Early termination technique for LDPC decoder architecture

1. A method for performing low-density parity-check (LDPC) decoding, the method comprising:

receiving log-likelihood ratios (LLRs) corresponding to bits encoded using LDPC encoding;

updating the LLRs corresponding to the encoded bits corresponding to columns of a parity check matrix (PCM), the LLRs indicating a probability of a value of each of the encoded bits, wherein each row of the PCM corresponds to a parity check for the encoded bits;

iteratively decoding the encoded bits using the PCM, wherein for each iteration of decoding, for each row of the PCM, the LLRs are updated based on a processing of the row of the PCM;

maintaining an overall parity bit;

setting the overall parity bit to true at a beginning of an iteration of decoding;

for each processing of the row of the PCM, adding a result of the parity check for the LLRs to the overall parity bit during the iteration of decoding;

maintaining a sign flip bit;

setting the sign flip bit to true at the beginning of the iteration of decoding;

setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding; and

stopping iteratively decoding the encoded bits when the sign flip bit and the overall parity bit are set to true at an end of the iteration of decoding.

2. The method of claim 1 , wherein iteratively decoding the encoded bits comprises utilizing a layered LDPC decoding technique.

3. The method of claim 1 , further comprising determining if a sign of any LLR changes based on a corresponding outdated a posteriori LLR and a corresponding updated a posteriori LLR.

4. The method of claim 3 , wherein processing of the row of the PCM comprises determining for each of the encoded bits an a posteriori LLR for the row of the PCM, wherein for a first set of the encoded bits, determining the a posteriori LLR comprises using a first number of outdated LLRs, and wherein for a second set of the encoded bits, determining the a posteriori LLR comprises using a second number of updated bit LLRs, wherein the first set of the encoded bits and the second set of the encoded bits comprise all of the encoded bits.

5. The method of claim 1 , wherein setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding comprises setting the sign flip bit to false when a sign of any LLR changes after processing of a first row of the PCM during the iteration.

6. An apparatus for performing low-density parity-check (LDPC) decoding, the apparatus comprising:

a memory; and

a processor, the memory and processor being configured to:

receive log-likelihood ratios (LLRs) corresponding to bits encoded using LDPC encoding;

update the LLRs corresponding to the encoded bits corresponding to columns of a parity check matrix (PCM), the LLRs indicating a probability of a value of each of the encoded bits, wherein each row of the PCM corresponds to a parity check for the encoded bits;

iteratively decode the encoded bits using the PCM, wherein for each iteration of decoding, for each row of the PCM, the LLRs are updated based on a processing of the row of the PCM;

maintain an overall parity bit;

set the overall parity bit to true at a beginning of an iteration of decoding;

for each processing of the row of the PCM, add a result of the parity check for the LLRs to the overall parity bit during the iteration of decoding;

maintain a sign flip bit;

set the sign flip bit to true at the beginning of the iteration of decoding;

set the sign flip bit to false when a sign of any LLR changes during the iteration of decoding; and

stop iteratively decoding the encoded bits when the sign flip bit and the overall parity bit are set to true at an end of the iteration of decoding.

7. The apparatus of claim 6 , wherein the memory and processor being configured to iteratively decode the encoded bits comprises the memory and processor being configured to utilize a layered LDPC decoding technique.

8. The apparatus of claim 6 , wherein the memory and processor are further configured to determine if a sign of any LLR changes based on a corresponding outdated a posteriori LLR and a corresponding updated a posteriori LLR.

9. The apparatus of claim 8 , wherein processing of the row of the PCM comprises determining for each of the encoded bits an a posteriori LLR for the row of the PCM, wherein for a first set of the encoded bits, determining the a posteriori LLR comprises using a first number of outdated LLRs, and wherein for a second set of the encoded bits, determining the a posteriori LLR comprises using a second number of updated bit LLRs, wherein the first set of the encoded bits and the second set of the encoded bits comprise all of the encoded bits.

10. The apparatus of claim 6 , wherein the memory and processor being configured to set the sign flip bit to false when a sign of any LLR changes during the iteration of decoding comprises the memory and processor being configured to set the sign flip bit to false when a sign of any LLR changes after processing of a first row of the PCM during the iteration.

11. A non-transitory computer readable medium having instructions stored thereon that when executed by an apparatus cause the apparatus to perform a method for performing low-density parity-check (LDPC) decoding, the method comprising:

receiving log-likelihood ratios (LLRs) corresponding to bits encoded using LDPC encoding;

updating the LLRs corresponding to the encoded bits corresponding to columns of a parity check matrix (PCM), the LLRs indicating a probability of a value of each of the encoded bits, wherein each row of the PCM corresponds to a parity check for the encoded bits;

iteratively decoding the encoded bits using the PCM, wherein for each iteration of decoding, for each row of the PCM, the LLRs are updated based on a processing of the row of the PCM;

maintaining an overall parity bit;

setting the overall parity bit to true at a beginning of an iteration of decoding;

for each processing of the row of the PCM, adding a result of the parity check for the LLRs to the overall parity bit during the iteration of decoding;

maintaining a sign flip bit;

setting the sign flip bit to true at the beginning of the iteration of decoding;

setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding; and

stopping iteratively decoding the encoded bits when the sign flip bit and the overall parity bit are set to true at an end of the iteration of decoding.

12. The non-transitory computer readable medium of claim 11 , wherein iteratively decoding the encoded bits comprises utilizing a layered LDPC decoding technique.

13. The non-transitory computer readable medium of claim 11 , wherein the method further comprises determining if a sign of any LLR changes based on a corresponding outdated a posteriori LLR and a corresponding updated a posteriori LLR.

14. The non-transitory computer readable medium of claim 13 , wherein processing of the row of the PCM comprises determining for each of the encoded bits an a posteriori LLR for the row of the PCM, wherein for a first set of the encoded bits, determining the a posteriori LLR comprises using a first number of outdated LLRs, and wherein for a second set of the encoded bits, determining the a posteriori LLR comprises using a second number of updated bit LLRs, wherein the first set of the encoded bits and the second set of the encoded bits comprise all of the encoded bits.

15. The non-transitory computer readable medium of claim 11 , wherein setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding comprises setting the sign flip bit to false when a sign of any LLR changes after processing of a first row of the PCM during the iteration.

16. An apparatus for performing low-density parity-check (LDPC) decoding, the apparatus comprising:

means for receiving log-likelihood ratios (LLRs) corresponding to bits encoded using LDPC encoding;

means for updating the LLRs corresponding to the encoded bits corresponding to columns of a parity check matrix (PCM), the LLRs indicating a probability of a value of each of the encoded bits, wherein each row of the PCM corresponds to a parity check for the encoded bits;

means for iteratively decoding the encoded bits using the PCM, wherein for each iteration of decoding, for each row of the PCM, the LLRs are updated based on a processing of the row of the PCM;

means for maintaining an overall parity bit;

means for setting the overall parity bit to true at a beginning of an iteration of decoding;

means for, for each processing of the row of the PCM, adding a result of the parity check for the LLRs to the overall parity bit during the iteration of decoding;

means for maintaining a sign flip bit;

means for setting the sign flip bit to true at the beginning of the iteration of decoding;

means for setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding; and

means for stopping iteratively decoding the encoded bits when the sign flip bit and the overall parity bit are set to true at an end of the iteration of decoding.

17. The apparatus of claim 16 , wherein iteratively means for decoding the encoded bits comprises means for utilizing a layered LDPC decoding technique.

18. The apparatus of claim 16 , further comprising means for determining if a sign of any LLR changes based on a corresponding outdated a posteriori LLR and a corresponding updated a posteriori LLR.

19. The apparatus of claim 18 , wherein processing of the row of the PCM comprises determining for each of the encoded bits an a posteriori LLR for the row of the PCM, wherein for a first set of the encoded bits, determining the a posteriori LLR comprises using a first number of outdated LLRs, and wherein for a second set of the encoded bits, determining the a posteriori LLR comprises using a second number of updated bit LLRs, wherein the first set of the encoded bits and the second set of the encoded bits comprise all of the encoded bits.

20. The apparatus of claim 16 , wherein means for setting the sign flip bit to false when a sign of any LLR changes during the iteration of decoding comprises means for setting the sign flip bit to false when a sign of any LLR changes after processing of a first row of the PCM during the iteration.