Method for stopping iteration in an iterative turbo decoder and an iterative turbo decoder

摘要

The present document discloses a method for stopping iteration in an iterative Turbo decoder and an iterative Turbo decoder. Hard decisions from the two convolutional decoders of the iterative Turbo decoder are used simultaneously to determining when to stop the iteration in the iterative Turbo decoder.

主权项

1. A method for stopping iteration in an iterative Turbo decoder comprising:
obtaining a first hard decision H(1, j, k) and a second hard decision H(2, j, k) for systematic bit number k at iteration step j=1;
wherein k=1, 2, . . . , N, and N is the number of systematic bits;wherein H(1, j, k) is obtained by performing hard decision on a first log-likelihood ratio output from a first convolutional decoder of the iterative Turbo decoder for systematic bit number k at iteration step j=1, and H(2, j, k) is obtained by performing hard decision on a deinterleaved second log-likelihood ratio for systematic bit number k at iteration step j=1, the deinterleaved second log-likelihood ratio is obtained by deinterleaving a second log-likelihood ratio output from a second convolutional decoder of the iterative Turbo decoder; determining whether A(j) is zero at j=1, wherein A(j) is calculated by:A⁡(j)=∑k=1N⁢H⁡(1,j,k)-H⁡(2,j,k), where j=1; and stopping the iteration if A(j) is zero at j=1; wherein in the determining step of A(j) at j=1, when A(j) is nonzero at j=1, the method further comprises: obtaining a first hard decision H(1, j, k) and a second hard decision H(2, j, k) for systematic bit number k at iteration step j, wherein j=2, 3, . . . ;
wherein H(1, j, k) is obtained by performing hard decision on a first log-likelihood ratio output from the first convolutional decoder of the iterative Turbo decoder for systematic bit number k at iteration step j, and H(2, j, k) is obtained by performing hard decision on a deinterleaved second log-likelihood ratio for systematic bit number k at iteration step j, the deinterleaved second log-likelihood ratio is obtained by deinterleaving a second log-likelihood ratio output from the second convolutional decoder of the iterative Turbo decoder; determining whether B(j) or C(j) is zero at j>1,
wherein B(j) is calculated by:B⁡(j)=∑k=1N⁢H⁡(2,j-1,k)-H⁡(1,j,k), where j=2, 3, . . . ,
wherein C(j) is calculated by:C⁡(j)=∑k=1N⁢H⁡(1,j-1,k)-H⁡(1,j,k), where j=2, 3, . . . , and stopping the iteration when B(j) or C(j) is zero.