Current Issue : October-December Volume : 2026 Issue Number : 4 Articles : 5 Articles
Customer churn remains a major challenge in the telecommunications industry, where retaining existing customers is signi7cantly more cost-e8ective than acquiring new ones. This study uses the publicly available IBM Telco Customer Churn dataset consisting of 5634 preprocessed records to develop a predictive model for customer churn classi7cation. The dataset was cleaned and prepared through preprocessing steps including handling missing values and encoding categorical variables to ensure suitability for machine learning analysis. Four classical machine learning models—decision tree, random forest, XGBoost, and logistic regression—were selected due to their proven e8ectiveness in classi7cation tasks, interpretability, and strong performance in structured tabular datasets commonly used in churn prediction studies. These models were evaluated using standard performance metrics, including accuracy, precision, recall, F1-score, and ROC-AUC. To enhance predictive performance and stability, a tuned approach was applied. The random forest achieved an accuracy of 84.73%, precision of 85.20% for churn, recall of 84.07% for churn, F1-score of 84.62%, and ROC-AUC of 93.86%. The results demonstrate that combining well-tuned classical machine learning models can produce reliable and robust churn prediction performance. This study contributes by providing a systematic evaluation of classical models on a benchmark dataset with standardized preprocessing and comprehensive performance analysis for telecom churn prediction....
We use pre‐existing fiber‐optic telecommunications fibers in Istanbul, Türkiye, to generate a seismic velocity model of the subsurface down to 100 m depth, and estimate site‐amplification in the region. We collect ambient noise, predominantly from urban traffic, along the 8 km segment fiber, extract Rayleigh wave dispersion curves, and use a trans‐dimensional Monte‐Carlo algorithm to estimate shear wave velocities along the profile that is consistent with local expectations of geology. We then estimate site‐amplification in three different ways: (a) using 1D estimates drawn from the velocity model, (b) through 2D earthquake simulations across the model and (c) by comparing to real earthquake data, which were dominated by the Kahramanmaraş sequence in 2023. All three estimates of site response roughly agree in both spatial distribution and amplification factor, and the study demonstrates the efficacy of using telecommunications fibers in urban environments for seismic hazard estimates....
An efficient and mathematically sound method for simulating the distributions of telecommunication signals is the star-like semigroup convolution model. A rigorous operator-theoretic measure of signal strength is provided by the (S 3I), which is obtained via semigroup transformations. This study combines empirical field data with the mathematical tools of spectral convolution and star-like transformation semigroups to create a corrective-predictive model for the functioning of telecommunication networks in Gombe State, Nigeria. A poll on signal strength and Quality of Experience (QoE) was conducted using an 11-point star-like scale with 615,510 respondents spread over 11 Local Government Areas (LGAs) and three major providers (MTN, GLO, and Airtel). A star-like signal spectral index (S 3I) that generalizes empirical RSS distributions under a continuous semicharacter model was derived by analyzing the data using both formal statistical measures and semigroup-based spectral transformations. Semigroup theory is used with actual telecom data to provide a mathematically sound but practically useful model of network resilience.While GLO urgently needs infrastructure densification, MTN is demonstrated to dominate across LGAs, and Airtel exhibits localized strengths in Billiri and Shongom. Regulators such as the Nigerian Communications Commission are urged to use (S 3I) as an impartial performance metric. The model guarantees that transformational semigroups can direct predictive modeling of large-scale communication systems and can be extended to additional Nigerian states and next-generation (5G) networks....
Mode coupling in multimode step-index polymer optical fibers (SI POFs) plays a critical role in determining signal integrity and bandwidth performance in optical communication systems. It originates from intrinsic random perturbations that influence power distribution among propagating modes, making accurate prediction of steady-state distributions (SSDs) essential for reliable system design. In this work, we model mode coupling as a stochastic process using the Langevin equation, incorporating simulated Langevin forces to numerically evaluate modal power evolution and steady-state behavior. The proposed approach demonstrates strong agreement with previously reported experimental results, validating its capability to capture energy redistribution mechanisms induced by fiber imperfections. From a telecommunications perspective, the model provides valuable insights into modal dispersion, bandwidth limitations, and signal degradation in SI POFbased links. These results establish a robust and efficient framework for analyzing and optimizing multimode SI POFs, supporting their application in high-speed data transmission and short-reach optical communication networks....
Telecommunication networks are highly intricate, with numerous interacting components influenced by various known and unknown factors. Among these, Radio Access Networks (RANs) play a critical role in wireless communication. Unraveling causal relationships within such complex systems is essential for their improvement and optimization through statistical and data analysis techniques. This study pioneers the application of Transfer Entropy (TE) and Granger Causality (GC) to real-world telecommunication networks, marking the first exploration of these methods in this context. We assess their effectiveness in identifying relationships among key network attributes, offering insights into potential system optimization in practical scenarios. Furthermore, our research extends to a comparative analysis of TE and GC under varying network load conditions, utilizing live data collected from multiple base stations to uncover common patterns across identified network attributes....
Loading....