Intelligent Radio Planning and Connectivity Optimization for Underground Public Transportation Systems Using Distributed Antenna Networks

Abstract

Reliable wireless connectivity in underground public transportation can vary with train position because rolling stock modifies propagation paths and local interference conditions. This study proposes a transport-state-aware radio-planning framework for a three-node distributed antenna system (DAS) in a reference underground metro station. A three-dimensional 3.5 GHz reference station model, three operating states (empty station, train at platform, and train entering), a discrete set of 69 candidate antenna positions, and multi-objective placement/power optimization are combined to evaluate received power, SNIR, joint service coverage, and serving-area balance. The optimized configuration increased worst-state joint service coverage from 66.12% to 68.26%, improved worst-state P5 received power from −71.92 to −71.49 dBm, and reduced aggregate transmit power from 753.57 to 682.49 mW. The mean serving-area coefficient of variation decreased from 0.202 to 0.114. Threshold analysis showed that the optimized layout was advantageous at moderate and high SNIR requirements but not at a relaxed 5 dB criterion. The results support treating underground DAS deployment as a transport-state-aware infrastructure-planning problem rather than a static geometric-spacing problem.