Experimental Demonstration of Multi-Target Ranging Using Parallel Photonic Reservoir Computing With Enhanced Parameter Robustness in Chaotic Laser System

To tackle the critical challenge of traditional synchronized chaotic lidar systems, which demand exacting parameter alignment between drive and response lasers, this study introduces a pioneering multi-target ranging approach leveraging parallel photonic reservoir computing. By establishing a three-channel distributed feedback (DFB) laser network, we seamlessly integrate mutually coupled chaotic dynamical systems with delayed feedback photonic reservoirs, thereby crafting a robust chaotic synchronization prediction learning mechanism resilient to parameter variations. Empirical findings reveal that, even under substantial parameter discrepancies in the drive-response system, such as a 100% bias current disparity and a 0.2 nm wavelength shift, the trained reservoir sustains chaotic lag synchronization with a normalized mean square error (NMSE) below 0.1, significantly surpassing the stringent device uniformity constraints of conventional synchronization techniques. Employing Hilbert transform phase demodulation, we adeptly accomplish simultaneous ranging of three targets. The system demonstrates an absolute ranging error of no more than 22.1 mm and a relative error under 0.84% for targets within the 1–5 m range, marking a precision enhancement exceeding 30% over traditional synchronized chaotic radar systems previously documented. This investigation is the first to experimentally substantiate the augmentative impact of photonic reservoirs on multi-channel chaotic radar systems, offering a novel theoretical framework and technological trajectory for the advancement of high-precision laser ranging systems in intricate environments.