TWI682636B - Ldpc code decoding method for communication system and communication device using the same - Google Patents

Abstract

A LDPC code decoding method for a communication system and a communication device using the same are provided. A parity-check matrix represents a relationship among a plurality of variable nodes and a plurality of check nodes. One of the variable nodes is a target variable node. At least one of the variable nodes is a relating variable node. The relating variable node is related to the target variable node. The decoding method includes the following steps. A plurality of log likelihood ratios (LLRs) of the target variable node and the relating variable node are received. A minimum, a secondary minimum, a maximum and a degree of the LLRs are obtained. A correction value is obtained according to the minimum, the secondary minimum, the maximum and the degree. The LLR of the target variable node is updated according to the correction value.

Description

Decoding method and application of low density parity check code in communication system Its communication device

The present invention relates to a decoding method and a communication device using the same, and particularly relates to a decoding method of a low-density parity check code of a communication system and a communication device using the same.

In communication systems, encoding and decoding are indispensable technologies. Among them, the low-density parity check code (Low-density parity check code, LDPC code) decoding technology has received attention in recent years and will be used in future 5G communication systems. Message Passing Algorithm (Message Passing Algorithm) is the core architecture of low-density parity check codes. It uses the probability that the variable node other than itself is 0 or 1 to estimate the probability that its variable node is 0 or 1. To decode the received value as 0 or 1.

In the traditional decoding process of low-density parity check codes, the normalized min-sum algorithm (normalized min-sum algorithm) can be used algorithm) or offset min-sum algorithm to update the probability.

However, the researchers found that the normalized minimum-sum algorithm requires a lot of computing resources to find the normal value, and the offset minimum-sum algorithm also requires a lot of resources to find the offset value. Therefore, the decoding technology of low-density parity check codes is currently applied to 5G communication systems with a relatively large data transmission volume. It has encountered major bottlenecks and is difficult to make breakthroughs.

The invention relates to a decoding method of a low-density parity check code of a communication system and a communication device using the same. The correction value is estimated in advance for various situations and recorded in a look-up table, so that it can operate at a relatively low computing resource Realize high-precision decoding results, especially for 5G communication systems with a large amount of data transmission, and have very good performance.

According to the first aspect of the present invention, a decoding method for a low-density parity check code (LDPC code) of a communication system is proposed. A parity-check matrix represents the relationship between a plurality of variable nodes and a plurality of check nodes. One of the variable nodes is a target variable node. At least one of the variable nodes is a related variable node. The at least one related variable node is related to the target variable node. The decoding method includes the following steps. Receive the logarithmic likelihood ratio of the target variable node and the related variable node (log likelihood ratio, LLR). Obtain a minimum value, a small value, a maximum value, and a degree of the log-likelihood ratios. Based on the minimum value, the next-smallest value, the maximum value, and the quantity, a correction value is obtained. The log-likelihood ratio of the target variable node is updated according to the correction value.

According to the first aspect of the present invention, a communication device is proposed. The communication device uses a low-density parity check code (LDPC code) for decoding. A parity-check matrix represents the relationship between a plurality of variable nodes and a plurality of check nodes. One of the variable nodes is a target variable node. At least one of the variable nodes is a related variable node. The at least one related variable node is related to the target variable node. The communication device includes a data receiving unit, a data comparing unit, a correction value obtaining unit and a data updating unit. The data receiving unit is used to receive a plurality of log likelihood ratios (LLRs) of the target variable node and the related variable node. The data comparison unit is used to obtain a minimum value, a small value, a maximum value, and a degree of the log-likelihood ratios. The correction value obtaining unit is used for obtaining a correction value according to the minimum value, the sub-small value, the maximum value and the quantity. The data updating unit is used to update the log-likelihood ratio of the target variable node according to the correction value.

In order to have a better understanding of the above and other aspects of the present invention, the following examples are specifically described in conjunction with the accompanying drawings as follows:

100‧‧‧Communication device

110‧‧‧Data receiving unit

120‧‧‧Data comparison unit

130‧‧‧Modified value acquisition unit

140‧‧‧Data update unit

150‧‧‧Verification unit

160‧‧‧storage unit

C1, C2, C3‧‧‧ inspection nodes

C2V‧‧‧ correction value

C6‧‧‧Uniform distribution curve

C71, C72, C73, C74, C81, C82 ‧‧‧ curve

deg‧‧‧ quantity

G1‧‧‧Tan Nengtu

LLR1, LLR1’, LLR3, LLR6 ‧‧‧ log likelihood ratio

LB1, LB2, LB3 ‧‧‧ Estimated lower limit

M1‧‧‧Parity check matrix

max‧‧‧Max

min1‧‧‧min

min2‧‧‧ times small value

S110, S120, S130, S140, S150, S160

T1‧‧‧Look-up table

UB1, UB2, UB3 ‧‧‧ estimated upper limit

V1, V2, V3, V4, V5, V6 ‧‧‧ variable node

φ(x)‧‧‧Conversion function

Figure 1 shows a schematic diagram of a parity-check matrix and a Tanner graph.

Figure 2 shows a schematic diagram of updating the log likelihood ratio (LLR) of variable nodes.

Figure 3 shows a schematic diagram of a conversion function.

FIG. 4 is a flowchart of a method for decoding a low-density parity check code according to an embodiment.

FIG. 5 is a schematic diagram of a communication device according to an embodiment.

Figure 6 shows a schematic diagram of a uniform distribution curve.

Figure 7 illustrates the accuracy comparison results of the low density parity check code decoding method.

FIG. 8 illustrates the comparison result of the iteration speed of the decoding method of the low density parity check code.

In the communication system of this embodiment, the low-density parity check code (LDPC code) decoding method used can achieve high-precision decoding results under very low computing resources, especially in applications The 5G communication system with relatively large data transmission capacity has a very good performance.

Please refer to FIG. 1, which shows a schematic diagram of a parity-check matrix M1 and a Tanner graph G1. The parity check matrix M1 represents several variable nodes The relationship between V1, V2, V3, V4, V5, V5 and several check nodes (C1, C2, C3). In the parity check matrix M1, "1" indicates that there is correlation, and "0" indicates that there is no correlation.

As shown in FIG. 1, according to the content of the parity check matrix M1, a Tanneng graph G1 can be drawn. In Tan Nengtu G1, each line segment has correlation. For example, in the parity check matrix M1, the variable node V1 and the check node C3 correspond to "1", so in the Tanneng graph G1, there is a connection between the variable node V1 and the check node C3. In the parity check matrix M1, the variable node V2 and the check node C3 correspond to "0", so in the Tanneng graph G1, there is no connection between the variable node V2 and the check node C3. The variable node V3 and the check node C3 correspond to "1", so in the Tanneng graph G1, there is a connection between the variable node V3 and the check node C3. In the parity check matrix M1, the variable node V4 and the check node C3 correspond to "0", so in the Tanneng graph G1, there is no connection between the variable node V4 and the check node C3, and so on.

Please refer to FIG. 2, which shows a schematic diagram of updating the log likelihood ratio (LLR) LLR1 of the variable node V1. Figure 2 only shows the variable nodes V1, V3, V6 connected to the inspection node C3 and its connection. The variable node V1 has a log-likelihood ratio LLR1, and the log-likelihood ratio LLR1 represents the probability that the value of the variable node V1 is 0 or 1. The variable node V3 also has a log-likelihood ratio LLR3. The log-likelihood ratio LLR3 represents the probability that the value of the variable node V3 is 0 or 1. The variable node V6 also has a log-likelihood ratio LLR6. The log-likelihood ratio LLR6 represents the probability that the value of the variable node V6 is 0 or 1.

In the low-density parity check decoding method of this embodiment, the logarithmic likelihood ratio LLR1 of the variable node V1 can be updated using the feedback of the logarithmic likelihood ratio LLR3 and the logarithmic likelihood ratio LLR6 to converge to a more accurate value.

Please refer to FIG. 3, which shows a schematic diagram of a transfer function φ(x). The transfer function φ(x) is symmetrical to the line φ(x)= x . The relationship between the conversion function φ(x) and the variable x is, for example, formula (1):

Theoretically, when the log-likelihood ratio LLR1 is updated, a correction value C2V needs to be calculated first, and the calculation formula is as formula (2). According to the symmetrical characteristic of the transfer function φ(x), when the X axis and the Y axis are inverted, the function is still the same. As shown in the upper graph of Figure 3, when the log-likelihood ratio LLR3 and the log-likelihood ratio LLR6 are brought into the transfer function φ (x), φ (LLR3) and φ (LLR6) can be obtained, respectively. As shown in the lower graph of Figure 3, when φ (LLR3)+ φ (LLR6) is substituted into the transfer function φ (x), φ ( φ (LLR3)+ φ (LLR6)) can be obtained. φ ( φ (LLR3)+ φ (LLR6)) is the correction value C2V. It can be seen from the comparison between the upper graph and the lower graph in Figure 3 that this correction value C2V must be smaller than the log-likelihood ratio LLR3 and the log-likelihood ratio LLR6.

C2V=φ(φ(LLR3)+φ(LLR6))........................(2)

Then, the log-likelihood ratio LLR1 is updated according to formula (3) to obtain the updated log-likelihood ratio LLR1': LLR1 ' =LLR1+C2V........................ ...................(3)

Because the calculation of the above formula (2) is too complicated, a large amount of computing resources are required, and it has a considerable bottleneck to be implemented in the 5G communication system. This embodiment proposes The following low-density parity check code decoding method enables the log-likelihood ratio to be updated with relatively low computing resources, thereby achieving high-precision and low-computing decoding results.

Please refer to FIG. 4 and FIG. 5, FIG. 4 shows a flowchart of a low-density parity check code decoding method according to an embodiment, and FIG. 5 shows a schematic diagram of a communication device 100 according to an embodiment. The communication device 100 is, for example, a smart phone, a notebook computer, a base station, or a server. The communication device 100 includes a data receiving unit 110, a data comparison unit 120, a correction value obtaining unit 130, a data updating unit 140, a verification unit 150, and a storage unit 160. The data receiving unit 110, the data comparing unit 120, the correction value obtaining unit 130, the data updating unit 140, and the verification unit 150 are, for example, a chip, a circuit, a circuit board, or a storage device that stores array code. The storage unit 160 is, for example, a memory, a hard disk, or a register. The following describes the operation of each component in detail with a flowchart.

The following is an example of how to update the log-likelihood ratio LLR1 of the variable node V1 by taking figures 1 and 2 as examples. As shown in Figure 1, the parity check matrix M1 represents the relationship between the variable nodes V1~V6 and the check nodes C1~C3. As shown in FIG. 2, the variable node V1 is a target variable node to be updated, and the variable node V1 is connected to the check node C3. The variable nodes V3 and V6 connected to the test node C3 are related variable nodes.

In step S110, the data receiving unit 110 receives the log-likelihood ratio (eg, log-likelihood ratio LLR1, LLR3, LLR6 of the target variable node (eg, variable node V1) and related variable nodes (eg, variable nodes V3, V6) ). In an embodiment, the logarithms of all the variable nodes V1~V6 are approximately proportional to LLR1~ LLR6 is received, and in this step, only the log-likelihood ratio LLR1, LLR3, LLR6 is taken out for subsequent calculation.

Next, in step S120, the data comparison unit 120 obtains a minimum value min1, a minimum value min2, a maximum value max A degree (degree) deg. In this example, the range of comparison is the log-likelihood ratio LLR1, LLR3, LLR6, which includes the log-likelihood ratio LLR1 of the variable node V1. In this example, the log-likelihood ratio LLR1, LLR3, and LLR6 is three. The minimum value min1 may be the log-likelihood ratio of the target variable node, or it may be the log-likelihood ratio of the related variable node. The next smallest value min2 may be the log-likelihood ratio of the target variable node, or it may be the log-likelihood ratio of the related variable node. The maximum value max may be the log-likelihood ratio of the target variable node, or it may be the log-likelihood ratio of the related variable node. The number deg may also be a value above 2 or 3. Alternatively, the minimum value min1 and the maximum value may be equal. The next smallest value min2 may be equal to the maximum value max. It is also possible that there is only the minimum value min1, and there is no sub-minimum value min2. The handling of various situations will be described in the following paragraphs.

Then, in step S130, the correction value obtaining unit 130 obtains a correction value C2V based on the minimum value min1, the second minimum value min2, the maximum value max, and the number deg. In one embodiment, the correction value C2V is recorded in a look-up table T1. The lookup table T1 is stored in the storage unit 160. Through the look-up table T1, the correction value C2V can be retrieved with the minimum value min1, the second minimum value min2, the maximum value max, and the number deg. In this way, without performing complicated calculation of the conversion function φ(x), the correction value C2V can be quickly found, which greatly reduces the consumption of calculation resources.

Next, in step S140, the data updating unit 140 updates the log-likelihood ratio (eg, log-likelihood ratio LLR1) of the target variable node (eg, variable node V1) according to the correction value C2V. For example, this step is calculated according to the above formula (3). This calculation has a low complexity and does not need to consume a large amount of computing resources.

Then, in step S150, the verification unit 150 determines whether a convergence condition is satisfied. If the convergence condition is satisfied, then step S160 is entered; if the convergence condition is not satisfied, then return to step S120 for iteration. In one embodiment, the convergence condition is, for example, whether the change in the correction value C2V is lower than a predetermined value (or a predetermined ratio). In one embodiment, the convergence condition is, for example, whether the change of the log-likelihood ratio LLR1 is lower than a predetermined value (or a predetermined ratio). In one embodiment, the convergence condition is, for example, whether the number of iterations exceeds a predetermined number of times.

In step S160, the data updating unit 140 outputs the updated log-likelihood ratio (for example, log-likelihood ratio LLR1') of the target variable node (for example, the variable node V1).

In the above step S130, different processes can be performed according to the following various situations.

The first case: if the minimum value min1 is 0, the correction value C2V is 0. According to the characteristics of the transfer function φ(x), the correction value C2V must be less than the minimum value min1. Once the minimum value min1 is already 0, the correction value C2V can be directly regarded as 0.

The second case: if the number deg is 2, the correction value C2V is the minimum value min1. When the number deg is 2, there is only one target variable node and one related variable node, and a related variable node does not need to perform the calculation of formula (2) at all, and the minimum value min1 can be directly used as the correction value C2V.

The third case: if the minimum value min1 is equal to the maximum value max, the correction value C2V is obtained based on the minimum value min1. For example, once the minimum value min1 is equal to the maximum value max, it means that all logarithmic likelihood ratios of the relevant variable nodes are the minimum value min1, so the correction value C2V can be obtained by using formula (4).

C2V = φ ( φ ( min 1)*( deg -1))........................(4)

The fourth case: if the subminimum value min2 is equal to the maximum value max, the correction value C2V is obtained based on the minimum value min1 and the subminimum value min2. For example, once the subminimum value min2 is equal to the maximum value max, it means that one of the logarithmic likelihood ratios of the relevant variable nodes is the minimum value min1, and the rest is the subminimum value min2, so the correction value C2V can use the formula (5) to obtain.

C2V = φ ( φ ( min 1)+ φ ( min 2)*( deg -2))............(5)

Fifth case: If the log-likelihood ratio of the target variable node is the second smallest value min2, and the estimated upper limit UB1 and the estimated lower limit LB1 of these log-likelihood ratios are equal, the correction value C2V is the estimated lower limit LB1. For example, once the log-likelihood ratio of the target variable node is the second smallest value min2, it means that there is no sub-minimum value min2 among all the log-likelihood ratios of the related variable nodes. When estimating the estimated lower limit LB1 of the log-likelihood ratio, all the log-likelihood ratios of the relevant variable nodes can be regarded as the minimum value min1, and calculated by formula (6). When estimating the estimated upper limit UB1 of the log-likelihood ratio, it can be considered that there is only one minimum value min1 in the log-likelihood ratio of the relevant variable nodes, and the rest are the maximum value max, which is calculated by formula (7).

LB 1 = φ ( φ ( min 1)*( deg -1))........................(6)

UB 1 = φ ( φ ( min 1)+ φ ( max )*( deg -2))............(7)

Sixth case: If the log-likelihood ratio of the target variable node is not the minimum value min1 or the next-smallest value min2, and one of these log-likelihood ratios is estimated to be the upper limit UB2 And an estimated lower limit phase LB2, etc., the correction value C2V is the estimated lower limit LB2. For example, once the log-likelihood ratio of the target variable node is not the minimum value min1 or the next-smallest value min2, it means that there is a minimum value min1 and the next-smallest value min2 in all log-likelihood ratios of the relevant variable nodes. When estimating the estimated lower limit LB2 of the log-likelihood ratio, it can be considered that there is only one sub-minimum value min2 in the log-likelihood ratio of the relevant variable nodes, and the rest are all the minimum value min1, which is calculated by formula (8). When estimating the upper limit UB2 of the log-likelihood ratio, the log-likelihood ratio of the related variable nodes can be regarded as only a minimum value min1 and a sub-minimum value min2, and the rest are the maximum values, and the formula (9) Calculation.

LB 2 = φ ( φ ( min 1)*( deg -2)+ φ ( min 2)).........(8)

UB 2 = φ ( φ ( min 1)+ φ ( min 2)+ φ ( max )*( deg -3))............................... .................................................. ...(9)

Seventh case: Under the conditions other than the first to sixth cases above, the correction value C2V can be set to the average of an estimated lower limit LB3 and an estimated upper limit UB3. The calculation method of the estimated upper limit LB3 and the estimated lower limit LB3 may adopt the above formulas (6) to (9), or adopt other methods.

Alternatively, the correction value C2V can be obtained according to the estimated value of the uniform distribution of the log-likelihood ratio of the relevant variable nodes. For example, please refer to FIG. 6, which shows a schematic diagram of a uniform distribution curve C6. The values of the log-likelihood ratio of the related variable nodes can be assumed to be evenly distributed between the minimum value min1 and the maximum value max, and several uniform distribution estimates can be obtained. The correction value C2V is calculated based on these uniform distribution estimates.

Please refer to FIG. 7, which illustrates the accuracy comparison result of the decoding method of the low density parity check code. Curve C71 is 32 iterations using the traditional decoding method As a result, curve C72 is the result of three iterations using the decoding method of this embodiment, curve C73 is the result of four iterations using the decoding method of this embodiment, and curve C74 is the decoding method of this embodiment. The result of 8 iterations. It can be clearly seen from FIG. 7 that the result of the iteration of the decoding method 3 times in this embodiment is already superior to the result of the iteration of the traditional decoding method 32 times. Therefore, the decoding method of this embodiment has a significant improvement in accuracy.

Please refer to FIG. 8, which illustrates the comparison result of the iteration speed of the decoding method of the low density parity check code. Curve C81 is the iteration result using the traditional decoding method, and curve C82 is the iteration result using the decoding method of this embodiment. It can be clearly seen from FIG. 8 that under the same SNR value, the number of iterations required by the decoding method of this embodiment is significantly lower than the number of iterations of the conventional decoding method. Therefore, the decoding method of this embodiment has a significant improvement in calculation speed.

In summary, although the present invention has been disclosed as above with examples, it is not intended to limit the present invention. Those with ordinary knowledge in the technical field to which the present invention belongs can make various modifications and retouching without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be deemed as defined by the scope of the attached patent application.

S110, S120, S130, S140, S150, S160

Claims (20)

A decoding method of a low-density parity check code (LDPC code) of a communication system, in which a parity-check matrix represents a plurality of variable nodes and a plurality of check nodes ( check node), one of the variable nodes is a target variable node, at least one of the variable nodes is a related variable node, and the at least one related variable node is related to the target variable node, The decoding method includes: receiving a plurality of log likelihood ratios (LLRs) of the target variable node and the related variable nodes; obtaining a minimum value, a small value, and a maximum value of the log likelihood ratios A degree (degree); based on the minimum value, the sub-smallest value, the maximum value, and the amount, obtain a correction value; and update the log-likelihood ratio of the target variable node according to the correction value. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein the correction value is recorded in a look-up table. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the minimum value is 0, the correction value is 0. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the number is 2, the correction value is the minimum value. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the minimum value is equal to the maximum value, the correction value is based on the minimum value The value is obtained. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the minor value is equal to the maximum value, the correction value is based on the The minimum value and the minimum value are obtained. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the log-likelihood ratio of the target variable node is the second smallest value, And the estimated upper limit and the estimated lower limit of the log-likelihood ratios are equal, the correction value is the estimated lower limit, the estimated upper limit is obtained based on the minimum value and the sub-small value, and the estimated lower limit is obtained based on the minimum value . The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, if the logarithmic ratio of the target variable node is not the minimum value or the The second smallest value, and the upper and lower estimates of the log-likelihood ratios are equal, then the correction value is the estimate The lower limit, the estimated upper limit is obtained based on the minimum value, the sub-small value and the maximum value, and the estimated lower limit is obtained based on the minimum value and the sub-small value. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, the correction value is an average of an estimated lower limit and an estimated upper limit. The method for decoding a low-density parity check code of a communication system as described in item 1 of the patent scope, wherein in the step of obtaining the correction value, the correction value is based on the at least one logarithm of the at least one related variable node The estimated value of the uniform distribution of the ratio is obtained. A communication device adopts a low-density parity check code (LDPC code) for decoding, and a parity-check matrix represents a plurality of variable nodes and a plurality of test nodes (check node), one of the variable nodes is a target variable node, at least one of the variable nodes is a related variable node, and the at least one related variable node is related to the target variable node The communication device includes: a data receiving unit for receiving a plurality of log likelihood ratio (LLR) of the target variable node and the related variable node; a data comparison unit for obtaining the logarithm Probability ratio of one minimum value, one small value, one maximum value and one degree (degree); A correction value obtaining unit for obtaining a correction value based on the minimum value, the sub-smallest value, the maximum value and the quantity; and a data updating unit for updating the logarithm of the target variable node according to the correction value Probably ratio. The communication device as described in item 11 of the patent application scope further includes: a storage unit for storing a look-up table, and the correction value is recorded in the look-up table. The communication device as described in item 11 of the patent application scope, wherein if the minimum value is 0, the correction value is 0. The communication device as described in item 11 of the patent application scope, wherein if the number is 2, the correction value is the minimum value. The communication device as described in item 11 of the patent application scope, wherein if the minimum value is equal to the maximum value, the correction value is obtained based on the minimum value. The communication device as described in item 11 of the patent application scope, wherein if the sub-small value is equal to the maximum value, the correction value is obtained based on the minimum value and the sub-small value. The communication device as described in item 11 of the patent application scope, wherein if the log-likelihood ratio of the target variable node is the sub-small value, and one of the estimated upper limit and the lower estimated limit of the log-likelihood ratios are equal, then The correction value is the estimated lower limit, the estimated upper limit is obtained based on the minimum value and the sub-small value, and the estimated lower limit is obtained based on the minimum value. The communication device as described in item 11 of the patent application scope, wherein if the log-likelihood ratio of the target variable node is not the minimum value or the sub-smallest value, and one of the estimated upper limit and an estimate of the log-likelihood ratio If the lower limit is equal, the correction value is the estimated lower limit, the estimated upper limit is obtained based on the minimum value, the sub-small value and the maximum value, and the estimated lower limit is obtained based on the minimum value and the sub-small value. The communication device as described in item 11 of the patent application scope, wherein the correction value is the average of an estimated lower limit and an estimated upper limit. The communication device according to item 11 of the patent application scope, wherein the correction value is obtained based on the at least one logarithmic likelihood ratio of the at least one related variable node.

Images