Wavelength transmission system and method using 3-dimensional LDPC-coded modulation
View Patent ↗A transmitter and method include a LDPC encoder configured to encode source data, and a mapper configured to generate three coordinates in accordance with a 3D signal constellation where the coordinates include an amplitude coordinate and two phase coordinates. A laser source is modulated in accordance with each of the three coordinates to provide a transmission signal. A receiver, includes a demapper receives an input signal from three branches to demap the input signal using a three-dimensional signal constellation having three coordinates. The three branches include a direct detection branch, and two coherent detection branches such that the direct detection branch detects an amplitude coordinate of the input signal and the two coherent detection branches detect in-phase and quadrature coordinates of the input signal. A bit prediction module and at least one LDPC decoder are configured to iteratively decode bits by feeding back extrinsic LLRs to the demapper.
1. A transmitter, comprising:
at least one low density parity check (LDPC) encoder configured to encode source data;
a mapper configured to generate three coordinates in accordance with a three-dimensional (3D) signal constellation where the coordinates include an amplitude coordinate and two phase coordinates; and
a power source outputting signals, where the signals are modulated in accordance with each of the three coordinates to provide a transmission signal.
2. The transmitter as recited in claim 1 , wherein the at least one low density parity check (LDPC) encoder includes a plurality of encoders and the transmitter further comprises an interleaver to interleave encoded data source channels.
3. The transmitter as recited in claim 1 , wherein the three-dimensional signal constellation coordinates include respective coefficients defined by orthonormal basis functions Φ 1 , Φ 2 , and Φ 3 , where T is a symbol duration and 0<t<T where
Φ
1
(
t
)
=
1
T
sin
(
2
π
t
T
)
,
Φ
2
(
t
)
=
1
T
cos
(
2
π
t
T
)
,
Φ
3
(
t
)
=
1
T
.
4. The transmitter as recited in claim 1 , wherein the three-dimensional signal constellation is formed using identical two-dimensional (2D) signal-constellation points constructed on equally spaced parallel layers.
5. The transmitter as recited in claim 4 , wherein the equally spaced parallel layers are indexed by the amplitude coefficients and the 2D signal-coordinate points represent in-phase and quadrature coordinates.
6. A modulation method, comprising:
encoding a signal using low density parity checking code;
mapping the encoded signal to three coordinates using a three-dimensional constellation, the three coordinates including one associated with amplitude and two associated with phase; and
modulating the encoded signal for transmission in accordance with said three coordinates.
7. The method as recited in claim 6 , further comprising interleaving data from a plurality of encoders for input to a mapper for mapping the encoded signals.
8. The method as recited in claim 6 , wherein the three coordinates coefficients are defined by orthonormal basis functions Φ 1 , Φ 2 and Φ 3 , where T is a symbol duration and 0<t<T where
Φ
1
(
t
)
=
1
T
sin
(
2
π
t
T
)
,
Φ
2
(
t
)
=
1
T
cos
(
2
π
t
T
)
,
Φ
3
(
t
)
=
1
T
.
9. The method as recited in claim 6 , wherein the three-dimensional coordinates are is formed using identical two-dimensional (2D) signal-coordinate points constructed on equally spaced parallel layers.
10. The method as recited in claim 9 , wherein the equally spaced parallel layers are indexed by the amplitude coordinates and the 2D signal-coordinates points representing in-phase and quadrature coordinates.
11. A demodulation method, comprising:
demapping an input signal from three branches using a three-dimensional signal constellation, the three branches including a direct detection branch, and two coherent detection branches such that the direct detection branch detects an amplitude coordinate of the input signal and the two coherent detection branches detect in-phase and quadrature coordinates of the input signal;
predicting bits using log-likelihood ratios (LLRs) for demapped bits; and
iteratively decoding using at least one LDPC decoder configured to iteratively decode bits by feeding back extrinsic LLRs for demapping.
12. The method as recited in claim 11 , wherein each of the three branches are associated with a corresponding three-dimensional constellation coordinate coefficient where the amplitude detection branch has an output that is proportional to |S i | 2 , and the in-phase and quadrature detection branches have outputs proportional to Re{S i L*} and Im{s i L*} where S i is a complex envelope of a received electrical field and L is a complex envelope of a local source electrical field.
13. The method as recited in claim 11 , wherein the three-dimensional signal constellation is formed using identical two-dimensional (2D) signal-constellation points constructed on equally spaced parallel layers.
14. The method as recited in claim 13 , wherein the equally spaced parallel layers are indexed by the amplitude coefficients and the 2D signal-constellation points representing in-phase and quadrature coordinates.