Improvement of ranging precision in a FMCW LiDAR by digital Kalman Filtering

Due to the coherent detection, Frequency Modulated Continuous Wave (FMCW) LiDAR system can not only be immune to the background light or the Sun light, but also can measure the distance and velocity of a target simultaneously. However, its ranging precision is fundamentally limited by the intrinsic frequency noise of the laser diode, which arises from spontaneous emission and cannot be suppressed by conventional pre-distortion methods. In this work, we successfully suppressed this stochastic noise by modeling this stochastic noise as a Gaussian distribution and introducing a digital Kalman filtering approach. Experimental results on static targets show a significant precision improvement at least one order of magnitude by implement of KF algorithm, reducing the measurement uncertainty from 2.497 10−4 m to 1.034 10−5m on static point with 10,000 measurement data, and from 1.734 10−4 m to 1.628 10−5m on a static object with 3,000 measurement data, respectively. Future work will extend this framework to moving targets.