Viterbi decoding apparatus and viterbi decoding method

1. A Viterbi decoding apparatus, comprising:

a main decoder, performing a Viterbi decoding process on input data to generate a set of main decoded results;

a re-encoder, performing a convolutional encoding process on the set of main decoded results to generate a set of re-encoded results;

an adjusting module adjusting the input data according to the set of re-encoded results to generate adjusted input data corresponding to a predetermined path in a Viterbi trellis diagram;

a secondary decoder, generating a plurality of symbols according to the adjusted input data; and

a secondary result generating module generating a set of secondary decoded results according to the plurality of symbols and the set of main decoded results.

2. The Viterbi decoding apparatus according to claim 1 , wherein the secondary decoder selects a survival path, which is traced back from a plurality of nodes of the predetermined path and different from the predetermined path, and for each of the plurality of nodes, calculates an accumulated branch metric difference from the selected survival path and the predetermined path to that node to accordingly generate the plurality of symbols.

3. The Viterbi decoding apparatus according to claim 2 , wherein the predetermined path is an all-zero path.

4. The Viterbi decoding apparatus according to claim 2 , wherein the secondary decoder comprises:

a recording unit, storing a plurality of symbol records and a metric index corresponding to each of the plurality of symbol records;

a comparing unit, receiving a latest accumulated branch metric difference, and comparing the latest accumulated branch metric difference with a currently recorded metric index in the recording unit;

an updating unit, when a determination result of the comparing unit indicates that the latest accumulated branch metric difference is better than the currently recorded metric index, updating the currently recorded symbol record and metric index according to the latest accumulated branch metric difference and the corresponding survival path; and

a symbol establishing unit, determining whether to establish a new symbol record and a metric index corresponding to the new symbol record; when it is determined that the new symbol record and the metric index corresponding to the new symbol record are to be established, the symbol establishing unit establishing the new symbol record and the metric index corresponding to the new symbol record in the recording unit according to the latest accumulated branch metric difference and the corresponding survival path.

5. The Viterbi decoding apparatus according to claim 3 , wherein the input data comprises N sets of data, the set of re-encoded results comprise N re-encoded results, where N is an integer greater than 1, and the adjusting module comprises:

a positive/negative sign determining unit, obtaining N positive/negative signs in the N re-encoded results; and

a multiplier, multiplying an i th positive/negative sign of the N positive/negative signs by an i th set of data of the input data to generate N sets of adjusted data as the adjusted input data, where i is an integral index between 1 and N.

6. A Viterbi decoding method, comprising:

a) performing a Viterbi decoding process on input data to generate a set of main decoded results;

b) performing a convolutional encoding process on the set of main decoded results to generate a set of re-encoded results;

c) adjusting the input data to generate adjusted input data corresponding to a predetermined path in a Viterbi trellis diagram;

d) generating a plurality of symbols according to the adjusted input data; and

e) generating a set of secondary decoded results according to the plurality of symbols and the set of main decoded results

wherein the step of generating the plurality of symbols comprises:

selecting a survival path, which is traced back from a plurality of nodes of the predetermined path and different from the predetermined path, and for each of the plurality of nodes, calculating an accumulated branch metric difference from the selected survival path and the predetermined path to that node to accordingly generate the plurality of symbols, and

wherein step (d) comprises:

storing a plurality of symbol records and a metric index corresponding to each of the plurality of symbol records;

receiving a latest accumulated branch metric difference, and comparing the latest accumulated branch metric difference with a currently recorded metric index;

when it is determined that the latest accumulated branch metric difference is better than the currently recorded metric index, updating the currently recorded symbol record and metric index according to the latest accumulated branch metric difference and the corresponding survival path;

determining whether to establish a new symbol record and a metric index corresponding to the symbol record; and

when it is determined that the new symbol record and the metric index corresponding to the new symbol record are to be established, establishing the new symbol record and the metric index corresponding to the new symbol record according to the latest accumulated branch metric difference and the corresponding survival path.

7. The Viterbi decoding method according to claim 6 , wherein the predetermined path is an all-zero path.

8. The Viterbi decoding method according to claim 7 , wherein the input data comprises N sets of data, the set of re-encoded results comprise N re-encoded results, where N is an integer greater than 1, and step (c) comprises:

obtaining N positive/negative signs in the N sets of re-encoded results; and

multiplying an i th positive/negative sign of the N positive/negative signs by an i th set of data of the input data to generate N sets of adjusted data as the adjusted input data, where i is an integral index between 1 and N.