“…In particular, reflectometer-based distributed OFS provide a huge amount of sensing points over many kilometers of range [ 1 , 2 , 3 , 4 ]. Conventional signal processing methods for such distributed OFS include optical pulse coding [ 5 , 6 ], Fourier transform [ 7 , 8 , 9 ] or wavelet transform processing [ 10 , 11 ]. In recent years, however, more and more advanced digital signal processing concepts, for example based on image processing [ 12 ] and machine learning [ 13 ], have been applied to further enhance the performance of distributed OFS.…”