Hardware efficient implementation of decision feedback phase estimation for an ultra high speed coherent receiver

摘要

A method for estimating the phase of a modulated complex optical signal, according to which, the phase of the signal is isolated and the complex amplitude of the signal is digitized by a block of P samples. An adaptive filter minimizes phase-error between the phase of the received symbol and the phase of a sample, rotated by a phase correction factor, by iteratively performing a Block-Wise Phase LMS estimation on the samples using a step size parameter. During each iteration, the resulting correction factor consists of the sum of the estimated errors multiplied by this parameter, and the correction factor of previous iteration, which is updated, until obtaining the final correction factor from the last iteration. The samples constituting the received signal are recovered by adding the most updated correction factor to the phase of each sample and performing a decision regarding the phase of each sample.

主权项

1. A method for estimating a phase of a modulated complex carrier signal transmitted by a laser source to a coherent receiver, via an optical channel, comprising:
a) isolating the phase of said modulated complex carrier signal, in a phase plane of the modulated complex carrier signal; b) determining a block of P samples for digitizing a complex amplitude of said modulated complex carrier signal, according to a rate of phase variations of said optical channel and an amount of noise in said optical channel; c) determining a digitized phase-error being a difference between a phase of a digitized decoded symbol received by said coherent receiver and a phase of a sample being the digitized complex amplitude detected by said coherent receiver, rotated by a phase correction factor on a complex plane; d) constructing an adaptive filter to minimize said digitized phase-error, by:
d.1) iteratively performing a Block-Wise Phase Least Mean Square (LMS) estimation on said P samples according to a step size parameter, determined by a desired convergence rate, while during each iteration, the resulting correction factor consists of a sum of estimated errors of the P samples in a current iteration multiplied by said step size parameter, and a correction factor of previous iteration;d.2) updating the resulting correction factor during each iteration until obtaining a final correction factor from a last iteration; andd.3) recovering the P samples constituting a received signal by adding the most updated correction factor to the phase of each sample removing 2π cycles from the resulting phase and performing a decision regarding the phase of each sample.