Polarization independent frequency domain equalization (FDE) for chromatic dispersion (CD) compensation in PolMux coherent systems
View Patent ↗A method for the polarization independent frequency domain equalization (FDE) chromatic dispersion compensation on polarization multiplexing (POLMUX) coherent systems. Operationally, time domain signals are converted to frequency domain signals such that time domain convolution can be done as simple multiplications in the frequency domain. These frequency domain signals then converted back to time domain for subsequent use. The input signal size and FFT size are advantageously designed so that the output signals can be continuous with some overlap between two successive frames.
1. A method for polarization independent, frequency domain equalization-based chromatic dispersion compensation of a polarization multiplexed (PolMux) optical signal comprising:
applying the PolMux optical signal to an optical hybrid such that four output signals X′ I , X′ Q , Y′ I , Y′ Q are produced wherein X′ and Y′ represent two random polarizations orthogonal to one another and I and Q represent two orthogonal phases;
generating four digital electrical signals from the four output signals;
applying the four digital electrical signals to a digital signal processor wherein the followings are performed:
performing a frequency domain equalization-based chromatic dispersion compensation of the four digital electrical signals thereby producing two complex signals;
polarization demultiplexing the two complex signals; and
demodulating X and Y data from the two complex signals such that two signals Pol-X data and Pol-Y data are produced,
wherein said frequency domain equalization-based chromatic dispersion compensation includes:
converting the digital signals to parallel digital signals in a frame-by-frame manner;
forming a fast fourier transform (FFT) frame of size M from the converted frame;
transforming the FFT frame into a frequency domain;
multiplying the frequency domain frame with corresponding coefficients;
performing an inverse FFT to transform the frame into the time domain; and
adding the first N−1 output time domain signals from this frame with the last N−1 output time domain signals from the previous frame such that final output signals are produced.