Experimental Demonstration Of Adaptive Vff-Rls-Fde For Long-Distance Mode-Division Multiplexed Transmission

We experimentally demonstrate adaptive variable-forgetting-factor (VFF) recursive-least-square frequency-domain equalization (RLS-FDE) for mode-division multiplexing. The VFF-RLS-FDE algorithm improves convergence speed as transmission distance increases. For MDM transmission over a 1,000-km few-mode fiber, the convergence speed was increased by 18.7 times in comparison with LMS-FDE. In the meantime, the convergence error performance of VFF-RLS is also much superior to LMS while fixed-forgetting-factor RLS has a 0.2 dB penalty compared with LMS. The proposed VFF-RLS algorithm achieves better performances than conventional RLS in terms of convergence speed and convergence error