Low density parity check codec and method of the same
The present invention provides a low-complexity and multi-mode Low-density Parity-check (LDPC) codec, in which the decoding operations are divided into small tasks and a unified hardware is implemented so that the hardware resources can be reused in different modes. In addition, memory access is achieved via routing networks with fixed interconnections and memory address generators, the complexity of the hardware implementation is reduced accordingly. Further, the present invention provides an early termination function with which the iterative operations can be terminated early when a threshold is reached so that the power consumption can be thus reduced. The hardware resources for early termination shares a part of hardware resources with an encoder according to the present invention so that the complexity of the hardware implementation can also be reduced.
1. A method for encoding and decoding a low-density parity-check code, comprising steps of:
(a) layering a parity-check matrix of a quasi-cyclic low-density parity-check code into a plurality of layers in a row permutation manner;
(b) partitioning each of the layers into a plurality of sub-layers by rows;
(c) partitioning each of the sub-layers into a plurality of tasks, wherein each of the layers comprises the tasks; and
(d) iteratively processing the tasks of each of the layers in a sequential manner.
2. The method for encoding and decoding a low-density parity-check code according to claim 1 , wherein the row permutation manner in step (a) enables that the l th layer after the layering comprises the (l+a 0 )mod(z) th row, the (z+(l+a 1 )mod(z)) th row, . . . , and the ((M b −1)z+(l+a Mb-1 )mod(z)) th row of the parity-check matrix, and wherein 0□l<z, and wherein a 0 , a 1 , . . . , and a Mb-1 are selected from non-negative integers, and wherein M b is the row number of a binary base matrix of the parity-check matrix.
3. The method for encoding and decoding a low-density parity-check code according to claim 1 , wherein the k th iteration of the iterative processing in step (d) comprises the following steps undertaken synchronously:
(d1) based on a plurality of check nodes and variable nodes corresponding to the tasks of the 1 th layer, using a plurality of APP values of the k th iteration of the (l-1) th layer and a plurality of “check-to-variable (C2V)” messages of the (k-1) th iteration to compute a plurality of “variable-to-check (V2C)” messages of the k th iteration;
(d2) computing a plurality of updated APP values of the k th iteration of the l th layer; and
(d3) using the “variable-to-check (V2C)” messages of the k th iteration to compute the “check-to-variable (C2V)” messages of the k th iteration.
4. The method for encoding and decoding a low-density parity-check code according to claim 3 , wherein in step (d1), the updated “check-to-variable (C2V)” messages of the k th iteration are used to replace the “check-to-variable (C2V)” messages of the (k-1) th iteration to compute the “variable-to-check (V2C)” messages of the k th iteration.
5. The method for encoding and decoding a low-density parity-check code according to claim 3 further comprising steps of:
(e1) according to step (d2), computing the signs of the updated APP values of each sub-layer of the k th iteration to obtain a value; and
(e2) terminating the iteration processing of step (d) if the values of the layers are successive zeros and the number of the successive zeros is equal to a threshold.
6. The method for encoding and decoding a low-density parity-check code according to claim 1 , wherein in step (a), a plurality of message bits of the parity-check matrix are used to compute a plurality of parity bits for each of the layers.
7. The method for encoding and decoding a low-density parity-check code according to claim 1 , wherein in step (b), a plurality of message bits of the parity-check matrix are used to compute a plurality of parity bits for each of the layers.
8. The method for encoding and decoding a low-density parity-check code according to claim 1 , wherein the low-density parity-check code is a quasi-cyclic low-density parity-check code conforming to WiMAX standards.