Joint decoupling method of Mach-Zehnder modulation index and local oscillator power for improving range profile detection sensitivity of FMCW coherent LiDAR

Frequency modulated continuous wave (FMCW) coherent LiDAR based on a Mach-Zehnder modulator suffers from coupled sideband and noise power effects that limit the system signal-to-noise ratio (SNR) and detection sensitivity. The modulation index inherently trades off sideband power against spectral purity, while the local oscillator power dictates the transition of the dominant noise source; these two mechanisms are also nonlinearly coupled, rendering single-parameter optimization fundamentally insufficient. A comprehensive effective SNR model is therefore established that rigorously couples the MZM sideband characteristics with multi-regime noise evolution, based on which a globally convergent joint optimization framework is developed to precisely co-optimize the modulation index and local oscillator power. The validity of the method is verified through experiments on one-dimensional range profiles of typical spatial targets and the range-time spectrum of the micro-motion target. Experimental results show that, under the optimal modulation index, the goodness of fit between the measured SNR and the theoretical model reaches 0.979. The results of joint optimization are better than single-parameter optimization. Compared with the single optimization state, the proposed method improves the range profile detection sensitivity by an average of approximately 4.93 dB, with a maximum improvement of 7.1 dB, and enhances the range-time spectrum detection sensitivity by an average of approximately 7.735 dB, with a maximum improvement of 11.64 dB. This research can provide an effective theoretical basis and algorithmic support for the design and performance optimization of FMCW coherent LiDAR systems.