Improved adaptive equalization with fixed step size CMA for DP-QPSK DWDM system

Advanced modulation format like dual-polarization quadrature phase-shift keying (DP-QPSK) along with digital coherent receiver is the key technology behind the current high-speed dense wavelength division multiplexing (DWDM) infrastructure. The adaptive equalization (AE), frequency offset estimation (FOE) and carrier phase estimation (CPE) are among the most important functions of the digital signal processing (DSP) unit of the digital coherent receiver and AE-FOE-CPE is the conventional process flow. The performance of AE is dependent on the step size of the constant modulus algorithm (CMA) and currently, the variable step size algorithm is in use to compensate for residual dispersion and PMD. In our recent work, we have presented that an improvement in the performance can be achieved if FOE is performed before AE. In this paper, we have extended our work further and exhibited with the help of numerical simulations that for the new process flow (FOE-AE-CPE) there is no need of complex variable step size approach in CMA algorithm and a fixed value of step size can also provide better system performance thereby making the process flow simpler in the DSP unit. We have also shown that under this scheme (FOE-AE-CPE + fixed step size CMA) an improved performance for an eight-channel 112 Gb/s DP-QPSK DWDM system with 50 GHz channel spacing can be achieved at relatively lesser launch power per channel compared to the conventional scheme. The tolerance of this new scheme is also analyzed for the linewidth of the local oscillator (LO).