Channel selfadapting iteration Turbo decoder system and method
Channel selfadapting iteration Turbo decoder system and method
 CN 102,171,937 B
 Filed: 09/16/2009
 Issued: 05/18/2016
 Est. Priority Date: 09/16/2008
 Status: Active Grant
First Claim
1. a channel selfadapting iteration Turbo decoder system, comprising:
 The first and second MAP decoders, for calculating iteratively a group of window of received dataBranch metric, calculates forward recursive path status tolerance, calculates backward recursive path status tolerance, andAnd from the loglikelihood ratio of described forward direction and backward recursive path status metric calculation 1 and 0;
At least one interleaver, for the decision bits that interweaves;
Iteration count, for what count by the iteration of at least one execution of described MAP decoderNumber of times;
Nonconvergence detector, for only after the relevant iterations of predetermined channel, identifies nonconvergenceThe window of received data in those decision bits, and, thereafter, only for described nonconvergenceDecision bits, at least one that make described MAP decoder can be calculated received dataOne set of branch metrics, calculates forward direction and backward recursive path status tolerance, and from described forward direction and antiTo 1 and 0 loglikelihood ratio of each nonconvergence decision bits of recurrence path status metric calculation,Described at least one interleaver described nonconvergence bit that interweaves;
AndWatchdog device, for the nonconvergence judgement ratio in response to from least one MAP decoderSpecial quantity, and in the time that the quantity of described nonconvergence decision bits is less than threshold value, described in makingAt least one MAP decoder can be limited to each nonconvergence decision bits path status degree aroundThe wicket reducing of amount, and 1 He of calculating each nonconvergence decision bits from those path metrics0 loglikelihood ratio.
Chinese PRB Reexamination
Abstract
A kind of channel selfadapting iteration Turbo decoder system, for calculate a set of branch metrics of the window of received data with MAP decoder, calculate forward recursive path status tolerance, calculate backward recursive path status tolerance, and from the loglikelihood ratio of forward recursive path status tolerance and backward recursive path status metric calculation 1 and 0, and the decision bits that interweaves; And, identify nonconvergent those MAP decoder decision bits, calculate a set of branch metrics of the data that receive, and from 1 and 0 loglikelihood ratio of forward direction and each nonconvergence decision bits of backward recursive path status metric calculation, and the nonconvergence decision bits that interweaves.
22 Claims

1. a channel selfadapting iteration Turbo decoder system, comprising:

The first and second MAP decoders, for calculating iteratively a group of window of received dataBranch metric, calculates forward recursive path status tolerance, calculates backward recursive path status tolerance, andAnd from the loglikelihood ratio of described forward direction and backward recursive path status metric calculation 1 and 0; At least one interleaver, for the decision bits that interweaves; Iteration count, for what count by the iteration of at least one execution of described MAP decoderNumber of times; Nonconvergence detector, for only after the relevant iterations of predetermined channel, identifies nonconvergenceThe window of received data in those decision bits, and, thereafter, only for described nonconvergenceDecision bits, at least one that make described MAP decoder can be calculated received dataOne set of branch metrics, calculates forward direction and backward recursive path status tolerance, and from described forward direction and antiTo 1 and 0 loglikelihood ratio of each nonconvergence decision bits of recurrence path status metric calculation,Described at least one interleaver described nonconvergence bit that interweaves;
AndWatchdog device, for the nonconvergence judgement ratio in response to from least one MAP decoderSpecial quantity, and in the time that the quantity of described nonconvergence decision bits is less than threshold value, described in makingAt least one MAP decoder can be limited to each nonconvergence decision bits path status degree aroundThe wicket reducing of amount, and 1 He of calculating each nonconvergence decision bits from those path metrics0 loglikelihood ratio.


2. channel selfadapting iteration Turbo decoder system according to claim 1, whereinInitialize described backward recursive path metric with the probability equating.

3. channel selfadapting iteration Turbo decoder system according to claim 1, whereinDescribed in each around described nonconvergence decision bits, wicket initializes with the probability equatingOppositely and forward recursive path metric.

4. channel selfadapting iteration Turbo decoder system according to claim 1, whereinUse LogMax* () to calculate described forward direction, reverse and LLR.

5. channel selfadapting iteration Turbo decoder system according to claim 1, whereinUse MaxLogMax () to calculate described forward direction, reverse and LLR.

6. channel selfadapting iteration Turbo decoder system according to claim 1, whereinDescribed Turbo decoder is binary system Turbo decoder.

7. channel selfadapting iteration Turbo decoder system according to claim 1, whereinDescribed Turbo decoder is duobinary system Turbo decoder.

8. channel selfadapting iteration Turbo decoder system according to claim 1, wherein,The scope of the iterations that described predetermined channel is relevant is from 1 to 4 iteration.

9. channel selfadapting iteration Turbo decoder system according to claim 1, wherein,After the relevant iterations of described predetermined channel, described at least one interleaver described nonreceipts that only interweaveHold back bit.

10. channel selfadapting iteration Turbo decoder system according to claim 1, wherein,The wicket reducing of described path status tolerance uses fewer than the sum of the state of all data windowStage quantity.

11. channel selfadapting iteration Turbo decoder systems according to claim 1, wherein,The wicket reducing of described path status tolerance uses 40 stages in every side in main stage.

12. 1 kinds of channel selfadapting iteration Turbo decoder methods, comprising:

Use MAP decoder iteration ground to calculate a set of branch metrics of the window of received data, countCalculate forward direction and backward recursive path status tolerance, and from described forward direction and backward recursive path status degreeAmount is calculated 1 and 0 loglikelihood ratio; Described decision bits interweaves; The number of times of the iteration that counting is carried out by described MAP decoder; Only, after the relevant iterations of predetermined channel, identify nonconvergent those MAP decodersDecision bits, and, thereafter, only for described nonconvergent decision bits, calculate the number receivingAccording to a set of branch metrics, from described forward direction and the each nonconvergence of backward recursive path status metric calculation1 and 0 loglikelihood ratio (LLR) of decision bits; Described nonconvergence decision bits interweaves; In response to the quantity of the nonconvergence decision bits of MAP decoder, and when described MPA decoderWhen the quantity of nonconvergence decision bits is less than threshold value, be limited to each nonconvergence decision bits road aroundThe wicket reducing of footpath state measurement;
AndCalculate 1 and 0 loglikelihood ratio of each nonconvergence decision bits from those path metrics.


13. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn initialize described backward recursive path metric with the probability equating.

14. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn initialize fenestella described in each around described nonconvergence decision bits with the probability equatingReverse and forward recursive path metric in mouthful.

15. channel selfadapting iteration Turbo decoder methods according to claim 12, itsMiddle use LogMax* () calculates described forward direction and backward recursive path status tolerance and LLR.

16. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, use MaxLogMax () to calculate described forward direction and backward recursive path status tolerance and LLR.

17. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, described Turbo decoder method is binary system Turbo decoder method.

18. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, described Turbo decoder method is duobinary system Turbo decoder method.

19. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, the scope of the iterations that described predetermined channel is relevant is from 1 to 4 iteration.

20. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, the described nonconvergence decision bits that interweaves comprises the described nonconvergence bit that only interweaves.

21. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, the wicket reducing of described path status tolerance uses fewer than the sum of the state of all data windowStage quantity.

22. channel selfadapting iteration Turbo decoder methods according to claim 12, itsIn, the wicket reducing of described path status tolerance uses 40 stages in every side in main stage.
Specification(s)