Skip to Main Content
We experimentally investigate the performance of the signal power autocorrelation-based method for chromatic dispersion (CD) estimation in a polarization-multiplexing quadrature phase-shift keying (PM-QPSK) 100G coherent optical system conveying optical channel transport unit level-4 (OTU4) frames. It is shown that the typical laboratory setup, where the signal components are generated from delayed versions of the same pseudorandom binary sequence (PRBS), is inadequate for experimental validation because of artifacts in the signal auto-correlation function. This problem is circumvented by the use of a commercial line card transporting independent data sequences. The algorithm is used to estimate accumulated CD values from 0 to 50 000 ps/nm under up to 80-ps differential group delay (DGD). We further evaluate its convergence properties in terms of the length of the sample vector required for correct CD estimation and address the hardware resource requirements. The CD-shifted version of the algorithm yielded a maximum estimation error of 186 ps/nm in all tested conditions.