Deep neural network-assisted high-accuracy microwave instantaneous frequency measurement with a photonic scanning receiver
Yuewen Zhou, Fangzheng Zhang, Jingzhan Shi, Shilong Pan
Abstract
A microwave instantaneous frequency measurement system with a photonic scanning receiver is proposed in which deep neural network (DNN)-assisted frequency estimation is used to deal with the system defects and improve the accuracy. The system performs frequency-to-time mapping by optical-domain frequency scanning and electrical-domain intermediate frequency envelop detection. Thanks to the optical frequency multiplication, the system can measure high frequency signals in a large spectral range. The DNN establishes an accurate mapping between the digital samples and real frequencies, based on which high-accuracy measurement is achieved. The measurement of signals from 43 to 52 GHz is experimentally demonstrated. Compared with the direct measurements, the DNN-assisted method achieves obviously reduced average errors of about 3.2 MHz.