Abstract The rapid diversification of mobile devices poses significant challenges for market analysis, application deployment, and system level design optimization. This study presents a specification driven segmentation framework that integrates cluster profiling with principal component analysis (PCA) based visualization to characterize heterogeneous mobile device ecosystems. Using a structured Kaggle telecommunications dataset encompassing hardware specifications, display characteristics, camera features, connectivity options, and pricing attributes, we apply unsupervised... The rapid diversification of mobile devices poses significant challenges for market analysis, application deployment, and system level design optimization. This study presents a specification driven segmentation framework that integrates cluster profiling with principal component analysis (PCA) based visualization to characterize heterogeneous mobile device ecosystems. Using a structured Kaggle telecommunications dataset encompassing hardware specifications, display characteristics, camera features, connectivity options, and pricing attributes, we apply unsupervised clustering to identify distinct device categories. Z-score standardization ensures feature comparability, while centroid based profiling captures group-level specification patterns. Dual visualization strategies are employed: radar plots enable intuitive multivariate comparison of normalized feature means across clusters, revealing interpretable trade,offs between performance, physical design, and economic positioning; two dimensional PCA projections validate cluster separability and expose dominant latent factors governing device differentiation, with explained variance ratios quantifying the reliability of reduced dimension representations. Results demonstrate clear segmentation into entry level, mid range, and feature rich device categories, with discriminant features including RAM, battery capacity, camera resolution, and processing power. Radar plots facilitate rapid qualitative assessment of cluster personalities, while PCA confirms mathematical distinctness among groups and identifies performance capability and multimedia features as primary differentiation axes. The framework offers a scalable, interpretable methodology for analyzing mobile device diversity, supporting data informed decision making for product positioning, adaptive application deployment, and market intelligence. By bridging traditional engineering perspectives with modern data science practices, this approach contributes to more responsive, user centric innovation in mobile and wireless system design, with implications for digital engineering and context aware mobile computing applications. Read More Read Less

Abstract This paper presents a novel feature engineering framework for time series classification of capacitive sensor data to distinguish between water and oil immersion conditions. Capacitive sensors are widely used in industrial monitoring and human machine interaction applications, yet accurately classifying short sequential signals remains challenging due to temporal dependencies and subtle pattern variations. The dataset consists of ten consecutive intensity measurements from sensor electrodes,... This paper presents a novel feature engineering framework for time series classification of capacitive sensor data to distinguish between water and oil immersion conditions. Capacitive sensors are widely used in industrial monitoring and human machine interaction applications, yet accurately classifying short sequential signals remains challenging due to temporal dependencies and subtle pattern variations. The dataset consists of ten consecutive intensity measurements from sensor electrodes, labeled according to immersion medium. While raw sequential values contain discriminative information, their high correlation and limited length make it difficult to extract meaningful temporal dynamics. To address this, we design a comprehensive feature set encompassing trend based, complexity-based, stability, and position specific descriptors. Key features include linear slope, first last difference, entropy, variance, stability, and change point counts, which collectively encode global trends, variability, and structural irregularities within each sequence. Experimental evaluation using Logistic Regression, Random Forest, SVM, and k-NN demonstrates that engineered features substantially enhance classification performance. Notably, combining raw and engineered features yields near perfect performance, achieving 99.93% accuracy and a ROC-AUC of 0.99999 with Logistic Regression, confirming that the proposed descriptors provide complementary discriminative power. Feature importance and coefficient analyses further reveal that temporal trend strength, entropy, and signal stability are critical in differentiating water from oil immersion, reflecting underlying dielectric property differences. The proposed framework balances predictive accuracy and interpretability, offering a scalable and transparent solution for short sequence binary sensor classification problems. Read More Read Less

Abstract This study characterized electromagnetic nano-network performance through packet level traffic analysis of 876 simulated transmission events within the Terahertz band (0.1-10 THz), offering critical insights for Internet of Nano Things (IoNT) development. Key findings revealed that traditional networking assumptions often fail at the nanoscale: packet size showed no significant correlation with latency (r=0.019, p=0.579), indicating propagation and processing delays dominate over serialization time. Similarly,... This study characterized electromagnetic nano-network performance through packet level traffic analysis of 876 simulated transmission events within the Terahertz band (0.1-10 THz), offering critical insights for Internet of Nano Things (IoNT) development. Key findings revealed that traditional networking assumptions often fail at the nanoscale: packet size showed no significant correlation with latency (r=0.019, p=0.579), indicating propagation and processing delays dominate over serialization time. Similarly, signal strength did not predict error rates (r=-0.009, p=0.792), suggesting environmental factors like molecular absorption or interference outweigh simple path loss models in THz communications. Bandwidth utilization demonstrated minimal discriminatory power between normal and anomalous traffic, with only a 6.1% mean difference lacking statistical significance (p=0.155), underscoring the need for multi feature machine learning approaches in nano network security rather than single metric thresholds. Protocol comparison highlighted distinct trade offs: TDMA achieved the lowest latency (2.717 ms) and highest bandwidth (5.764 Mbps), while CSMA minimized error rates (0.478%). These results suggest protocol selection must align with specific application requirements, such as latency sensitive healthcare monitoring versus error critical data transmission. While based on simulated data, this work establishes a foundational understanding of nanocommunication dynamics, emphasising that optimising electromagnetic nano networks requires embracing the unique physical constraints of the Terahertz band. Future research should prioritize experimental validation using physical testbeds and explore advanced multivariate models to enhance anomaly detection, ultimately enabling robust, scalable IoNT ecosystems for next generation 6G healthcare applications and beyond. Read More Read Less