Non-linear Equalization in 112 Gb/s PONs Using Kolmogorov-Arnold Networks
arXiv:2411.19631v11 citationsh-index: 6
Originality Incremental advance
AI Analysis
This addresses signal distortion in high-speed optical networks for telecommunications, but appears incremental as it applies a novel method to a known bottleneck.
The paper tackled non-linear equalization in 112 Gb/s PAM4 passive optical networks by using Kolmogorov-Arnold networks, achieving better performance than linear equalizers and convolutional neural networks with low computational complexity.
We investigate Kolmogorov-Arnold networks (KANs) for non-linear equalization of 112 Gb/s PAM4 passive optical networks (PONs). Using pruning and extensive hyperparameter search, we outperform linear equalizers and convolutional neural networks at low computational complexity.