Optimized receivers for faster than nyquist (FTN) transmission rates in high spectral efficiency satellite systems

1. An apparatus comprising:

a receiver configured to receive a signal transmitted over a wireless channel via a high-power amplifier of a transponder,

wherein the received signal comprises a discrete carrier signal of a respective carrier phase and frequency, and wherein the carrier signal reflects a plurality of source data symbols of a respective source signal with which it was modulated, and

wherein the carrier signal includes inter-symbol interference (ISI) effects induced based on a faster-than-Nyquist (FTN) signaling rate and a tight frequency roll-off applied to the respective source signal for transmission over the wireless channel;

a sampler circuit configured to sample the carrier signal at the FTN signaling rate to generate a sequence of sampled estimates of the source data symbols of the respective source signal; and

a receive signal processor configured to process the sequence of sampled source data symbol estimates to determine original source data symbols of the respective source signal;

wherein the receive signal processor comprises:

an equalizer configured to compensate for the ISI effects induced based on the FTN signaling rate and the tight frequency roll-off applied to the source signal; and

a decoder configured to decode an output of the equalizer to determine and regenerate the respective source signal for the carrier signal.

2. The apparatus according to claim 1 , wherein the FTN signaling rate is approximately 6% and the tight frequency roll-off is approximately 5%.

3. The apparatus according to claim 1 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the equalizer is configured to use a priori soft information of every code bit for the carrier on which the source signal was transmitted.

4. The apparatus according to claim 3 , wherein the a priori soft information comprises log-likelihood ratios (LLRs).

5. The apparatus according to claim 3 , wherein:

the decoder is further configured to generate a set of a posteriori log-likelihood ratios (LLRs) based on the output of the equalizer; and

the a priori soft information used by the equalizer comprises the a posteriori LLRs generated by the decoder.

6. The apparatus according to claim 3 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the receive signal processor is configured to process the received signal via a plurality of processing iterations, and wherein:

for a one processing iteration, the decoder is further configured to generate a set of a posteriori soft information based on the output of the equalizer; and

for a processing iteration subsequent to the one processing iteration, the a priori soft information used by the equalizer comprises the a posteriori soft information generated by the decoder for the one processing iteration.

7. The apparatus according to claim 3 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the receive signal processor is configured to process the received signal via a plurality of processing iterations, and wherein the receive signal processor further comprises:

a log-likelihood ratio (LLR) processor configured to generate, for a one processing iteration, a set of a posteriori log-likelihood ratios (LLRs) based on the output of the equalizer; and

wherein the decoder is further configured to decode, for the one processing iteration, the a posteriori LLRs generated by the LLR processor, and to provide the decoded a posteriori LLRs back to the equalizer; and

wherein the a priori soft information used by the equalizer, for a processing iteration subsequent to the one processing iteration, comprises the decoded a posteriori LLRs provided by the decoder from the one processing iteration.

8. The apparatus according to claim 7 , further comprising:

a deinterleaver configured to deinterleave the a posteriori LLRs generated by the LLR processor before being decoded by the decoder; and

an interleaver configured to interleave the decoded a posteriori LLRs generated by the decoder prior to being provided back to the equalizer.

9. A method comprising:

receiving, by a communications terminal, a signal transmitted over a wireless channel via a high-power amplifier of a transponder, wherein the received signal comprises a discrete carrier signal of a respective carrier phase and frequency, and wherein the carrier signal reflects a plurality of source data symbols of a respective source signal with which it was modulated, and wherein the carrier signal includes inter-symbol interference (ISI) effects induced based on a faster-than-Nyquist (FTN) signaling rate and a tight frequency roll-off applied to the respective source signal for transmission over the wireless channel;

sampling the carrier signal at the FTN signaling rate to generate a sequence of sampled estimates of the source data symbols of the respective source signal; and

processing the sequence of sampled source data symbol estimates to determine original source data symbols of the respective source signal;

wherein the processing of the sequence of sampled source data symbol estimates comprises:

equalizing the sequence of sampled source data symbol estimates to compensate for the ISI effects induced based on the FTN signaling rate and the tight frequency roll-off applied to the source signal; and

decoding an output of the equalizing step to determine and regenerate the respective source signal.

10. The method according to claim 9 , wherein the FTN signaling rate is approximately 6% and the tight frequency roll-off is approximately 5%.

11. The method according to claim 9 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the equalizing of the received signal comprises:

utilizing a priori soft information of every code bit for the carrier on which the source signal was transmitted.

12. The method according to claim 11 , wherein the a priori soft information comprises log-likelihood ratios (LLRs).

13. The method according to claim 11 , wherein the decoding of the output of the equalizing step comprises:

generating a set of a posteriori log-likelihood ratios (LLRs) based on the output of the equalizing step, wherein the a priori soft information utilized in the equalizing step comprises the a posteriori LLRs generated by the decoding of the output of the equalizing step.

14. The method according to claim 11 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the processing of the received signal comprises decoding the received signal via a plurality of processing iterations, and wherein:

for a one processing iteration, the decoding of the output of the equalizing step comprises generating a set of a posteriori soft information based on the output of the equalizing step; and

for a processing iteration subsequent to the one processing iteration, the a priori soft information utilized in the equalizing step comprises the a posteriori soft information generated by the decoding of the output of the equalizing step for the one processing iteration.

15. The method according to claim 11 , wherein, in compensating for the ISI effects induced by the FTN signaling rate and tight frequency roll-off, the processing of the received signal comprises decoding the received signal via a plurality of processing iterations, and wherein the method further comprises:

for a one processing iteration, generating a set of a posteriori log-likelihood ratios (LLRs) based on the output of the equalizing step, and decoding the a posteriori LLRs; and

wherein, for a processing iteration subsequent to the one processing iteration, the a priori soft information utilized in the equalizing step comprises the decoded a posteriori LLRs from the one decoding iteration.

16. The method according to claim 15 , further comprising:

deinterleaving the a posteriori LLRs before being decoded; and

interleaving the decoded a posteriori LLRs prior to being utilized in the equalizing step.

17. The method according to claim 9 , wherein the carrier signal reflects the plurality of source data symbols of the respective source signal with which it was modulated wherein the modulation was based on an 8-ary, 1+7APSK signal constellation, including an outer ring with 7 bit positions, in accordance with the following bit labeling and [x, y] bit coordinate positions:

Bit Label

[x, y] Coordinates

000

[0.0, 0.0]

001

[{square root over ((8.0 * ε x /7.0)}, 0.0]

010

[{square root over (8.0 * ε x /7.0)} * cos(4.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(4.0 * π/7.0)]

011

[{square root over (8.0 * ε x /7.0)} * cos(2.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(2.0 * π/7.0)]

100

[{square root over (8.0 * ε x /7.0)} * cos(12.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(12.0 * π/7.0)]

101

[{square root over (8.0 * ε x /7.0)} * cos(10.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(10.0 * π/7.0)]

110

[{square root over (8.0 * ε x /7.0)} * cos(6.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(6.0 * π/7.0)]

111

[{square root over (8.0 * ε x /7.0)} * cos(8.0 * π/7.0), {square root over (8.0 * ε x /7.0)} * sin(8.0 * π/7.0)]

where ε x represents average energy per symbol, and

wherein each of the [x, y] bit coordinate positions is rotated by a same rotation factor (from 0 to 359°, inclusive), and/or each bit label is altered by interchanging the 0's and 1's, and/or a uniform swapping of bit positions is applied within each bit label.

18. The method according to claim 9 , wherein the carrier signal reflects the plurality of source data symbols of the respective source signal with which it was modulated wherein the modulation is based on a 16-ary, 6+10APSK signal constellation, including an inner ring with 6 bit positions and an outer ring with 10 bit positions, in accordance with the following bit labeling and [x, y] bit coordinate positions:

Bit Label

[x, y] Coordinates

0000

[R1* {square root over (ε x )}, 0.0]

0001

[R1 * {square root over (ε x )} * cos(5.0 * π/3.0), R1 * {square root over (ε x )} * sin(5.0 * π/3.0)]

0010

[R1 * {square root over (ε x )} * cos(π/3.0), R1 * {square root over (ε x )} * sin(π/3.0)]

0011

[R2 * {square root over (ε x )} * cos(13.0 * π/30.0), R2 * {square root over (ε x )} * sin(13.0 * π/30.0)]

0100

[R2 * {square root over (ε x )} * cos(55.0 * π/30.0), R2 * {square root over (ε x )} * sin(55.0 * π/30.0)]

0101

[R2 * {square root over (ε x )} * cos(49.0 * π/30.0), R2 * {square root over (ε x )} * sin(49.0 * π/30.0)]

0110

[R2 * {square root over (ε x )} * cos(π/30.0), R2 * {square root over (ε x )} * sin(π/30.0)]

0111

[R2 * {square root over (ε x )} * cos(7.0 * π/30.0), R2 * {square root over (ε x )} * sin(7.0 * π/30.0)]

1000

[−R1* {square root over (ε x )}, 0.0]

1001

[R1 * {square root over (ε x )} * cos(4.0 * π/3.0), R1 * {square root over (ε x )} * sin(4.0 * π/3.0)]

1010

[R1 * {square root over (ε x )} * cos(2.0 * π/3.0), R1 * {square root over (ε x )} * sin(2.0 * π/3.0)]

1011

[R2 * {square root over (ε x )} * cos(19.0 * π/30.0), R2 * {square root over (ε x )} * sin(19.0 * π/30.0)]

1100

[R2 * {square root over (ε x )} * cos(37.0 * π/30.0), R2 * {square root over (ε x )} * sin(37.0 * π/30.0)]

1101

[R2 * {square root over (ε x )} * cos(43.0 * π/30.0), R2 * {square root over (ε x )} * sin(43.0 * π/30.0)]

1110

[R2 * {square root over (ε x )} * cos(31.0 * π/30.0), R2 * {square root over (ε x )} * sin(31.0 * π/30.0)]

1111

[R2 * {square root over (ε x )} * cos(25.0 * π/30.0), R2 * {square root over (ε x )} * sin(25.0 * π/30.0)]

where ε x represents average energy per symbol, 6*R1 2 +10*R2 2 =16, and R1 represents the radius of the inner ring and R2 represents the radius of the outer ring, and

wherein each of the [x, y] bit coordinate positions is rotated by a same rotation factor (from 0 to 359°, inclusive), and/or each bit label is altered by interchanging the 0's and 1's, and/or a uniform swapping of bit positions is applied within each bit label.

19. The method according to claim 9 , wherein the carrier signal reflects the plurality of source data symbols of the respective source signal with which it was modulated wherein the modulation is based on a 32-ary, 4+12+16APSK signal constellation, including an inner ring with 4 bit positions, a middle ring with 12 bit positions and an outer ring with 16 bit positions, in accordance with the following bit labeling and [x, y] bit coordinate positions:

Bit Label

[x, y] Coordinates

00000

[−R3 * {square root over (ε x )} * cos(5.0 * π/16.0), R3 * {square root over (ε x )} * cos(3.0 * π/16.0)]

00001

[−R3 * {square root over (ε x )} * cos(7.0 * π/16.0), R3 * {square root over (ε x )} * cos(π/16.0)]

00010

[R3 * {square root over (ε x )} * cos(5.0 * π/16.0), R3 * {square root over (ε x )} * cos(3.0 * π/16.0)]

00011

[R3 * {square root over (ε x )} * cos(7.0 * π/16.0), R3 * {square root over (ε x )} * cos(π/16.0)]

00100

[−R2 * {square root over (ε x )} * sin(π/4.0), R2 * {square root over (ε x )} * sin(π/4.0)]

00101

[−R2 * {square root over (ε x )} * sin(π/12.0), R2 * {square root over (ε x )} * sin(5.0 * π/12.0)]

00110

[R2 * {square root over (ε x )} * sin(π/4.0), R2 * {square root over (ε x )} * sin(π/4.0)]

00111

[R2 * {square root over (ε x )} * sin(π/12.0), R2 * {square root over (ε x )} * sin(5.0 * π/12.0)]

01000

[−R3 * {square root over (ε x )} * cos(3.0 * π/16.0), R3 * {square root over (ε x )} * cos(5.0 * π/16.0)]

01001

[−R3 * {square root over (ε x )} * cos(π/16.0), R3 * {square root over (ε x )} * cos(7.0 * π/16.0)]

01010

[R3 * {square root over (ε x )} * cos(3.0 * π/16.0), R3 * {square root over (ε x )} * cos(5.0 * π/16.0)]

01011

[R3 * {square root over (ε x )} * cos(π/16.0), R3 * {square root over (ε x )} * cos(7.0 * π/16.0)]

01100

[−R2 * {square root over (ε x )} * sin(5.0 * π/12.0), R2 * {square root over (ε x )} * sin(π/12.0)]

01101

[−R1 * {square root over (ε x )} * sin(π/4.0), R1 * {square root over (ε x )} * sin(π/4.0)]

01110

[R2 * {square root over (ε x )} * sin(5.0 * π/12.0), R2 * {square root over (ε x )} * sin(π/12.0)]

01111

[R1 * {square root over (ε x )} * sin(π/4.0), R1 * {square root over (ε x )} * sin(π/4.0)]

10000

[−R3 * {square root over (ε x )} * cos(5.0 * π/16.0), −R3 * {square root over (ε x )} * cos(3.0 * π/16.0)]

10001

[−R3 * {square root over (ε x )} * cos(7.0 * π/16.0), −R3 * {square root over (ε x )} * cos(π/16.0)]

10010

[R3 * {square root over (ε x )} * cos(5.0 * π/16.0), −R3 * {square root over (ε x )} * cos(3.0 * π/16.0)]

10011

[R3 * {square root over (ε x )} * cos(7.0 * π/16.0), −R3 * {square root over (ε x )} * cos(π/16.0)]

10100

[−R2 * {square root over (ε x )} * sin(π/4.0), −R2 * {square root over (ε x )} * sin(π/4.0)]

10101

[−R2 * {square root over (ε x )} * sin(π/12.0), −R2 * {square root over (ε x )} * sin(5.0 * π/12.0)]

10110

[R2 * {square root over (ε x )} * sin(π/4.0), −R2 * {square root over (ε x )} * sin(π/4.0)]

10111

[R2 * {square root over (ε x )} * sin(π/12.0), −R2 * {square root over (ε x )} * sin(5.0 * π/12.0)]

11000

[−R3 * {square root over (ε x )} * cos(3.0 * π/16.0), −R3 * {square root over (ε x )} * cos(5.0 * π/16.0)]

11001

[−R3 * {square root over (ε x )} * cos(π/16.0), −R3 * {square root over (ε x )} * cos(7.0 * π/16.0)]

11010

[R3 * {square root over (ε x )} * cos(3.0 * π/16.0), −R3 * {square root over (ε x )} * cos(5.0 * π/16.0)]

11011

[R3 * {square root over (ε x )} * cos(π/16.0), −R3 * {square root over (ε x )} * cos(7.0 * π/16.0)]

11100

[−R2 * {square root over (ε x )} * sin(5.0 * π/12.0), −R2 * {square root over (ε x )} * sin(π/12.0)]

11101

[−R1 * {square root over (ε x )} * sin(π/4.0), −R1 * {square root over (ε x )} * sin(π/4.0)]

11110

[R2 * {square root over (ε x )} * sin(5.0 * π/12.0), −R2 * {square root over (ε x )} * sin(π/12.0)]

11111

[R1 * {square root over (ε x )} * sin(π/4.0), −R1 * {square root over (ε x )} * sin(π/4.0)]

where ε x represents average energy per symbol, 4*R1 2 +12*R2 2 +16*R3 2 =32, and R1 represents the radius of the inner ring, R2 represents the radius of the middle ring and R3 represents the radius of the outer ring, and

wherein each of the [x, y] bit coordinate positions is rotated by a same rotation factor (from 0 to 359°, inclusive), and/or each bit label is altered by interchanging the 0's and 1's, and/or a uniform swapping of bit positions is applied within each bit label.