A3PNet: Antijam and Accurate Antenna Parameters Measuring Network for Mobile Communication Base Station Using UAV

Whether the down-tilt angle of the mobile communication base station antenna is reasonable directly affects the coverage effect and communication quality of the whole network. Measurement of traditional antenna parameters mainly relies on engineers climbing the base station for manual measurements, confronting the dilemma of low measurement efficiency and a dangerous working environment. Moreover, due to the interference of the wind field in real unmanned aerial vehicle (UAV) measurement scenes, there is attitude disturbance during observation, which leads to a sharp increase in measurement error. To tackle these problems, we propose an antijam and accurate antenna parameters' measurement framework, named A(3)PNet, which aims to realize fully automatic mobile communication base station antenna parameters' measurement through UAV. Above all, for measuring in windy environment, an adaptive feature recovery module (AFRM) is presented to restore the antenna attitude while reducing the measurement error. Also, contour modulation strategy (CMS) is cooperated to smooth the deformation path and decouple the offset movement of the vertex to speed up the learning process. Go one step further, a general antenna down-tilt measurement module is proposed to replace the cumbersome numerical fitting analysis method, and it is embedded in the mainstream model to realize the task of down-tilt angle measurement. Extensive experiments over both the newly established Antenna-disturb dataset and the original Antenna-scapes dataset demonstrate the effectiveness of A(3)PNet. Specifically, it outperforms the state-of-the-art (SOTA) approaches by 2.40% on the Antenna-scapes dataset and 7.23% on the Antenna-disturb dataset, and meanwhile has 27.1 frames/s real-time processing speed, improving the measurement method of antenna parameters.