Currently, the sub-6 GHz band remains the mainstream frequency choice for Internet of Things(IoT) applications due to its advantages in wide coverage and stable propagation. However, with the explosive growth of massive data transmission demands, the millimeter-wave (mmWave) band is emerging as a crucial development direction for future IoT communication, owing to its ultra-large bandwidth characteristics. During subway fires, the high-density water vapor environment generated by activated sprinkler systems causes complex attenuation of mmWave signals, significantly increasing the difficulty of channel analysis. This paper focuses on this emergency scenario to systematically investigate the impact of water vapor on mmWave channel characteristics in subway IoT systems. Through ray-tracing simulations conducted in a subway station environment, both the sub-6 GHz band (3~5 GHz) and the mmWave band(24~26 GHz) are examined. Key parameters including path loss, shadow fading, Rician K-factor, Root Mean Square(RMS) delay spread, and angular spread are extracted, with particular emphasis on analyzing their variation patterns under water vapor influence. The results demonstrate that the water vapor environment introduces significant additional attenuation to mmWave signals. Specifically, in the mmWave band, the path loss increases by approximately 12~18 dB on average compared to the sub-6 GHz band, the Rician K-factor decreases notably, and multipath components become more dispersed. Based on these findings, this paper proposes a channel modeling methodology and communication optimization strategies suitable for water-vapor-interference scenarios, providing theoretical foundations and technical references for mmWave communication design in subway emergency IoT systems.