A denoising method of all-fiber pulsed coherent Doppler LIDAR (CDL) is investigated. The goal is to enhance the signal-to-noise ratio (SNR) in the weak signal regime. Based on differential detection theory, the total noise expression of CDL with a dual-balanced detector is introduced and analyzed. The conclusion is drawn that the total noise can be acquired under the local oscillator laser exposure conditions by reasonable simplification. Using the actual measured data, the ratio of the standard deviation to the mean value of the total noise in each range gate is obtained and is up to approximately 11%. In order to suppress the jitter of the noise, an effective noise modeling by the trend of each range gate is developed. The feasibility of this method is verified by a long set of measured data. The spatial and temporal distribution of wind speed is illustrated with 400 pulses accumulation. Compared to the noise modeling of the tail, the detection range of wind speed using the proposed method can be improved by 35.3%.
Read full abstract