基于端到端优化的可编程微波光子滤波器

Key Points

• The research aims to optimize the performance of microwave photonic filters using an end-to-end approach. It focuses on managing the complexities of the weighted delay structure to mitigate distortion.
• Developed an end-to-end optimization strategy for the microwave photonic filter as a black box system.
• Monitored the output spectral shapes in real time to adjust the filter coefficients dynamically.
• Calculated discrepancies between actual and ideal tap weights to refine filter performance.
• Achieved a spectral reconstruction accuracy of 0.05 dB.
• Realized a high out-of-band rejection ratio of 47 dB for the RF low-pass filter.