Hybrid low-complexity baud-rate sampling and adaptive equalization technique for 400 G/800 G coherent metropolitan area networks.

Baud-rate sampling and adaptive equalization (BSAE) techniques digitize and adaptively equalize received optical signals at the baud-rate and thus can concurrently reduce ADC and equalization power consumption compared with oversampling and fractionally-spaced adaptive equalization (OFAE) techniques operating at a higher sampling rate. Although suitable for short-reach intensity modulation-direct detection (IM-DD) based access networks or intra-datacenter networks, the BSAE techniques are generally regarded as not suitable for metropolitan area networks (MANs), owing to their limited capability to compensate for the large chromatic dispersion (CD) and in-phase/quadrature (IQ) skew distortions encountered in MANs. This constraint renders them inferior to the OFAE techniques able to compensate for arbitrary CD and IQ skew distortions. In this paper, we propose a BSAE technique combining a baud-rate frequency and time domain hybrid adaptive equalizer (AEQ) and Nyquist filtering for dynamic 400 G/800 G coherent MANs. Numerical simulations and experiments show that the BSAE technique can overcome the limitations and compensate for large CD and IQ skew distortions encountered in 400 G/800 G coherent MANs with spans up to 120 km. Furthermore, compared with the current state-of-the-art BSAE and OFAE techniques, the proposed BSAE technique can reduce the equalization computation complexity by about 40% and 80%, respectively, while maintaining a superior performance as the OFAE techniques. These merits make it a highly attractive solution for the dynamic 400 G/800 G coherent MANs.