Method and apparatus for buffering an encoded signal for a turbo decoder
View Patent ↗A method and apparatus for buffering an encoded signal having a plurality of codewords for a turbo decoder is provided. The method comprises de-interleaving each sub-block of the codeword received at the turbo-decoder; and storing LLRs of the de-interleaved codeword LLRs into an input buffer. Thereafter, each of punctured locations, if any, in the de-interleaved codeword is indicated to a read logic for enabling the latter to fill in each of those locations with a pre-determined LLR value as and when a read request corresponding to one of those locations arrives. This method obviates the need for storing the pre-determined LLRs at the punctured locations into the input buffer and thereby cuts down the input latency of turbo decoder significantly for higher code rates.
1. A method for buffering an encoded signal received by a turbo decoder in a communication network, the encoded signal having been generated by a turbo encoder after a puncturing process and including a codeword having sub-blocks, the method comprising:
de-interleaving the sub-blocks of the codeword received at the turbo decoder, each sub-block being either a data sub-block or a parity sub-block and including a plurality of Log Likelihood Ratios (LLRs), each of the plurality of LLRs being either a punctured LLR or a valid LLR;
storing the codeword received by the turbo decoder in an input buffer; and
indicating to a read logic, for at least one punctured parity sub-block of the codeword, a position of each punctured LLR and each valid LLR, wherein
each punctured LLR corresponds to a party bit removed during the puncturing performed by the turbo encoder, and
each valid LLR corresponds to a parity bit retained during the puncturing performed by the turbo encoder.
2. The method of claim 1 , wherein
the indicating includes creating an array corresponding to one punctured parity sub-block of the at least one punctured parity sub-block,
the one punctured parity sub-block includes a punctured LLR, and
the size of the array is associated with the size of the one punctured parity sub-block.
3. The method of claim 2 , wherein flags are set in the array for indicating the position of each valid LLR and each punctured LLR in the one parity sub-block of the at least one punctured parity sub-block.
4. The method of claim 2 , wherein the array is created simultaneous with storing a data sub-block of the codeword in the input buffer.
5. The method of claim 4 , wherein the read logic accesses the array corresponding to the one punctured parity sub-block to check if a current read address corresponds to one of a valid LLR or a punctured LLR.
6. The method of claim 5 , wherein the read logic accesses the array corresponding to the one punctured parity sub-block simultaneous with reading a sub-block from the input buffer.
7. The method of claim 6 , wherein, if the current read address corresponds to a valid LLR, the read logic accepts the valid LLR read from the input buffer.
8. The method of claim 6 , further comprising, if the current read address corresponds to a punctured LLR, performing de-puncturing for the current read address.
9. The method of claim 8 , wherein the de-puncturing includes outputting a pre-defined value to the read logic for the punctured LLR.
10. The method of claim 2 , wherein, if turbo encoder performed the puncturing in a random pattern, the array is created for each punctured parity sub-block of the at least one punctured parity sub-block.
11. The method of claim 10 , wherein, if the turbo encoder performed the puncturing in compliance with IEEE 802.16, a common array is created for a plurality of punctured parity sub-blocks of the at least one punctured parity sub-block.
12. The method of claim 11 , wherein, if the plurality of punctured parity sub-blocks are completely punctured, a one-bit array is created for the plurality of punctured parity sub-blocks.
13. A turbo decoder in a communication network that receives an encoded signal generated by a turbo encoder after a puncturing process, the encoded signal including a codeword having sub-blocks, the turbo decoder comprising:
a read logic;
a de-interleaving module configured to de-interleave the sub-blocks of the codeword, each sub-block being either a data sub-block or a parity sub-block and including a plurality of Log Likelihood Ratios (LLRs), each of the plurality of LLRs being either a punctured LLR or a valid LLR;
an input buffer configured to store the codeword; and
an indicating module configured to indicate to the read logic, for at least one punctured parity sub-block of the codeword, a position of each punctured LLR and each valid LLR, wherein
each punctured LLR corresponds to a parity bit removed during the puncturing performed by the turbo encoder, and
each valid LLR corresponds to a parity bit retained during the puncturing performed by the turbo encoder.
14. The turbo decoder of claim 13 , wherein the indicating module is configured to create an array corresponding to one punctured parity sub-block of the at least one punctured parity sub-block,
the one punctured parity sub-block includes a punctured LLR, and
the size of the array is associated with the size of the one punctured parity sub-block.
15. The turbo decoder of claim 14 , wherein the indicating module is configured to set flags in the array for indicating the position of each valid LLR and each punctured LLR in the one parity sub-block of the at least one punctured parity sub-block.
16. The turbo decoder of claim 14 , wherein the read logic is configured to access the array corresponding to the one punctured parity sub-block to check if a current read address corresponds to one of a valid LLR or a punctured LLR.
17. The turbo decoder of claim 16 , wherein the read logic is configured to access the array corresponding to the one punctured parity sub-block simultaneous with reading a sub-block from the input buffer.
18. The turbo decoder of claim 16 , wherein, if the current read address corresponds to a valid LLR, the read logic is configured to accept the valid LLR read from the input buffer.
19. The turbo decoder of claim 16 , further comprising a de-puncturing module configured to perform de-puncturing of the current read address if the current read address corresponds to a punctured LLR.
20. The turbo decoder of claim 19 , wherein the de-puncturing includes outputting a pre-defined value to the read logic for the punctured LLR.