Аннотация
Rapid urbanization has intensified air pollution, noise, and connectivity disparities in smart cities. This study integrates GPS-tagged IoT sensor networks and AI analytics to evaluate air quality, noise, and mobile internet performance in Astana. From October to November 2024, 840 noise measurements and PM₂.₅ data from fixed and mobile sensors were collected, along with 2G–5G network tests at ten locations. Data analysis employed Pearson correlation to examine links between meteorological factors and pollutants, and seven machine-learning models, including Random Forest, XGBoost, and Support Vector Regression, for multi-step pollutant forecasting. Spatiotemporal clustering and anomaly detection were applied to assess network reliability and identify coverage gaps. Findings show average noise levels of 73 dB, frequent PM₂.₅ exceedances of EPA thresholds, and a throughput–stability trade-off in 5G: Kcell/Activ reached peak speeds (up to 471 Mbps), while Altel/Tele2 maintained lower latency (≈29 ms), jitter, and no packet loss. The framework provides high-resolution insights to support targeted urban planning and policy.