Method and apparatus for forward error correction
View Patent ↗A convolutional decoder for a wired or wireless receiver receiving an encoded datastream. The branch and path metric components compute branch and path metrics for the encoded datastream stream. The traceback unit couples to the path metric component to trace in each traceback interval a single path through the path metrics to determine correction factors for decoding at least two corresponding successive sub-symbols/symbols from the encoded datastream in each traceback interval.
1. An apparatus for error correction of a convolutionally encoded datastream received from a communication medium and the apparatus comprising:
a branch metric component to determine branch metrics for the encoded datastream;
a path metric component coupled to the branch metric to determine path metrics from the branch metrics determined by the branch metric component; and
a traceback component coupled to the path metric component to trace in each traceback interval a single path through the path metrics to determine correction factors for decoding at least two corresponding successive sub-symbols/symbols from the encoded datastream in each traceback interval.
2. The apparatus of claim 1 , wherein the single path exhibits a length expressed as:
N=hm+j
where m is a maximul memory order associated with encoding of the encoded datastream, h is a maximul memory order scale factor, and j is a number of sub-symbols/symbols that will be decoded from the single path.
3. The apparatus of claim 2 , wherein further j is an integer fraction of N.
4. The apparatus of claim 1 , further comprising:
a plurality of history buffers coupling the path metric component to the traceback component to buffer successive sets of path metrics from the path metric component for traceback by the traceback component.
5. The apparatus of claim 1 , further comprising:
a subset demapper coupled to the traceback component for combining the correction factors from the traceback component with the at least two corresponding successive sub-symbols/symbols from the encoded datastream to generate resultant error corrected sub-symbols/symbols corresponding thereto; and
a demapper coupled to the subset demapper to demap each error corrected sub-symbol/symbol to the corresponding bits.
6. The apparatus of claim 1 , wherein the communication medium comprises one of a wired medium and a wireless medium.
7. The apparatus of claim 1 , wherein the encoded datastream exhibits at least one of: a discrete multi-tone (DMT) communication protocol, a carrierless AM/PM (CAP) communication protocol, a quadrature amplitude modulation (QAM) communication protocol, a quadrature phase amplitude modulation (QPAM) communication protocol, a code division multiple access (CDMA) communication protocol, a binary shift keying (BPSK) communication protocol, a quaternary phase shift keying (QPSK) communication protocol.
8. A pipelined Viterbi decoder for decoding a multi-tone modulated datastream encoded with convolutional encoder with a maximul memory order “m”, and the pipelined Viterbi decoder comprising:
a two-dimensional branch metric component accepting in each of successive pipeline processing intervals an input of a pair of tones and determining there from a quantity m two-dimensional branch metrics;
a four-dimensional branch metric component coupled to the two-dimensional branch metric component to determine in each successive pipeline processing interval a quantity 2m four-dimensional branch metrics from the m two-dimensional branch metrics;
a path metric component coupled to the four-dimensional branch metric component to determine in each successive pipeline processing interval a quantity m 2 path metrics from the 2m four-dimensional branch metrics; and
a traceback component coupled to the path metric component to repetitively traceback through a succession of the path metrics provided by the path metric to determine in each traceback interval correction factors for decoding at least two corresponding ones of the tones.
9. The pipelined Viterbi decoder of claim 8 , wherein each of the two-dimensional, four-dimensional and path metric components include corresponding sub-components determining in parallel the corresponding metric and numbering in a quantity m, 2m and m 2 respectively.
10. The pipelined Viterbi decoder of claim 8 wherein the traceback component further traces in each repetitive traceback, a single path through the path metrics to determine correction factors for decoding at least two corresponding tones from the input datastream.
11. The pipelined Viterbi decoder of claim 10 , wherein the single path exhibits a length expressed as:
N=hm+j
where m is a maximul memory order associated with encoding of the encoded datastream, h is a maximul memory order scale factor, and j is a number of tones that will be decoded from the single path.
12. The pipelined Viterbi decoder of claim 11 , wherein further j is an integer fraction of N.
13. The pipelined Viterbi decoder of claim 8 , further comprising:
a plurality of history buffers coupling the path metric component to the traceback component to buffer successive sets of path metrics from the path metric component for traceback by the traceback component.
14. The pipelined Viterbi decoder of claim 8 , further comprising:
a subset demapper coupled to the traceback component for combining the correction factors from the traceback component with the at least two corresponding tones from the encoded datastream to generate resultant error corrected tones corresponding thereto; and
a demapper coupled to the subset demapper to demap each error corrected tone to the corresponding bits.
15. The pipelined Viterbi decoder of claim 8 , wherein the datastream is received from one of a wired communication medium and a wireless communication medium.
16. The pipelined Viterbi decoder of claim 8 , wherein the encoded datastream exhibits at least one of: a discrete multi-tone (DMT) communication protocol, a carrierless AM/PM (CAP) communication protocol, a quadrature amplitude modulation (QAM) communication protocol, a quadrature phase amplitude modulation (QPAM) communication protocol, a code division multiple access (CDMA) communication protocol, a binary shift keying (BPSK) communication protocol, a quaternary phase shift keying (QPSK) communication protocol.
17. A method for error correction of a convolutionally encoded datastream received from a communication medium and the method comprising:
determining branch and path metrics for the encoded datastream;
tracing back in each of successive traceback intervals a single traceback path through the path metrics; and
determining from each single traceback path correction factors for decoding at least two corresponding successive sub-symbols/symbols from the encoded datastream.
18. The method of claim 17 , wherein the single traceback path exhibits a length expressed as:
N=hm+j
where m is a maximul memory order associated with encoding of the encoded datastream, h is a maximul memory order scale factor, and j is a number of sub-symbols/symbols that will be decoded from the single traceback path.
19. The method of claim 18 , wherein further j is an integer fraction of N.
20. The method of claim 17 , further comprising the act intermediate to the first determining and the tracing back acts of:
buffering successive sets of path metrics for traceback in the act of tracing back.
21. The method of claim 17 , further comprising:
error correcting at least two corresponding successive sub-symbols/symbols from the encoded datastream with the correction factors from the second determining act; and
demapping each error corrected sub-symbol/symbol from the act of error correcting to corresponding bits.
22. The method of claim 17 , wherein the communication medium comprises one of a wired medium and a wireless medium.
23. The method of claim 17 , wherein the encoded datastream exhibits at least one of: a discrete multi-tone (DMT) communication protocol, a carrierless AM/PM (CAP) communication protocol, a quadrature amplitude modulation (QAM) communication protocol, a quadrature phase amplitude modulation (QPAM) communication protocol, a code division multiple access (CDMA) communication protocol, a binary shift keying (BPSK) communication protocol, a quaternary phase shift keying (QPSK) communication protocol.
24. A method for decoding a multi-tone modulated datastream encoded with convolutional encoder with a maximul memory order “m”, and the method comprising:
determining from an input of a pair of tones in each of successive pipeline processing intervals a quantity m two-dimensional branch metrics;
determining in each successive pipeline processing interval a quantity 2m four-dimensional branch metrics from the m two-dimensional branch metrics;
determining in each successive pipeline processing interval a quantity m 2 path metrics from the 2m four-dimensional branch metrics; and
repetitively tracing back through a succession of the path metrics to determine in each traceback interval correction factors for at least two corresponding ones of the tones.
25. The method for decoding of claim 24 , wherein the act of repetitively tracing back further comprises:
tracing back a single path through the path metrics; and
determining correction factors from the single path at least two corresponding tones from the input datastream.
26. The method of claim 25 , wherein the single path exhibits a length expressed as:
N=hm+j
where m is a maximul memory order associated with encoding of the encoded datastream, h is a maximul memory order scale factor, and j is a number of tones that will be decoded from the single path.
27. The method of claim 26 , wherein further j is an integer fraction of N.
28. The method of claim 24 , further comprising the act intermediate to the third determining and the tracing back acts of:
buffering successive sets of path metrics for traceback in the act of tracing back.
29. The method of claim 24 , further comprising:
error correcting at least two corresponding tones from the encoded datastream with the correction factors from the act of repetitively tracing back; and
demapping each error corrected sub-symbol/symbol from the act of error correcting to corresponding bits.
30. The method of claim 24 , further comprising one of the acts of:
receiving the datastream from a wired medium; and
receiving the datastream from a wireless medium.
31. The method of claim 24 , wherein the encoded datastream exhibits at least one of: a discrete multi-tone (DMT) communication protocol, a carrierless AM/PM (CAP) communication protocol, a quadrature amplitude modulation (QAM) communication protocol, a quadrature phase amplitude modulation (QPAM) communication protocol, a code division multiple access (CDMA) communication protocol, a binary shift keying (BPSK) communication protocol, a quaternary phase shift keying (QPSK) communication protocol.