Scaling with Ranked Subsampling (SRS) is a normalization method for ecological count data (see Beule & Karlovsky, 2020).
SRS is implemented as the 'SRS' R package and the 'q2-srs' QIIME 2 plugin (see Heidrich et al., 2021). Refer to their particular repositories for more specific information:
You can also run SRS natively in Python using SRS.py. Check the usage example.