Optimized adaptation algorithms for low-complexity adaptive equalizers.

ABSTRACT

In order to address the high-power consummation issue of conventional multi-input and multi-output (MIMO) adaptive equalizer (AEQ) for short-reach coherent transmissions, several state-of-the-art low-complexity AEQs have been proposed. In our work, optimized adaptation algorithms for low-complexity real-valued (RV) AEQs with different structures are analyzed. Moreover, an approach to avoid introducing additional computational complexity due to the optimized adaptation process is presented here. The advantages of proposed optimized adaptation algorithms are experimentally demonstrated in a 25 Gbaud dual-polarization 16-quadrature-amplitude-modulation (DP-16QAM) back-to-back (BtB) intradyne system with an overall bandwidth of 14 GHz. Experimental results show that a similar performance as the conventional AEQ could be achieved by using proposed adaptation algorithms and reducing the number of multiplications with up to ∼65%.