphor for exploring packet capture (PCAP) files through visualizations of linked filter steps. We have created PCAPFunnel, a novel tool that improves the user experience and speeds up packet capture data analysis. The tool provides an overview of the communication, intuitive data filtering, and details of individual network nodes and connections between them. The qualitative usability study with nine domain experts confirmed the usability and usefulness of our approach for the initial data exploration in a wide range of tasks and usage scenarios, from educational purposes to exploratory network data analysis.

As sustainability becomes a requirement in network operations, accurately quantifying the carbon footprint of Internet traffic is essential. While energy-aware networking has seen significant attention, the ability to trace carbon emissions at the flow level remains an open challenge due to the complexity of shared infrastructure and lack of related telemetry. In this paper, we present a methodology to obtain fine-grained per-flow carbon emissions from traffic statistics. To this end, we collect power measurements from three switches under varying traffic conditions, including synthetic and real-world traces. From these measurements, we derive a regression model that accurately estimates instantaneous router power consumption using only throughput and packet rate counters--achieving >96% accuracy across all switch types and traces. We then extend this model to compute per-flow carbon emissions, distinguishing between consequential and attributional perspectives, and validate the results using traces from CAIDA and Google services. Our findings uncover actionable insights into how flow and network characteristics such as packet size, packet rate, and network utilization influence carbon cost. Finally, we propose feasible deployment strategies for flow-level carbon estimation frameworks. This work provides a foundational step towards enabling carbon-aware flow-level decision-making for users, applications, and network operators.

Creating, maintaining, and archiving accurate station metadata is critical for successful seismic network operations, data discovery, and research. The Station Information System (SIS) is a centralized repository of seismic station equipment inventory, instrument response, and site information of stations operated by regional seismic networks (RSNs) of the Advanced National Seismic System (ANSS; Filson and Arabasz, 2017). It has a web‐based user interface that enables the creation and manipulation of the corresponding metadata. The system can track the installation, maintenance, and removal of equipment from a site, which often results in the creation of new metadata epochs. SIS also computes the overall response, including gain, of a data channel by combining the responses of the underlying hardware components. SIS distributes this information in standard formats such as Federation of Digital Seismic Networks StationXML and dataless Standard for the Exchange of Earthquake Data. SIS can also be used to manage inventory of field equipment such as power, telemetry, or Global Positioning System antenna, as well as links to other site‐related repositories external to SIS to give the network operator the most complete view of a site and the overall network. This article summarizes the main features in SIS. We present its basic infrastructure, holdings, workflow, and how RSNs retrieve data from it. We also explain the reasoning to pursue one centralized repository and why it supports the goals of SIS and the ANSS. We demonstrate that by providing the ANSS network operator with a comprehensive site view, SIS enables the production of high‐quality metadata, a necessary prerequisite for producing high‐quality seismic data.

As the penetration level of embedded generation in the electricity distribution network rises, network operators' need for real time and geographically accurate information about the Low Voltage (LV) network increases. Smart meters and network monitors are the main providers of Smart data to the network operators. However, the Smart data need to be presented to the users in meaningful ways, that are helpful, relevant, and accurate. This paper proposes a model that identifies the granularity of data required by network operators and the best ways in which the Smart data need to be visualized in order to enhance network operation applications such as network planning, asset management, and fault management, leading to smarter grid operation.

Federated network operations become more and more important and the central challenge for Future Internet testbeds around the world. Federated network operations on Future Internet need to meet at least two conditions based on the general definition of federation [1]. First, multiple distributed network domains agree upon standards of network operations. Second, internal network operation of each domain is autonomous under its own control. Under the fundamental requirements, this paper presents a framework that federates and operates the autonomous distributed networks (ADNs) in the perspective of virtual network operations and end-user oriented network management on end-to-end network slices. In addition, we propose a virtual federated network (VFN) environment as a common federation interface between ADNs, associated with three key functionalities such as core network resource schema and representation, operational data exchanges, and end-to-end federated network operations. The performance consideration of this scheme includes the fast convergence of state updates on each ADN through VFN interface, as well as reliable and scalable operational dataset management of the federated Future Internet testbeds.

The advent of 5G technology marks a significant transformation in the telecommunications landscape, promising to deliver ultra-fast data speeds, minimal latency, and widespread connectivity that supports many new services and applications, including IoT, augmented reality, and autonomous systems. However, achieving the full potential of 5G networks involves overcoming substantial infrastructure challenges. Deploying and managing 5G's intricate architecture, encompassing the Radio Access Network (RAN), Core network, and Service-Based Interfaces (SBI)—requires a shift from traditional, labor-intensive methods to more advanced, automated processes. Infrastructure-as-Code (IaC) has emerged as a game-changing methodology that addresses these needs by codifying infrastructure setup and management into programmable scripts. This comprehensive review delves into how IaC facilitates the efficient deployment and management of 5G components, enhancing agility, consistency, and scalability. By automating resource provisioning, configuration, and updates, IaC empowers network operators to rapidly deploy infrastructure that meets the demanding requirements of 5G services. The paper discusses how IaC enables seamless integration and scaling of network functions, improves operational efficiency through reduced human error, and supports consistent infrastructure management across distributed environments. Key tools such as Terraform, Ansible, and Kubernetes are explored, showcasing their applications in orchestrating various layers of the 5G network, from RAN deployment to core network orchestration and SBI configurations. Despite its advantages, adopting IaC in 5G deployment comes with its challenges. These include managing the complexities inherent in multi-vendor environments, ensuring robust security for automated provisioning scripts, and maintaining compliance with regional regulatory standards. The paper explores these obstacles in-depth, offering insights into best practices and potential solutions to navigate these hurdles. This review highlights emerging trends and research opportunities, such as the potential for zero-touch automation, which leverages IaC for autonomous network operations with minimal human intervention. It also discusses the role of artificial intelligence (AI) and machine learning (ML) in predictive resource management and automated fault detection, proposing a future where AI-enhanced IaC can revolutionize the deployment and operation of 5G networks. By examining these elements, the review aims to provide a holistic view of the current state and future trajectory of IaC in 5G RAN, Core, and SBI deployment, setting the stage for ongoing innovation and development in the field.

This paper addresses network management and operations issues for WDM networks. New WDM components, such as the optical spectrum analyzer, P(/spl lambda/), inexpensive WDM multiplexers with flat passband, and optical amplifiers with uniform gain, will be important technologies to enable cost effective and robust operation of WDM networks.

Traffic visualization tools help network operators to maintain awareness of the status of a network, including anomalous activities. Unfortunately, the network operator may look away from the visualizer when beginning network forensics, such as launching a terminal application, logging into a server, and analyzing log files. Thus, the eyesight of the network operator will move from the visual screen even if valuable information is displayed. Our motivation is to develop the ability to use visualization tools as a network operation console. Whereas previous tools focused on outputting packet information, we herein extend the visualizer to accept inputting for operators to start their operations. Since little such software exists for our intent, we develop PACKTER, which is able to visualize traffic based on per-packet information in real time. We also extend PACKTER to have a function of negotiating to a network forensic system, which allows the operator to select an individual packet using a mouse, to start network forensics using a keyboard, and to receive results without looking away from the PACKTER viewer.

Sequential recommendation systems predict the next item that a user will interact with according to the sequential patterns inferred from historical interacted items. Benefiting from the self-attention mechanism, BERT has achieved success in sequential recommendation. Nevertheless, it is still challenging to leverage multiple types of information (e.g., item IDs, attributes, timestamps) effectively under the BERT framework. Existing methods fuse different types of information as the input of BERT by directly adding or concatenating their embedding vectors. This way of fusion ignores the differences of different types of information in sequential pattern inference, which can cause the mutual interference problem and hinder the effective utilization of multi-information. In this work, we propose a Dual Disentangled Attention (DDA) based BERT model, called DDA-BERT, for better leveraging multi-information. Our model performs disentangled information modeling from two aspects: 1) different views to describe items using content information (e.g., using IDs or using attributes) are modeled in different attention heads, respectively; 2) different sequential information (time intervals, relative position distances) modeling is disentangled from content information modeling while calculating the attention scores. As a result, DDA-BERT can effectively prevent the mutual interference problem to improve the prediction accuracy. Extensive experiments on three benchmark datasets show that DDA-BERT consistently outperforms the state-of-the-art baselines by up to 30%.