Enhancing Transmission Network Reliability: AI-Powered Fault Diagnosis in Smart Grid

Surekha U. Ade, Ajinkya D. Salpe — Thu, 10 Sep 2026 00:00:00 +0000

A smart grid is a modern electricity system that addresses the problems of the traditional power grid. It becomes "smart" when all its parts work intelligently, including power generation, transmission, distribution, control, and smart meters. This paper mainly focuses on smart transmission and distribution to improve the performance of the smart grid. The system was tested under different fault conditions, such as single-line-to-ground faults and double-line-to-ground faults, to check its reliability. The performance was also verified using MATLAB Simulink simulations. This paper explains the cybersecurity of important devices used in Active Distribution Networks (ADNs), such as Phasor Measurement Units (PMUs), smart meters, Advanced Metering Infrastructure (AMI), and protection relays. It also discusses the main technologies that support ADNs, including microgrids, Electric Vehicles (EVs), the Internet of Things (IoT), and smart homes, and explains how they help improve the power system. Traditional Fault Detection and Diagnosis (FDD) methods are not enough for modern power systems because today's power networks are more complex and require fast, real-time monitoring. Deep Learning (DL) provides a better solution by automatically learning important patterns from large amounts of data. This helps detect and identify faults more accurately, quickly, and reliably. This situation increases the risk of cyberattacks because attackers can target the trusted third party (TTP), which may provide services to multiple smart grid operators. To improve security and reliability, the paper proposes a fault-tolerant multi-subset data aggregation scheme that continues to collect and process data correctly even if some parts of the system fail or are attacked. To solve the challenges of fault detection in modern power systems, this study introduces a new method that combines spectral graph theory with deep learning. This approach improves the speed and accuracy of fault detection, making the power system more reliable and efficient. Enhancing transmission network reliability ai-powered fault diagnosis in smart grid.

Ethical AI and Personal Data Protection for AI-Driven Mobile Network Operators

Cattleya Delmaire — Tue, 25 Aug 2026 00:00:00 +0000

Artificial intelligence now influences both network engineering and subscriber-facing decisions within mobile network operators (MNOs). The data used for these tasks—call detail records, device identifiers, radio measurements, location histories, and inferred profiles—can reveal far more about a person than the original service requires. This study uses a problem-oriented integrative review to examine four interconnected questions: which personal data are required for common MNO use cases; how privacy, fairness, and opacity risks arise; which legal and governance instruments are relevant; and which controls provide auditable evidence. Thailand’s Personal Data Protection Act provides the principal jurisdictional setting, while the EU General Data Protection Regulation and Artificial Intelligence Act are used where they may apply directly or offer a mature benchmark. The review identifies three recurring weaknesses in operator practice: treating pseudonymization as anonymization, treating location as an optional data field even though radio operation makes it inferable, and adopting ethical principles without assigning measurable controls or owners. In response, the study develops an MNO-specific four-layer governance framework that joins regulatory duties, ethical principles, technical safeguards, and lifecycle assurance under the NIST Govern–Map–Measure–Manage functions. Its distinctive contribution is a traceability chain from an AI use case and its data dependencies to risk, control, retained evidence, accountable ownership, and lifecycle action. A structured mobility-demand forecasting demonstration shows how the framework can record purpose, data granularity, privacy tests, performance thresholds, human responsibility, and post-deployment monitoring. The result is a decision-oriented reference for MNOs; it is not an empirical assessment of any operator or a substitute for jurisdiction-specific legal advice.

International Journal of Artificial Intelligence of Things (AIoT) in Communication Industry

Enhancing Transmission Network Reliability: AI-Powered Fault Diagnosis in Smart Grid

Ethical AI and Personal Data Protection for AI-Driven Mobile Network Operators