Chunghwa Telecom (Taiwan)

Global Navigation Satellite Systems (GNSSs) observables, such as those of the Global Positioning System (GPS), are frequently affected by multipath effects that cause unpredictable signal interference at the receiver, posing significant challenges for accurate state estimation in complex environments with non-Gaussian noise or outliers. The traditional extended Kalman filter (EKF), based on the minimum mean square error (MMSE) criterion, assumes Gaussian noise distributions and exhibits degraded performance under non-Gaussian conditions. To overcome this limitation, the minimum error entropy (MEE) criterion was proposed to reduce random uncertainty in estimation error distributions; however, due to its translation invariance property, MEE may inadvertently increase bias when errors contain systematic offsets, leading to poor convergence. In contrast, the maximum correntropy criterion (MCC) concentrates the error probability density function (PDF) around zero, enabling effective entropy adjustment even in the presence of bias and achieving superior error convergence. This paper presents the centered error entropy (CEE) extended Kalman filter (CEE-EKF) that integrates the complementary merits of both MEE and MCC approaches to overcome their individual limitations. Experimental validation in complex nonlinear GPS environments with non-Gaussian noise demonstrates that the CEE-EKF significantly outperforms individual algorithms in noise suppression, particularly exhibiting enhanced robustness and accuracy when handling outliers. These results offer an effective approach to enhancing the reliability of GPS navigation in challenging real-world environments, and the algorithm can be readily extended to other GNSS applications.

As network infrastructures become more complex, determining whether a failure of network management command is caused by a device malfunction or a network black hole has become a major challenge for Communication Service Providers (CSPs). In this paper, we propose a Network Black hole Detection (NBD) method to detect whether a network black hole exists in the telecommunication network. By analyzing historical incident reports and largescale network traffic data, we design two sets of features-entropy-based and ratio-based-that capture traffic variations among neighboring devices surrounding a suspected faulty node. We use machine learning algorithms such as XGBoost (XGB) and Logistic Regression (LR) to train predictive models based on these features. These models are then evaluated using real-world incident data collected from 208 access-layer Gigabit Ethernet Switch (GESW) devices. Experimental results show that, compared with the traditional rule-based method and the LR model, XGB model achieves the highest prediction accuracy ($66.7 \%$). In addition, the traditional rule-based method achieves $61.7 \%$ accuracy using our custom-designed features, which is significantly higher than the $50 \%$ accuracy from random guessing. Our experimental results demonstrate that the proposed NBD system can effectively distinguish between device failures and network black holes, offering practical value for network fault diagnosis.

This paper presents an adaptive modulation method specifically for open all-photonics networks, which represent future optical transport architectures focusing on eliminating the need for optical-electrical-optical conversions. Unlike proprietary optical systems, open all-photonics networks lack built-in mechanisms for dynamically adjusting modulation formats in response to real-time changes in signal quality. This shortcoming makes them vulnerable to drops in performance due to variations in optical signal-to-noise ratio, attenuation, and link issues, especially in long-distance or multi-span settings. To address these challenges, we propose a method for real-time modulation reconfiguration based on telemetry inputs such as optical signal-to-noise ratio and forward error correction metrics. The proposed method is deployed and validated on a practical, 380 -kilometer open all-photonics network spanning from northern to southern Taiwan. The approach supports dynamic modulation adaptation across multi-vendor optical equipment using standardized, open network interfaces. Experimental results demonstrate that the proposed approach reduces the recovery time from several hours to less than two minutes under degraded OSNR conditions. These findings confirm the feasibility of implementing adaptive modulation in disaggregated photonic infrastructures and underscore the importance of real-world validation in the development of scalable and resilient all-photonics network systems.

Telecom network services are transforming to cloud-native network functions with greater flexibility and faster scaling capabilities. However, this enhanced flexibility brings new challenges for resource allocation and energy efficiency, especially in large-scale CNF deployments. We proposed a dynamic resource allocation algorithm based on Kubernetes pod autoscaling to optimize CNF utilization while maximizing energy efficiency. In addition, CNFs are scaled efficiently by adopting approaches integrating energy and performance monitoring, event-driven scaling, and resource optimization. According to the pilot site results, the dynamic resource allocation approach enhances the CNF energy efficiency ratio and increases the power consumption of worker nodes. The CNF energy efficiency ratio increased from an initial $83.9 \%$ to $90 \%$ with the dynamic resource allocation approach.

With the growth of quantum computing hardware technology, more developers are becoming involved in the design and development of quantum algorithms, leading to the creation of practical applications that align with real-world needs through the combination of quantum hardware and software. Among quantum algorithms, the most well-known include Grover’s algorithm and Shor’s algorithm. Grover’s algorithm can quickly provide search results, making it highly valuable for search-related applications. In light of this, this study primarily examines the theoretical usability of Grover’s algorithm and explores its limitations. Additionally, on the practical side, this research implements Grover’s algorithm using IBM Quantum Computing and the IBM Qiskit package. By comparing simulation results with those obtained from execution on real quantum computers, the study discusses the current limitations of quantum computing hardware technology.

We give a self–contained derivation that upgrades the Dirac–Born–Infeld+Chern–Simons action of a single Type–IIB D3–brane to a four–dimensional, quaternionic and PT –symmetric spacetime model with only two free parameters. A long–wavelength NS–NS two–form induces exactly two linear, SU(2)-valued deformations of the open–string metric, ϵ(t) e1Tμν and ϵ(r) e2Rμν, where ϵ(t) = ϵ0 cosωt and ϵ(r) = ϵ1/r play the role of geometric activators. With the minimal prescription P : xi 7→ −xi, T : t 7→ −t, i 7→ −i the full Dirac operator becomes pseudo–Hermitian and the metric remains PT –invariant. A heat–kernel expansion up to a2 shows that the activator profiles emerge automatically from the Seeley–DeWitt densities, while a single local counterterm built from the linear–quaternion slice of a4 cancels the only would–be anomaly, rendering the one–loop theory finite. The resulting spectral action predicts a narrow phenomenological window, |ϵ0| ≲ 10−7, |ϵ1| ≲ 101, already constrained by PLANCK 2018 CMB data, low–surface–brightness rotation curves, and present atom–interferometer limits. Forthcoming measurements with MAGIS–100, ELGAR and the Einstein Telescope can tighten these bounds by one–to–two orders of magnitude, providing a decisive test of the framework. Conceptually, the work closes the loop DBI(10D) → quaternionic (4D) → spectral dynamics → laboratory/astrophysical observables, and offers a minimal template for exploring higher–dimensional quantum structures in gravity with falsifiable predictions.

In real-time video streaming systems, NAT(Network Address Translator) traversal technology is very important, as it can significantly reduce system operating bandwidth costs.In past studies, four NAT classifications and nine NAT classifications were proposed for NAT behavior. At the same time, STUN, TURN, P2P, ICE, ALG, C2C, etc. protocols were proposed to implement NAT traversal. Although the above NAT classifications and traversal methods have effectively solved the current problems, there are still unknown NAT classifications waiting to be discovered and traversed.This paper discovers a new NAT behavior for the first time and proposes an effective NAT traversal method. This NAT behavior is called Port Knocking Sensitive (PKS) NAT. The PKS NAT will block the mapping port for external active invading packets, which helps prevent DDoS(Distributed Denial of service) attacks. Regardless, this newly discovered NAT behavior will render existing NAT traversal techniques inoperable.This paper also proposes a Synchronous Client to Client(SC2C) NAT traversal protocol technology for the first time. This protocol first requires both parties to test and predict the Mapping Port of their own NAT and notify the other party. The server then coordinates both parties to send NAT traversal packets synchronously, which can successfully traverse this newly discovered NAT. The experimental results prove that this new NAT does block the existing NAT traversal protocol, and also confirm that the SC2C NAT traversal protocol proposed in this paper can indeed effectively traverse this newly discovered NAT.

Since the amazing effect of ChatGPT has been recognized by the industry, towards integrating with employees’ workflow to create more business applications. At the same time, many artificial intelligence scientists are actively looking for methods and platforms that are highly efficient and can quickly embed LLM and gain business value quickly. This study proposes two important innovations. First, it uses historical operating data to quickly and automatically generate LLMs in the company’s exclusive domain. Second, to quickly embed those models in ontology maps that to generate thinking and decision-making documents (as AIGC). In the paper, first step is use auto-insight engine to transfer input data from operational datasets. Secondly, to adjust parameters the training in deep learning framework to output models. Finally, integrate more API that they like conversation tools or other services. It can automatically generate and output high-precision event determination results in real time and can automatic notification and complete the processing flow that integrate daily report for frontline partners. In future, it will be able to develop more follow-up applications, such as artificial intelligence assistant for the security control center partner, which can determine the high level of daily incidents and automatically complete high-complexity daily operation procedures in a closed-loop security environment, and moderately connect high-level voice assistants to achieve important indicators such as health care, efficiency improvement and cost reduction.

This paper presents an adaptive proportional-derivative adjuster (APDA) integrated into a notch filter for processing ECG signals in wearable smart clothing. The APDA filter dynamically adjusts parameters - such as gain and cutoff frequency - in real time to mitigate environmental interference, including 60 Hz noise and electromyographic signals common in workplaces and fitness centres. Unlike traditional fixed-parameter filters, this adaptive method minimises noise and preserves ECG clarity under varying conditions. A notable contribution of this study is optimising the APDA algorithm for resource-constrained microcontrollers, reducing both memory requirements and processing power. This streamlined computation lowers complexity and extends battery life, making it suitable for continuous, long-term monitoring. Experimental results from a smart clothing prototype validate the filter's effectiveness in suppressing noise while maintaining signal integrity, offering an energy-efficient and cost-effective solution for wearable healthcare applications. This advancement addresses the limitations of conventional designs and shows promising potential for future clinical applications.