Similar to bulk Hi-C data, the frequency distribution of single-cell Hi-C data also adheres to a power law distribution in relation to the genomic distance between chromatin interaction endpoints. In light of this, we introduce an innovative normalization approach for single-cell Hi-C data that capitalizes on this power-law distribution. An extensive comparative study, employing three publicly accessible single-cell Hi-C datasets, underscores the robustness of PLNorm, critically evaluated against established normalization techniques including BandNorm, scVI-3D, and scHiCNorm. A diverse range of metrics, including changes in cell similarities, cell embeddings, and scalability, were utilized in the assessment. The results highlight the distinct advantages of PLNorm: it not only mitigates biases but also adeptly preserves cell-type information, enhances the precision of clustering outcomes, and demonstrates impressive scalability, making it a prime choice for large-scale data analysis. PLNorm is available at https://github.com/bignetworks2019/PLNorm/.