Modeling and predicting the transmission efficiency of communication devices under joint noise and vibration disturbances

In complex environments such as industrial sites and rail transit, communication equipment often faces multi-source interference from mechanical vibration and structural noise, which seriously affects its signal quality and transmission stability. Although previous studies have explored the influence mechanism of a single interference source, there is still a lack of in-depth understanding and quantitative modeling of the coupled interference effect of vibration and noise. To this end, this paper builds an experimental platform based on the ESP32 Wi-Fi communication module, which includes controllable electromagnetic vibration and sound pressure loading, and collects communication performance indicators (RSSI, BER, throughput and delay) and synchronous physical disturbance data under different interference conditions. Through multivariate statistics and variance analysis methods, the interaction law between vibration frequency, amplitude and noise sound pressure level is revealed for the first time. It is found that the combination of the two under medium and high intensity conditions will cause significant nonlinear amplification effects, which will have a synergistic degradation effect on communication performance. The long short-term memory neural network (LSTM) is further introduced to construct a time series prediction model under multi-disturbance environment. The results show that the model has excellent fitting accuracy (R2 > 0.97) in RSSI and throughput prediction tasks, which is better than the comparison models such as SVM and polynomial regression, and has good feedforward control potential. The study also proposed communication anti-interference optimization suggestions and equipment structure improvement strategies suitable for industrial and rail scenarios, providing a theoretical basis and experimental support for the intelligent adaptive design of wireless communication systems in high-interference environments.