Kolmogorov-Arnold Networks outperform MLPs in FTN signaling detection

A new research paper compares the effectiveness of Multilayer Perceptrons (MLPs) and Kolmogorov-Arnold Networks (KANs) for detecting Faster-than-Nyquist (FTN) signaling. The study generated a large dataset of nearly four million labeled windows for FTN BPSK detection under additive white Gaussian noise. Results indicate that KANs significantly outperform MLPs, achieving a much lower bit error rate with substantially fewer parameters. AI

IMPACT Kolmogorov-Arnold Networks demonstrate superior efficiency and performance over traditional MLPs for signal detection tasks, potentially influencing future research in communications.

RANK_REASON The cluster contains a research paper detailing a comparative analysis of neural network architectures for a specific signal processing task. [lever_c_demoted from research: ic=1 ai=0.7]

Read on arXiv cs.LG →

• additive white Gaussian noise
• BCJR algorithm
• BPSK
• Faster-than-Nyquist and beyond: how to improve spectral efficiency by accepting interference
• Kolmogorov--Arnold Networks
• multilayer perceptron

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →