Performance enhancement of microwave photonic-filter-based interrogation system using machine-learning algorithm

Microwave photonic filter (MPF) based interrogation solutions have attracted considerable research interest, through the sensing information conversion from the optical domain to the microwave domain and high-resolution electrical spectrum analyzing processing techniques. In order to overcome the trade-off between measurement accuracy and interrogation speed existing in the traditional direct-detection method, an efficient machine learning algorithm is introduced into the MPF-based interrogation system. Compared with the traditional direct-detection method, the proposed method can achieve better measurement accuracy under the sparsely sampled frequency response, whilst the interrogation speed is greatly improved. In addition, the well-trained model has strong adaptability to the amplitude variation of the microwave photonic filtering response.