laser-phase-noise-and-line-width

simulate the relation between laser phase frequency noise and line width. But the simulation is wrong. the results are as follows: 1、the result dose not conform to the fact that Sf(f)*pi=LW. 2、High frequency noise is high。

仿真激光相频噪声和线宽的关系，仿真结果不正确：频率功率谱密度和线宽不满足Sf(f)*pi=LW；并且频率功率谱密度噪声在高频处抖动很大。

Please refer to the python version calculation

Some differences I noticed:

• In Line 29 you multiply the sampling rate but should rather divide the sampling rate. Same for Line 20.

About smoothing

I did not include smoothing in my code. It only makes sense physically when we have a certain length of fiber to de-correlate the laser.

Example:

fiber_length= 20e3 # 20 km fiber
group_index = 1.46
c = 2.998e8 # speed of light in m/s
Npts = int(1/(fiber_length*group_index/c)/(freq[2]-freq[1]))

In case of 20 km fiber used for de-correlation, number of points for moving average smoothing is around $N_{pts} = 1/(t_{fiber}*df) = 1/(L_{fiber}n_gdf/c) \approx 43 $.