Abstract
To address the pressing demand for ultrafast and high-precision microwave frequency measurement, we propose and experimentally demonstrate an integrated photonic scheme based on thin-film lithium niobate (TFLN) and frequency-to-time mapping (FTTM). The system incorporates an on-chip Mach–Zehnder modulator (MZM) alongside an electro-optically tunable micro-ring resonator (MRR), wherein nanosecond-scale voltage-driven scanning of the MRR facilitates a linear frequency-to-time mapping. To circumvent the resolution bottleneck inherent to the MRR linewidth, we develop an adaptive multi-tone frequency extraction algorithm rooted in Lorentzian lineshape superposition. This algorithm enables precise deconvolution and autonomous identification of overlapping multi-tone components. Experimental results validate the approach, demonstrating accurate frequency identification across the 2–12 GHz range. At a single-shot acquisition time of 500 ns, measurement errors are bounded by ±50 MHz for single-tone signals and ±100 MHz for multi-tone signals. Extending the measurement duration to 1 µs further reduces these errors to ±40 MHz and ±70 MHz, respectively. This work offers a promising route toward ultrafast, low-power, real-time microwave photonic monitoring and provides a scalable foundation for the full integration of high-speed frequency measurement systems.