Blind Channel Equalization Using MCMA Algorithm with Adam Optimization
Armin SalimiKia, Saeed Mozaffari, Majid Ahmadi, Shahpour Alirezaee · 2025
This paper introduces Adam-MCMA, combining the Modified Constant Modulus Algorithm (MCMA) with Adam optimization for faster blind channel equalization. Leveraging Adam's adaptive updates, our approach significantly accelerates convergence compared to traditional methods. Simulation results for 16-QAM show Adam-MCMA achieved a lower final MSE (18.91 dB) than standard MCMA (-17.57 dB) with approximately$\mathbf{9 8. 6 \%}$fewer iterations and$\mathbf{9 6. 4 \%}$less processing time. While offering performance comparable to Multi Modulus Algorithm (MMA) ($\mathbf{- 1 9. 9 5 ~ d B), ~ A d a m - M C M A ~ c o n v e r g e s ~ o r d e r s ~ o f ~ m a g - ~}$nitude faster. This demonstrates Adam-MCMA's computational efficiency, making it highly suitable for applications requiring rapid equalization of high-order constellations.