Zero excess bandwidth modulation
View Patent ↗ZEB (zero excess bandwidth) modulation. The modulation rate of a signal is increased to the entirety of a communication channel's available bandwidth. Spectral zeroes at the edges of the available bandwidth now coincide with spectral zeroes at the edges of the Nyquist band and lead to ISI (intersymbol interference), which cannot be eliminated by equalization. Thus, in a ZEB system intersymbol interference (ISI) is intentionally allowed and dealt with by the known technique of TH (Tomlinson-Harashima) precoding. Comparison of conventional zero-ISI systems with ZEB systems, both exhibiting identical transmit spectra with finite spectral roll offs towards the edges of the available bandwidth, illustrate significant gains in channel throughput achievable by the ZEB systems. Similar gains can hardly be achieved by more sophisticated channel coding for zero-ISI systems. For ZEB systems an effective spectral shaping involving a simple infinite impulse response (IIR) of the overall channel is proposed.
1. A zero excess bandwidth modulation method, the method comprising:
encoding a plurality of information bits, thereby generating a plurality of encoded bits;
puncturing at least one of the plurality of encoded bits so that a plurality of remaining encoded bits includes a first encoded bit followed by a second encoded bit;
after performing the puncturing, rearranging an order of the plurality of remaining encoded bits so that the first encoded bit follows the second encoded bit, thereby generating a sequence of discrete-valued modulation symbols;
TH (Tomlinson-Harashima) precoding of the sequence of discrete-valued modulation symbols according to a predetermined overall channel symbol response having spectral zeroes at edges of a corresponding Nyquist band, thereby generating a plurality of discrete-time transmit signals at a modulation rate;
inserting the plurality of discrete-time transmit signals into means to generate a continuous-time transmit signal by appropriate discrete-time filtering, digital-to-analog conversion (DAC), and continuous-time filtering;
ensuring, within the means to generate the continuous-time transmit signal, that the continuous-time transmit signal has spectral zeroes at the edges of the corresponding Nyquist band, which equals a bandwidth of the available transmission band, and that any spectral components outside of the available transmission band are substantially suppressed; and
launching the filtered, continuous-time transmit signal into the communication channel.
2. The method of claim 1 , wherein:
the predetermined overall channel response is characterized as h(D)=1+h 1 D+h 2 D 2 + . . . ;
D=e −j2πfT (=z −1 ;
f is frequency;
T is an inverse of the bandwidth of the available transmission band;
h 1 , h 2 , . . . are constant valued coefficients; and
h(D) is zero when D=−1.
3. The method of claim 1 , further comprising:
mapping the rearranged plurality of remaining encoded bits into a plurality of modulation symbols according to a symbol constellation and a corresponding mapping function, thereby generating the sequence of discrete-valued modulation symbols.
4. The method of claim 1 , further comprising:
encoding a subset of information bits of the plurality of information bits into the plurality of encoded bits; and
mapping the rearranged plurality of remaining encoded bits and at least one uncoded information bits into a plurality of modulation symbols according to a symbol constellation and a corresponding mapping function, thereby generating the sequence of discrete-valued modulation symbols.
5. The method of claim 1 , wherein:
TH precoding operates on the discrete-valued modulation symbols to perform an inverse filtering operation in accordance with the predetermined overall channel symbol response and executes modulo operations to limit signals within a predetermined signal region, thereby generating the plurality of discrete-time transmit signals at the modulation rate.
6. The method of claim 1 , wherein:
the sequence of discrete-valued modulation symbols has a modulation type of at least one of PAM (pulse amplitude modulation), QPSK (quadrature phase shift keying), 16 QAM (quadrature amplitude modulation), and a higher-order QAM.
7. The method of claim 1 , wherein:
the encoding of the plurality of information bits thereby generating the sequence of discrete-valued modulation symbols involves at least one of uncoded modulation, TCM (trellis coded modulation), TTCM (turbo trellis coded modulation), LDPC (low density parity check) encoding and modulation, and concatenated encoding and modulation.
8. The method of claim 1 , wherein:
the method is performed cooperatively within a communication transmitter and a communication receiver that are communicatively coupled via the communication channel.
9. The method of claim 1 , further comprising:
receiving a continuous-time receive signal from the communication channel;
converting the continuous-time receive signal into a discrete-time signal by means to perform appropriate continuous-time filtering, ADC (analog-to-digital conversion), and discrete-time filtering, thereby obtaining a plurality of discrete-time receive signals at the modulation rate;
ensuring, within the means to perform appropriate continuous-time filtering, ADC, and discrete-time filtering, that any signal and noise components outside of the available transmission band is suppressed and that the discrete-time receive signal is shaped into a form corresponding to the predetermined overall channel symbol response that is assumed for the TH precoding, and
decoding the plurality of discrete-time receive signals to generate best estimates of the sequence of discrete-valued modulation symbols and the information bits encoded therein.
10. The method of claim 9 , further comprising:
performing adaptive equalization when ensuring that the discrete-time receive signal is shaped into a form corresponding to the predetermined overall channel symbol response that is assumed for the TH precoding.
11. The method of claim 1 , wherein:
the predetermined overall channel symbol response employed for TH precoding, referred to as h(D), is a finite impulse response (FIR) h(D)=1+h 1 D+h 2 D+ . . . h L D L for some finite positive integer L, or an infinite impulse response (IIR) h(D)=p(D)/q(D); and
p(D)=1+p 1 D+h P D P and q(D)=1+q 1 D+ . . . q Q D Q for some finite positive integers P and Q.
12. The method of claim 11 , wherein:
h(D)=(1+D)/(1−ρD) for 0<<ρ<1.
13. The method of claim 1 , further comprising:
after rearranging the order of the plurality of remaining encoded bits, puncturing at least one of the rearranged plurality of remaining encoded bits.
14. The method of claim 1 , wherein:
the sequence of discrete-valued modulation symbols has a modulation type of 32 QAM (quadrature amplitude modulation) (double square).
15. A zero excess bandwidth modulation communication transmitter, the transmitter comprising:
an encoder and symbol mapper that:
encodes a plurality of information bits, thereby generating a plurality of encoded bits;
performs puncturing at least one of the plurality of encoded bits so that a plurality of remaining encoded bits includes a first encoded bit followed by a second encoded bit; and
after performing the puncturing, rearranges an order of the plurality of remaining encoded bits so that the first encoded bit follows the second encoded bit, thereby generating a sequence of discrete-valued modulation symbols;
a TH (Tomlinson-Harashima) precoder that performs precoding of the sequence of discrete-valued modulation symbols according to a predetermined overall channel symbol response having spectral zeroes at edges of a corresponding Nyquist band, thereby generating a plurality of discrete-time transmit signals at a modulation rate;
means to generate a continuous-time transmit signal by appropriate discrete-time filtering, digital-to-analog conversion (DAC), and continuous-time filtering;
wherein the plurality of discrete-time transmit signals is inserted into the means; wherein the means ensures that the continuous-time transmit signal has spectral zeroes at the edges of the corresponding Nyquist band, which equals a bandwidth of the available transmission band, and that any spectral components outside of the available transmission band are substantially suppressed; and
wherein the filtered, continuous-time transmit signal is launched into the communication channel from the transmit filter.
16. The transmitter of claim 15 , wherein:
the predetermined overall channel response is characterized as h(D)=1+h 1 D+h 2 D 2 + . . . ;
D=e −j2πfT (=z −1 ;
f is frequency;
T is an inverse of the bandwidth of the available transmission band;
h 1 , h 2 , . . . are constant valued coefficients; and
h(D) is zero when D=−1.
17. The transmitter of claim 15 , wherein:
the encoder and symbol mapper maps the rearranged plurality of remaining encoded bits into a plurality of modulation symbols according to a symbol constellation and a corresponding mapping function, thereby generating the sequence of discrete-valued modulation symbols.
18. The transmitter of claim 15 , wherein:
the encoder and symbol mapper encodes a subset of information bits of the plurality of information bits into the plurality of encoded bits; and
the encoder and symbol mapper maps the rearranged plurality of remaining encoded bits and at least one uncoded information bits into a plurality of modulation symbols according to a symbol constellation and a corresponding mapping function, thereby generating the sequence of discrete-valued modulation symbols.
19. The transmitter of claim 15 , wherein:
the TH precoder operates on the discrete-valued modulation symbols to perform an inverse filtering operation in accordance with the predetermined overall channel symbol response and executes modulo operations to limit signals within a predetermined signal region, thereby generating the plurality of discrete-time transmit signals at the modulation rate.
20. The transmitter of claim 15 , wherein:
the sequence of discrete-valued modulation symbols has a modulation type of at least one of PAM (pulse amplitude modulation), QPSK (quadrature phase shift keying), 16 QAM (quadrature amplitude modulation), and a higher-order QAM.
21. The transmitter of claim 15 , wherein:
the encoder and symbol mapper performs encoding of the plurality of information bits, thereby generating the sequence of discrete-valued modulation symbols, that involves at least one of uncoded modulation, TCM (trellis coded modulation), TTCM (turbo trellis coded modulation), LDPC (low density parity check) encoding and modulation, and concatenated encoding and modulation.
22. The transmitter of claim 15 , wherein:
the communication transmitter is communicatively coupled to a communication receiver via the communication channel; and
the communication transmitter, the communication receiver, and the communication channel form a communication system.
23. The communication system of claim 22 , wherein:
the communication receiver receives a continuous-time receive signal from the communication channel;
the communication receiver converts the continuous-time receive signal into a discrete-time signal using means to perform appropriate continuous-time filtering, ADC (analog-to-digital conversion), and discrete-time filtering, thereby obtaining a plurality of discrete-time receive signals at the modulation rate;
the communication receiver ensures, within the means to perform appropriate continuous-time filtering, ADC, and discrete-time filtering, that any signal and noise components outside of the available transmission band is suppressed and that the discrete-time receive signal is shaped into a form corresponding to the predetermined overall channel symbol response that is assumed for the TH precoding, and
the communication receiver performs decoding of the plurality of discrete-time receive signals to generate best estimates of the sequence of discrete-valued modulation symbols and the information bits encoded therein.
24. The communication system of claim 23 , wherein:
the communication receiver performs adaptive equalization when ensuring that the discrete-time receive signal is shaped into a form corresponding to the predetermined overall channel symbol response that is assumed for the TH precoding.
25. The transmitter of claim 15 , wherein:
the predetermined overall channel symbol response employed by the TH precoder, referred to as h(D), is a finite impulse response (FIR) h(D)=1+h 1 D+h 2 D+h L D L for some finite positive integer L, or an infinite impulse response (IIR) h(D)=p(D)/q(D); and
p(D)=1+p 1 D+ . . . h P D P and q(D)=1+q 1 D+ . . . q Q D Q for some finite positive integers P and Q.
26. The transmitter of claim 25 , wherein:
for 0<<ρ<1.
27. The transmitter of claim 13 , wherein:
after rearranging the order of the plurality of remaining encoded bits, the encoder and symbol mapper punctures at least one of the rearranged plurality of remaining encoded bits.
28. The transmitter of claim 15 , wherein:
the sequence of discrete-valued modulation symbols has a modulation type of 32 QAM (quadrature amplitude modulation) (double square).