Multi-resolution Viterbi decoding technique
View Patent ↗A method for decoding an encoded signal. A first step generates a plurality of first precision state metrics for a decoder trellis in response to a plurality of first precision branch metrics. A second step generates a plurality of second precision state metrics for a selected subset of the first precision state metrics in response to a plurality of second precision branch metrics. A third step replaces the selected subset of first precision state metrics with the second precision state metrics. A fourth step stores the first precision state metrics and the second precision state metrics.
1. A method for decoding comprising the steps of:
(A) generating a plurality of first precision state metrics for a decoder trellis in response to a plurality of first precision branch metrics;
(B) generating a plurality of second precision state metrics for a selected subset of said first precision state metrics in response to a plurality of second precision branch metrics;
(C) replacing said selected subset of first precision state metrics with said second precision state metrics; and
(D) storing said first precision state metrics and said second precision state metrics.
2. The method according to claim 1 , further comprising the steps of:
generating said first precision branch metrics in response to a plurality of first precision symbols; and
generating said second precision branch metrics in response to a plurality of second precision symbols.
3. The method according to claim 2 , further comprising the steps of:
generating said first precision symbols in response to first precision quantizing a plurality of received symbols; and
generating said second precision symbols in response to second precision quantizing said received symbols.
4. The method according to claim 3 , further comprising the step of storing said second precision symbols for each of a number of columns of said decoder trellis.
5. The method according to claim 1 , further comprising the step of normalizing said second precision branch metrics prior to generating said second precision state metrics in response to a normalization factor.
6. The method according to claim 5 , further comprising the step of storing said normalization factor for each of a number of columns of said decoder trellis.
7. The method according to claim 5 , wherein the step of normalizing comprises the sub-steps of:
generating a difference between a smallest first precision branch metric of said first precision branch metrics and a smallest second precision branch metric of said second precision branch metrics; and
adjusting said second precision branch metrics by said difference.
8. The method according to claim 5 , wherein the step of normalizing comprises the sub-steps of:
generating a plurality of differences between a first subset of said first precision branch metrics and a second subset of said second precision branch metrics;
generating an average difference from said differences; and
adjusting said second precision branch metrics by said average difference.
9. The method according to claim 1 , further comprising the steps of:
determining a path from one of said second precision state metrics back to an earlier first precision state metric;
generating an earlier second precision branch metric leading into said earlier first precision branch metric;
generating an earlier second precision state metric in response to said earlier second precision branch metric; and
replacing said earlier first precision branch metric with said earlier second precision state metric.
10. The method according to claim 9 , further comprising the steps of:
generating a subsequent second precision branch metric from said earlier second precision state metric to a subsequent first precision state metric along said path;
generating a subsequent second precision state metric in response to said subsequent second precision branch metric; and
replacing said subsequent first precision state metric with said subsequent second precision state metric.
11. An apparatus comprising:
a processor circuit configured to (i) generate a plurality of first precision state metrics for a decoder trellis in response to a plurality of first precision branch metrics, (ii) generate a plurality of second precision state metrics for a selected subset of said first precision state metrics in response to a plurality of second precision branch metrics, and (iii) replace said selected subset of first precision state metrics with said second precision state metrics; and
a state metrics memory configured to store said first precision state metrics and said second precision state metrics.
12. The apparatus according to claim 11 , wherein said processor circuit is further configured to:
generate said first precision branch metrics in response to a plurality of first precision symbols; and
generate said second precision branch metrics in response to a plurality of second precision symbols.
13. The apparatus according to claim 12 , further comprising a quantizer circuit configured to (i) generate said first precision symbols in response to first precision quantizing a plurality of received symbols and (ii) generate said second precision symbols in response to second precision quantizing said received symbols.
14. The apparatus according to claim 13 , further comprising a symbol memory configured to store said second precision symbols for each of a number of columns of said decoder trellis.
15. The apparatus according to claim 11 , wherein said processor circuit is further configured to normalize said second precision branch metrics prior to generating said second precision state metrics in response to a normalization factor.
16. The apparatus according to claim 15 , further comprising a normalization memory configured to store said normalization factor for each of a number of columns of said decoder trellis.
17. The apparatus according to claim 15 , wherein normalizing comprises:
generating a difference between a smallest first precision branch metric of said first precision branch metrics and a smallest second precision branch metric of said second precision branch metrics; and
adjusting said second precision branch metrics by said difference.
18. The apparatus according to claim 15 , wherein normalizing comprises:
generating a plurality of differences between a first subset of said first precision branch metrics and a second subset of said second precision branch metrics;
generating an average difference from said differences; and
adjusting said second precision branch metrics by said average difference.
19. The apparatus according to claim 11 , further comprising a path memory configured to determine a path from one of said second precision state metrics back to an earlier first precision state metric, wherein said processor circuit is further configured to:
generate an earlier second precision branch metric leading into said earlier first precision branch metric;
generate an earlier second precision state metric in response to said earlier second precision branch metric; and
replace said earlier first precision branch metric with said earlier second precision state metric in said state metrics memory.
20. An apparatus comprising:
means for generating a plurality of first precision state metrics for a decoder trellis in response to a plurality of first precision branch metrics;
means for generating a plurality of second precision state metrics for a selected subset of said first precision state metrics in response to a plurality of second precision branch metrics;
means for replacing said selected subset of first precision state metrics with said second precision state metrics; and
means for storing said first precision state metrics and said second precision state metrics.