Superimposed training and digital filtering coherent optical receivers
View Patent ↗In a polarization-multiplexing optical communications system, a method and system for coherent reception of polarization-multiplexed optical communications. Training and data sequences are recovered from a plurality of orthogonally polarized signals, such that training and data sequences may be used on a single channel without loss of spectral efficiency. This is accomplished by estimating a channel response for each orthogonally polarized signal based on the mean of the signal and distinguishing between a data sequence and a training sequence in each orthogonally polarized signal. Such distinguishing is accomplished according to an iterative linear minimum mean-square equalization based on the respective channel response estimate.
1. A polarization multiplexing, optical receiver comprising:
a training/data recovery module configured to extract superimposed training signals and data signals from a plurality of orthogonally polarized signals such that training signals and data signals may be used on a single channel without loss of spectral efficiency, comprising:
a channel response estimation module which estimates a channel response based on the mean of each orthogonally polarized signal; and
a linear minimum mean-square error equalization module configured to distinguish a data sequence from a training sequence in each orthogonally polarized signal based on the respective channel response estimate;
wherein said channel response estimation includes a channel estimation responsive to Ĥ k =(C k H C k ) −1 C k H {circumflex over (D)} k where Ĥ k is the vector of the channel estimates, {circumflex over (D)} k is the vector of the {circumflex over (d)} mk 's. {circumflex over (d)} mk is a function of a received signal and is based on
d
^
mk
=
1
T
∑
n
=
0
T
-
1
y
(
n
)
ⅇ
-
jα
mk
n
.
where H shown as a superscript of C k indicates that C k H is a Hermitian transpose of C k , and an iterative updating of a multi-polarization channel keeps repeating until reaching the point where the incremental difference between two successive estimates of the multi-polarization channel reaches a predefined threshold.
2. The receiver of claim 1 , further comprising a polarization beam splitter configured to split a received optical signal into two orthogonally polarized signals.
3. The receiver of claim 2 , wherein the CD compensation modules are configured to compensate for second-order CD effects in the two orthogonally polarized signals.
4. The receiver of claim 2 , wherein the CD compensation modules comprise a plurality of all-pass infinite impulse response filters.
5. The receiver of claim 1 , further comprising a plurality of chromatic dispersion (CD) compensation modules.
6. The receiver of claim 1 , further comprising a plurality of finite impulse response filters configured to remove polarization-dependent impairments (PDIs) from each orthogonally polarized signal.
7. The receiver of claim 6 , wherein the finite impulse response filters are updated to reflect the time-varying PDIs.
8. A method for coherent reception of polarization-multiplexed optical communications, comprising:
recovering superimposed training and data sequences from a plurality of orthogonally polarized signals, such that training and data sequences may be used on a single channel without loss of spectral efficiency, by:
estimating a channel response for each orthogonally polarized signal based on the mean of the signal; and
distinguishing between a data sequence and a training sequence in each orthogonally polarized signal according to iterative linear minimum mean-square
equalization based on the respective channel response estimate;
wherein said channel response estimate includes a channel estimation responsive to Ĥ k =(C k H C k ) −1 C k H {circumflex over (D)} k where Ĥ k is the vector of the channel estimates, {circumflex over (D)} k is the vector of the {circumflex over (d)} mk 's. {circumflex over (d)} mk a function of a received signal and is based on
d
^
mk
=
1
T
∑
n
=
0
T
-
1
y
(
n
)
ⅇ
-
jα
mk
n
.
where H shown as a superscript of C k indicates that C k H is a Hermitian transpose of C k , and an iterative updating of a multi-polarization channel keeps repeating until reaching the point where the incremental difference between two successive estimates of the multi-polarization channel reaches a predefined threshold.
9. The method of claim 8 , further comprising the step of splitting a received signal into two orthogonally polarized signals.
10. The method of claim 9 , wherein said step of compensating includes compensating for second-order CD effects.
11. The method of claim 9 , further comprising the step of updating the PDI compensation to reflect time-varying PDIs.
12. The method of claim 8 , further comprising the step of compensating for chromatic dispersion (CD) effects.
13. The method of claim 8 , further comprising the step of compensating for polarization-dependent impairments (PDIs) in each orthogonally polarized signals.