Tunable photonic compressive sensing of microwave signals based on chirp mixing.

We propose a tunable photonic compressive sensing (PCS) scheme based on chirp mixing, in which a linear frequency-modulated (LFM) signal is adopted to replace the pseudo-random binary sequence (PRBS), enabling tunable band-selective reconstruction-a capability absent in the PRBS-based PCS approach. The proposed system supports both ultra-wideband scanning across the entire Nyquist frequency range and narrow-bandwidth operation for improved reconstruction quality. Photonic-assisted frequency doubling is utilized to lower the bandwidth requirement of the LFM signal during chirp mixing. Furthermore, an integrated array structure, a frequency-selective joint frequency and angle-of-arrival (AOA) measurement PCS system is developed. Experimental results demonstrate frequency-selective recovery of microwave signals over a 5 GHz bandwidth at a compression ratio of 20, as well as effective joint frequency and AOA measurement within predefined detection bands. This work introduces a new, to the best of our knowledge, application-oriented and spectrum-focused dimension to PCS.