Reconfigurable Fiber Random Laser for Computational Super-Resolution Spectroscopy

High-resolution spectral measurement is essential for revealing the intrinsic material and structural properties, with broad applications in gas detection, environmental monitoring, biosensing, and materials analysis. Conventional strategies for enhancing spectral resolution typically focus on improving spectrometer hardware or employing sophisticated computational techniques, but they often face trade-offs between the resolution and bandwidth, or the high cost and measurement complexity. Super-resolution spectroscopy enabled by random lasers recently emerged as a promising strategy, yet existing implementations suffering from uncontrolled sampling signals and inefficient data acquisition. Here, we propose a reconfigurable fiber random laser that enables computational super-resolution spectroscopy. By integrating a multimode-single mode fiber filtering structure with controllable input wavefront modulation inside the laser, the system generates narrow-linewidth, sparse, and randomized emission modes, enabling flexible and efficient one-path spectroscopy without relying on the extra reference path. Our experiment compares pre-recorded reference modes with selectively reloaded sampling modes to reconstruct high-resolution spectra. Experimental results demonstrate a 2.2-fold enhancement in spectral resolution compared to the baseline spectrometer, which originally had 70 pm resolution. This work offers a versatile and low-cost solution for super-resolution spectroscopy, with the significant potential for a wide range of optical sensing applications.