KR100317262B1 - Rate matching method for channelization code on up-link - Google Patents

Abstract

TECHNICAL FIELD The present invention relates to a next generation communication system, and in particular, used to adjust channel symbol rate to an optimal level during a coding and multiplexing procedure of a mobile communication system using a wideband code division multiple access (hereinafter, abbreviated as W-CDMA) scheme. A channelization code rate matching method in the uplink.

The present invention provides an optimal rate matching method for the turbo code in the uplink using a puncturing algorithm for adjusting the channel symbol rate to an optimal level during the coding and multiplexing procedures of the mobile communication system. Using a charging algorithm, an optimal rate matching method for channelization codes in the uplink is provided.

Description

Rate matching method for channelization code on up-link}

The present invention relates to a next generation communication system, and more particularly, to a channelization code rate matching method in uplink used to adjust channel symbol rate to an optimal level during coding and multiplexing procedures of a mobile communication system using a W-CDMA scheme. will be.

Recently, ARIB in Japan, ETSI in Europe, T1 in the US, TTA in Korea, and TTC in Japan are the core networks of the existing global system for mobile communications (GSM) that provide multimedia services such as voice, video and data. The next generation of mobile communication system based on wireless access technology was envisioned.

In order to present technical specifications for the next generation evolved mobile communication system, they agreed to joint research, and the project for this was called Third Generation Partnership Project (hereinafter abbreviated as 3GPP).

3GPP includes three major technical research areas.

The first is the 3GPP system and service sector, which is a study of the structure and service capabilities of the system based on the 3GPP specification.

Second, it is a research area for Universal Terrestrial Radio Access Network (UTRAN), where the Universal Terrestrial Radio Access Network (UTRAN) is based on W-CDMA according to Frequency Division Duplex (FDD) mode. Radio Access Network (RAN) using TD-CDMA according to Time Division Duplex (TDD) mode.

Third, it is a research section for core network that has evolved from the second generation mobile communication globalization system (GSM) and has third generation networking capability such as mobility management and global roaming.

In the above-mentioned technical research divisions of 3GPP, the research section for the global radio access network (UTRAN) describes the definition and description of the transport channel and the physical channel, in particular, the S1 series. S1.12 describes definitions and explanations of multiplexing, channel coding, and interleaving according to the frequency division duplex (FDD) mode.

1 is a diagram illustrating a coding and multiplexing procedure in an uplink according to a 3GPP radio access network (RAN) standard, and illustrates a coding and multiplexing procedure for a transport channel.

The data streams on the transport channel arrive at the coding / multiplexing blocks (1, 2) in the form of transport block sets. At this time, each transport block arrives through its own transport channel at regular time intervals.

In this case, CRC bits for cyclic redundancy check (hereinafter referred to as CRC) are added to each transport block arriving at each coding / multiplexing block (1, 2) (S1), and then the same quality of service ( Transmission channels of Quality of Service (hereinafter referred to as QoS) are multiplexed (S2).

CRC is applied to error detection of a transport block, and CRC bits are added to each transport block of all transport channels.

The multiplexed transmission channels of the same QoS are subjected to channel coding (S3), interleaving (S4), and rate matching (S5).

As the channel coding, either convolutional coding or turbo coding is used. In addition, specific coding may be used depending on a service.

In each coding / multiplexing block (1, 2), the transmission channels of different QoS adjusted to an optimal level of channel symbol rate by rate matching (S5) are multiplexed again, and then the data streams are respectively interleaved through the divided physical channels. Is sent.

FIG. 2 is a diagram illustrating a coding and multiplexing procedure in a downlink according to the 3GPP radio access network (RAN) standard. Although the coding and multiplexing procedure for a transport channel in this downlink is similar to the uplink, rate matching (S13). It is different that inserting discontinuous transmission indication bits S14 and interleaving S15 are performed after the.

Even in this downlink, the data stream on the transport channel arrives at the coding / multiplexing blocks 10 and 20 in the form of transport block sets. At this time, each transport block arrives through its own transport channel at regular time intervals.

At this time, CRC bits for CRC are added to each transport block arriving at each coding / multiplexing block (10, 20) (S10), and then the transport channels of the same QoS are multiplexed (S11).

CRC is applied to error detection of a transport block, and CRC bits are added to each transport block of all transport channels.

The multiplexed transmission channels of the same QoS are subjected to channel coding (S12), rate matching (S13), discontinuous transmission indication bit insertion (S14), and interleaving (S15).

As the channel coding, either convolutional coding or turbo coding is used. In addition, specific coding may be used depending on a service.

In each coding / multiplexing block (10, 20), after adjusting to the optimal level of the channel symbol rate by rate matching (S13), the transmission channels of different QoS through interleaving (S15) are multiplexed again, and then the data stream is divided. Interleaved and transmitted through the physical channel.

Of note in the coding and multiplexing procedure in the uplink or downlink according to the 3GPP radio access network (RAN) standard is rate matching.

This rate matching is a process of adjusting the optimal channel symbol rate by applying repetition and puncturing for different transmission channels.

Currently, in the 3GPP radio access network (RAN) standard using the W-CDMA scheme, puncturing schemes for convolutional codes are considered in uplink and downlink, and puncturing schemes for turbo codes are Considered in the downlink.

Here, the purpose of the puncturing technique for the convolutional code is to distribute the positions of the punctured code bits as evenly as possible. A puncturing algorithm that can meet this goal can be easily implemented in the downlink. However, in the uplink, as described above, puncturing is performed on the interleaved bit strings, and thus additional conditions must be added to the puncturing conditions in the downlink to satisfy uniform puncturing in the code bit unit.

Currently, puncturing techniques for convolutional codes in the uplink have been continuously proposed in view of such a point.

As an example, the puncturing procedure for convolutional code in the downlink is very simple. This is because only the condition of puncturing the code bits uniformly needs to be satisfied.

Therefore, after the puncturing distance N is obtained for the entire bit string in the code bit unit, an algorithm for puncturing every Nth bit may be configured.

If the puncturing distance is '5', puncture the (5xn) th bits for the entire bit string for the convolutional code (where n = 1, 2, 3, ...).

However, if the puncturing distance is' 5.5 'rather than an integer, the puncturing procedure when the puncturing distance is' 5' is performed for half of the total punctures, and for the other half, the puncturing distance is' Just execute the puncturing procedure when 6 '.

The puncturing technique for the turbo code is only considered in the downlink.

Unlike the convolutional code, the turbo code has different importance of each code bit. That is, in the turbo code having a specific code rate, systematic bits are relatively important in the decoding process, and the remaining parity bits are less important.

Therefore, in the puncturing technique for the turbo code, puncturing for the systematic bits should be excluded, and an algorithm for puncturing the remaining parity bits is necessary.

All of the puncturing algorithms for turbo code in the downlink under consideration are basically punctured in code symbol units rather than code bits.

The meaning of the code symbol is as follows. First, assume that the code rate of the channel code is 1/3. This code rate means that the channel code block outputs three bits of code bits with one information bit as input. In this case, if the output code bits are represented as x, y, z for convenience, the output string of the channel code block may be expressed using a time index as follows. {X0, y0, z0, x1, y1, z1,. xk, yk, zk, ....} That is, three code bits x0, y0, z0 are output for the 0 th input information bit. At this time, the group of three code bits is expressed as one code symbol. That is, x0, y0, and z0 together constitute the 0th code symbol. For example, one of the puncturing algorithms for turbo codes in the downlink proposed so far is that the puncturing occurs uniformly in code symbol units. After the puncturing position of the symbol is determined, the parity bits are alternately punctured. The other is an algorithm that increases the position where puncturing can occur in code symbol units by a double distance and then punctures parity bits for each code symbol in which puncturing will occur.

The puncturing algorithm for the turbo code described so far is considered in the downlink, and the puncturing procedure in the downlink may be performed independently of the interleaving procedure since the puncturing procedure occurs in the interleaving front end.

However, since the rate matching procedure is performed after the interleaving in the uplink, there are additional considerations in performing the puncturing procedure for the turbo code.

In order to solve these additional matters, an optimal puncturing algorithm for the turbo code in the uplink has to be proposed, and such a puncturing technique for the turbo code has not been considered yet.

In addition, a rate matching method using an integrated puncturing algorithm for a convolution code and a turbo code, which is a channelization code in the uplink, which has been proposed so far, has not been proposed until now.

SUMMARY OF THE INVENTION An object of the present invention has been made in view of the above, and in the uplink using a puncturing algorithm for adjusting the channel symbol rate to an optimal level during the coding and multiplexing procedure of a mobile communication system using the W-CDMA scheme To provide an optimal rate matching method for the turbo code.

It is still another object of the present invention to provide an integrated puncturing algorithm for optimally adjusting channel symbol rates during coding and multiplexing procedures in a mobile communication system using a W-CDMA scheme. It is to provide an optimal rate matching method.

A feature of the channelization code rate matching method in the uplink according to the present invention for achieving the above object is, in the uplink of the mobile communication system using the channelization code, by performing channel coding for a transport channel a plurality of Outputting a bit string in units of code symbols, grouping and interleaving the bit strings in units of code symbols so that uniform puncturing occurs at regular intervals with respect to the output plurality of bit sequences; The puncturing is performed on the coded bits in units of code symbols at a predetermined interval.

Preferably, the puncturing step is performed in units of code symbols at intervals for the remaining bit strings used for decoding, except for a specific bit string output by the channel coding, wherein the puncturing is performed in the channel. In consideration of the importance of each bit string output by coding, puncturing for the specific bit string having a relatively high importance is excluded.

In the puncturing step, a uniform puncturing algorithm of a predetermined interval is applied to each of the interleaved bit strings.

Another feature of the present invention for achieving the above object is that, in the uplink of a mobile communication system using a turbo code, a systematic bit and one or more parity bits are symbolized as a result of channel coding for a transport channel. Outputting in parallel in units, changing the order of the parity bits for each symbol unit outputted, and popping at intervals of each symbol unit with respect to the interleaved parity bits for the output code bit streams. It is a step of performing cheering.

Advantageously, a uniform number of puncturing occurs for each parallel bit string after the puncturing step is interleaved.

In order to achieve the above object, a final feature of the present invention is to output a systematic bit and one or more parity bits in symbol units as a result of channel coding for a transmission channel of a mobile communication system, and outputting The order of the parity bits is changed for each symbol unit, and the puncturing is performed at a predetermined interval of each symbol unit with respect to the interleaved parity bits for the output code bit streams.

1 is a diagram illustrating a coding and multiplexing procedure in the uplink according to the 3GPP Radio Access Network (RAN) standard.

2 is a diagram illustrating a coding and multiplexing procedure in downlink according to the 3GPP Radio Access Network (RAN) standard.

3 is a diagram illustrating a puncturing pattern when applying an optimized puncturing algorithm for a convolution code in uplink to a system using a turbo code.

4 is a view comparing an example of applying a puncturing algorithm for turbo code rate matching in uplink according to the present invention with an example of applying a puncturing algorithm in units of existing code symbols.

5 is a diagram illustrating an example of a sequence of serial bit strings output from a MIL interleaver when the number of columns of the interleaver K = 8.

FIG. 6 illustrates puncturing positions in code symbol units for realizing uniform puncturing in code symbol units for turbo codes in the present invention. FIG.

FIG. 7 illustrates an actual puncturing pattern punctured in code symbol units for channelization codes in the present invention. FIG.

8 is a diagram illustrating a pattern in code symbol units for puncturing only a specific bit (x) in a convolutional code to realize uniform puncturing in code symbol units according to the present invention.

9 is a view showing an actual puncturing pattern puncturing only a specific bit (x) in code symbol units in a convolutional code according to the present invention.

Hereinafter, a preferred embodiment of a channelization code rate matching method in uplink according to the present invention will be described with reference to the accompanying drawings.

In the present invention, the puncturing algorithm for the convolution code in the uplink is implemented to satisfy the following conditions.

First, the positions of punctured code bits are distributed as evenly as possible over the entire code bit string.

Second, the same number of code bits are punctured for each interleaved bit string. This is to allow uniform puncturing to occur even before the interleaved bit strings.

In the present invention, the puncturing algorithm for the turbo code in the uplink is implemented to satisfy the following conditions.

First, puncturing for systematic bits is excluded.

Secondly, for parity bits, puncturing occurs evenly in all parity bits. That is, when considering a turbo code of 1/3 rate, the puncturing amount of the y parity bits and the puncturing amount of the z parity bits are kept to be the same as much as possible.

Third, in the second condition, each interval of puncturing positions in each set of y parity bits and z parity bits is equalized.

In addition to these three conditions, the present invention should consider one of the following conditions. This is because the rate matching procedure is performed on the bit streams interleaved in the uplink.

Fourth, the same puncturing should occur for all interleaved frames. In other words, an even code bit should be punctured for each of the serial bit strings that are the interleaved bit strings.

The present invention proposes a puncturing algorithm for the turbo code in the uplink that satisfies all the puncturing conditions for the turbo code, which is the channelization code.

In addition, the present invention proposes an integrated puncturing algorithm in the uplink that satisfies not only the convolution code, which is the channelization code, but also the puncturing condition for the turbo code.

The following describes a puncturing algorithm for the turbo code in the uplink according to the present invention.

Prior to the description, a system using a turbo code uses two turbo decoders, a MAP decoder, or a Viterbi decoder indicating a soft decision decoding output.

These two turbo decoders are located at both ends of the turbo interleaver, and the code bits input to the first turbo decoder and the second turbo decoder are parity bits.

Therefore, it is very important to adjust so that uniform puncturing of the parity bits occurs on the transmitting side. It is also important to distribute the uniform puncturing for all channel coded bit streams.

Therefore, the present invention proposes a puncturing algorithm that satisfies the puncturing conditions for the turbo code in the uplink listed above.

FIG. 3 is a diagram illustrating a puncturing pattern when the optimized puncturing algorithm for a convolution code in the uplink is applied to a system using a turbo code. In FIG. 3, the code bits of the shaded portion are punctured.

3 shows a case in which the number of columns K of the interleaver is 8, where x represents a systematic bit, y represents a first parity bit, and z represents a second parity bit. As the result of the channel coding for the parity bits, the parity bits are the output of each RSC coder (Recursive systematic convolutional coder).

As can be seen in FIG. 3, puncturing for the systematic bits occurs, and puncturing occurs only for the second parity bits z of the parity bits.

This satisfies the uniform puncturing condition for the entire bit stream, the same puncturing condition for all interleaved frames, but excludes the puncturing condition for the systematic bits, and an equal amount for each parity bit. The puncturing condition of does not meet.

In contrast, the present invention proposes a puncturing algorithm for the turbo code in the uplink that satisfies all four puncturing conditions for the turbo codes listed above.

The present invention modifies the order of the interleaved bit strings in the coding / multiplexing block shown in FIG. 1, and causes the uniform puncturing to occur in each column in addition to the uniform puncturing per code symbol for each row. .

In the present invention, the interleaving order of channel coded bit strings is modified to fill each row of the interleaver memory in the following order. That is, 'x y z x z y x y z x z y... 'Save to the interleaver's memory in order.

The following describes the puncturing procedure for turbo code rate matching in the uplink according to the present invention in more detail. Hereinafter, the procedure to be described may be applied to each case where the number of columns K of the interleaver is 1, 2, and 4, but for convenience, only the case where the number of columns K of the interleaver is '8' will be described.

FIG. 4 is a diagram comparing an example of applying a puncturing algorithm for turbo code rate matching in an uplink with an example of applying a puncturing algorithm in units of existing code symbols.

4A illustrates an example in which an algorithm for puncturing parity bits is alternately applied after determining a puncturing position of a code symbol such that puncturing occurs uniformly in units of code symbols among puncturing algorithms for a turbo code. Here, a case in which a half rate is generated for each code symbol is described. As shown in FIG. 4A, one bit of parity bits is punctured for each code symbol, and a 0th code symbol is shown. For example, a method of puncturing y0 parity and z1 parity bits for the first parity bits is used. In addition, the output sequence of the channel encoder input to the interleaver in the coding / multiplexing block is in the order of 'x y z x y z x y z ...'.

Also, referring to FIG. 4A, among the turbo code puncturing conditions, the same puncturing condition for all frames that have undergone interleaving cannot be satisfied. In other words, the bit is not punctured uniformly for each of the column bit streams that are the interleaved bit streams. In FIG. 4A, puncturing does not occur in the 0th, 2nd, 4th, and 6th column bit streams.

However, in FIG. 4B to which the puncturing algorithm for the turbo code according to the present invention is applied, puncturing occurs in all of the column bit streams. Unlike the conventional puncturing algorithm, the output sequence of the channel encoder in the coding / multiplexing block shown in FIG. 1 is changed to 'x y z x z y x y z x z y... This is because it uses a method of changing the order in order and then filling the interleaver memory.

Here, it is easy to change the order of the output strings of the channel encoder by providing a switch at the output terminal of the encoder.

An example of applying a puncturing algorithm for a turbo code according to the present invention to obtain a code of 1/2 code rate at 1/3 code rate is shown in FIG. 4B. You can see the reason for changing the order of output string.

The following describes an integrated puncturing algorithm for turbo code according to the present invention. Prior to describing the puncturing algorithm according to the present invention, an interleaver used to realize the present invention will be described.

In the present invention, a multi-stage interleaver (hereinafter referred to as MIL interleaver) is used.

The MIL interleaver writes the input bit string into the interleaver memory in units of rows and reads in units of columns to form an interleaved output bit string. The peculiarity of this is that the order of reading the interleaver's memory applies the bit reversing order rule.

For example, if the bit value indicating the column number at which the interleaved bit string is output is 8 bits, the column bit string is not output in the column number '0 1 2 3 4 5 6 7' and the column number bit value is reversed. I.e., '0 (000) → 0 (000)', '1 (001) → 4 (100)', '2 (010) → 2 (010)', '3 (011) → 6 (110)' '4 (100) → 1 (001)', '5 (101) → 5 (101)', '6 (110) → 3 (011)', '7 (111) → 7 (111)' It outputs a serial bit string in the order of '0 4 2 6 1 5 3 7'.

Table 1 below shows the order of the column bit strings output from the MIL interleaver for each interleaver column.

R (k) K R (0) R (1) R (2) R (3) R (4) R (5) R (6) R (7) 2 0 One N.A N.A N.A N.A N.A N.A 4 0 2 One 3 N.A N.A N.A N.A 8 0 4 2 6 One 5 3 7

Table 1 shows the case where the columns of the interleaver are 2, 4, and 8, and R (k) shows the bit reversing result.

FIG. 5 is a diagram illustrating an example of a sequence of column bits output from the MIL interleaver when the number of columns of the interleaver K = 8.

In FIG. 5, the numbers of the shaded portions actually indicate the order in which the column bit strings are output from the interleaver. That is, the first column bit string is output first, the second column bit string is output fifth, the third column bit string is output third, the fourth column bit string is output seventh, and the fifth column is output. The bit string is output second, the sixth column bit string is output sixth, the seventh column bit string is output fourth, and the last eight column bit strings are output eighth. The code symbols are divided into groups of odd code symbols and groups of even code symbols. That is, if the bit input to the channel encoder has a value from 0 to N-1, and the code rate of the channel encoder is assumed to be 1/3, the index of the code bit has a value from 0 to 3N-1. Will be. The index of the code symbol will be from 0 to N-1. At this time, the code symbols are divided into the following two groups. First, the even-numbered code symbol group is a group of {0,2,4,6, ....} th code symbol indices, and the odd-numbered code symbol group is a {1,3,5,7 ...} th code According to the present invention, assuming that the output sequence of the turbo encoder is switched to x, y, z, x, z, y ... according to the present invention, the following output sequence can be obtained. y0, z0, x1, z1, y1, x2, y2, z2, x3, z3, y3, x4, y4, z4, x5, z5, y5, .......,} In the first code symbol group, only the y parity bits and the odd code symbol group are punctured only the z parity bits. The code symbol index to be punctured in each code symbol group is selected to be uniform so as to have a puncturing interval in each code symbol group. FIG. 6 is a code symbol for a turbo code according to the present invention. Code puncturing position in units of code symbols for realizing uniform puncturing in units. In FIG. 6A and FIG. 6B, and the index of code symbols to be punctured in each group of code symbols.

In this case, it is assumed that a turbo code of 1/3 rate is assumed and the total number of code symbols is 96. Therefore, the total number of code bits is 288 (96 * 3). The index of the code symbol has a value from 0 to 95. FIG. 6A illustrates an interleaver configured by extracting an even code symbol index. 6B shows an interleaver configured by extracting an odd code symbol index. And the total amount of puncturing we want is 2 bits for each column, that is, one column of the interleaver consists of 12 code symbols (36 code bits). Suppose you want The positions of puncturing in the first column of FIG. 6A are two indexes among 12 code symbol indices to satisfy uniformity constraints. That is, 16 code symbol indexes and 64 code symbol indexes are selected from among 12 code symbol indexes {0, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88}. It can be seen that the distance between these two indices is 6. In order to obtain the code symbol index to satisfy the equal interval condition, the following simple algorithm is used. Example 1) e = N = 12; (Initial error value) y = 2; (desired number of puncturing for each interleaver column) m = 1; do while m <= 12e = e-2 * yif (e <= 0) thenselect m; e = e + 2yend ifm = m + 1end do Is an increment value from 1 to 12, indicating the number of code symbol indices to select. In other words, when the above algorithm is executed, m = 3 and m = 9 are selected. As a result, among the {0, 8, 16, 24, 32, 40, 48, 56, 64, 72, 80, 88}, the code symbol index 16 and the code symbol index 64 are selected by the above algorithm. On all interleaver columns of 6a, the selected symbol index will be {18, 66} for the second column, {20, 68} for the third column, and {22, 70} for the fourth column. Selecting this symbol index will result in puncturing the y parity bits corresponding to that symbol index, but with this method, the position of the punctured y parity bits is equal to {y16 out of the 96 y parity bits. , y18, y20, y22, y64, y66, y68, y70}. In other words, in terms of equally spaced puncturing, this is not a desirable puncturing pattern, and some other action must be taken. The method used in the present invention is adapted to apply the algorithm of Example 1 to each interleaver column of FIG. 6A. Different initial error values e are to be used, i.e., if the initial error value e is used for each column of the interleaver, the symbol index is selected as an equally spaced condition by applying the algorithm of Example 1 to each column of the interleaver. In this case, the selected symbol indices can be evenly spaced even if they are seen in the original symbol index space before interleaving. For this purpose, the method used in the present invention is applied to each column of the interleaver. Calculate the shifting parameter S with respect to this shifting parameter The initial error value e to be used is calculated. The meaning of the shifting parameter S is as follows. If the shifting parameter S is calculated for a certain column and the value is 1, this means that the selected pattern is shifted by one below the symbol pattern selected in the first column having the shifting parameter of zero. That is, if the third column of FIG. 6a is assumed, the shifting parameter becomes 1, and if the initial error value e to be used in Example 1 is calculated by using (2 * 1 * 2 + 12) mod 24 and 16 is used. , the symbol index is selected when m = 4 and m = 9. This results in symbol index 28 and symbol index 76. In this case, the shifting parameter S to be used in each interleaver column is obtained through the following calculation (calculation algorithm of the shifting parameter S to be used in each interleaver's column) N: Total number of code symbol indexes for each column of the interleaver y: Desired number of punctures q = if (q <= 2) {for (k = 0; k <K / 2; k ++) {k is the serial number of the even or odd code symbol interleaver if (k = even) S [k] = 0; else S [k] = 1;}} else {if q is even q '= q- else q '= qfor (i = 0; i <K / 2; i ++) {k = S [k] = div K / 2}} In the above algorithm, q '= q- i = 0 i = 1 i = 2 i = 3 S [i] 0 4 One 2 e [i] 12 4 16 20 Substituting the above e [i] value into the algorithm of Example 1 and performing the algorithm for each column results in the selection of the shaded portion of FIG. 6A. The code symbol index selected through this process becomes {2, 16, 28, 38, 50, 64, 76, 86} among the symbol indices between 0 and 95, so that the punctured code bits are {y2, y16, y28, y38, y50, y64, y76, y86}. If the same procedure is applied to the odd-numbered code symbol group, the result of FIG. 6B can be obtained, and the punctured code bits are {z3, z17, z29. , z39, z51, z65, z77, z87}. As shown in the above process, the order of output sequence of the channel encoder is first set to x, y, z, x, z, y, x, y, z, ... And then store the code groups in the interleaver, dividing the entire code symbol groups into even and odd code symbols, and for each code symbol group compared to the number of columns of the original MIL interleaver. Shifting parameter S on a symbol interleaver with the number of columns reduced by one half After the calculation, the indexes of the code symbols where puncturing occurs are obtained. Next, the y parity bit corresponding to the code symbol index selected on the even-numbered symbol interleaver and the code symbol index selected on the odd-numbered symbol interleaver are calculated. Punch corresponding z parity bits. Through this process, all the puncturing conditions for the turbo code in the uplink may be satisfied. FIG. 7 illustrates an actual puncturing pattern punctured in code symbol units for the channelization code in the present invention. Figure 7 shows the puncturing pattern on the resulting original first MIL interleaver. A detailed algorithm for obtaining the puncturing pattern of FIG. 7 will now be described. [0039] Each of the columns on the MIL interleaver to which the rate matching process is actually performed on the shifting parameter S values on the respective interleaver columns obtained from FIGS. The mapping method is as follows. The method of obtaining the mapping rule is that the symbol interleaver for even-numbered symbols is for puncturing y parity bits, and the symbol interleaver for odd-numbered symbols is a function of z parity bits. Under the assumption that the data is intended to be equalized, each relation of the MIL interleaver and the even or odd symbol interleaver are compared with each other.

In the above equations, K denotes the number of columns of the MIL interleaver, k denotes the column number of each configured symbol interleaver, and the resulting Q [k] value represents the MIL interleaver that will use the shifting parameters obtained from the kth symbol interleaver. The two equations (Equation 1 and Equation 2) mean the following. First of all, considering the even-numbered code symbol group, the shifting parameters obtained at each column of the group interleaver are actually used. It is used in row 1, row 7, row 5 and row 3 of the MIL interleaver. The number in the uppermost shaded portion of Fig. 6A is a number representing this mapping. Similarly, considering the even-numbered code symbol groups, the shifting parameters obtained from the respective columns of the symbol group interleaver are actually columns 4 and 2 of the MIL interleaver. , Columns 0 and 6 are used. The number of the uppermost shaded portion of FIG. 6B is a number representing this mapping relationship. As a result, the final algorithm configured in consideration of the output order of the first MIL interleaver can be written in two steps as follows. Step 1a: Calculation of the shifting parameters for the order of the output strings of the MIL (for the turbo code) N: Number of code symbols per column of the MIL interleaver = (Nc: number of code bits per column of MIL interleaver) y: desired number of puncturing q = Average puncturing distance

'

for (k = 0; k <K / 2; k ++) {where K is the number of columns in the MIL interleaver

if (k = even)

if q is even

div K / 2

div K / 2}} from the algorithm above Represents the largest integer not greater than *, Represents the smallest integer greater than *. In the above algorithm, S [r] represents the shifting parameter to be used in the rate matching process for the output string of the r-th (r = 0,1,2, ... K-1) MIL interleaver, and R [.] Denotes the bit reversal interleaving rule of MIL interleaving given in Table 1. That is, S [R [(6k + 1) mod K]] means that the shifting parameter obtained on the interleaver for even-numbered code symbols is mapped to each serial number on the MIL interleaver, and then this value is converted to the MIL interleaver. It means to map in order of output.

Step 2: Determine the position to be punctured by the output string of each MIL interleaver S0 = {d1, d2, d3, .., dNc}: set of serial output code bits of the MIL interleaver

e [r] = (2 * S [r] * y + N) mod 2N; Initial error value to be used for the output string of the r th MIL interleaver. If e [r] = 0, use e [r] = 2N.

y = the amount of puncturing to be applied to the output string of each MIL interleaver

m = 1

a [r] = (K (R [r] -1) +11) mod 3do while m <= Ne = e-2 * yif (e <= 0) thenpuncture bit among e = e + 2yend ifm = m + 1end do In the above procedure of step 2, a [r] means that if a puncturing position is selected in code symbol units in the rth output string of the MIL interleaver, it is within one code symbol. Where the puncturing is actually performed, i.e. dividing each MIL interleaver into three code bits, assuming that it is one code symbol, and if the current code symbol is selected, This is the index that determines the number of code bits to puncture in the output of the MIL interleaver. Let's say Where index i is 1 It will have a value between. If you consider the leftmost column of the MIL interleaver, } Will be. In this case, the leftmost column of the MIL interleaver is calculated as a = 0. Therefore, when three code bits of each column of the MIL interleaver are grouped, this indicates that the last third bit of the MIL interleaver is actually punctured. Table 3 summarizes the shifting parameter S, the initial error e value, and the a value of each output string of the MIL interleaver used to obtain the puncturing pattern of FIG. 7. In short r 0 One 2 3 4 5 6 7 i 0 4 2 6 One 5 3 7 S [r] One 0 4 2 0 One 2 4 e [r] 16 12 4 20 12 16 20 4 a [r] 0 2 One 0 2 One 0 2 As a result, for the output strings of each MIL interleaver, the corresponding shifting parameter is calculated, and the initial error value to be used in the rate matching algorithm is calculated using the value, and then the value of a is calculated to calculate the algorithm of step 2. As a result, the puncturing pattern of FIG. 7 is obtained as a result. When the puncturing algorithm for the turbo code in the uplink is performed through the above process, the puncturing for the systematic bit strings is performed as described above. It is possible to exclude the sampling, and to apply an equal amount of puncturing for the z parity, y parity, which is the output parity from each RSC coder, and also seen on the respective y parity columns and z parity columns. You can select puncturing positions that are evenly spaced at the enemy. In addition, all the above conditions can be satisfied by applying the same amount of puncturing for each column on the MIL interleaver. The puncturing process for the turbo code in the uplink through steps 1 and 2 described above. Applies only if K, the number of columns of the MIL interleaver, is greater than 2. If K = 1, it can be considered that there is no MIL interleaver, and accordingly, the puncturing process used in the downlink can be performed. Next, the puncturing process in the uplink to the convolutional code is turbo. As in the case of the code, a method of performing the code symbol unit is described. Basically, the method to be used is to transform the steps 1 and 2 to the convolutional code. Generally, the convolutional code is different from the turbo code. Otherwise, the importance of the code bits constituting the code symbol is considered to be different. Accordingly, the rate matching process for the convolutional code performs the rate matching process in units of code bits and uses the assumption that any code bits constituting the code symbol can be punctured. The puncturing process uses a method that satisfies the equal interval puncturing conditions seen in the entire code bit stream. However, in the case of the currently used 1/3 rate convolution code, when puncturing is performed for each code symbol unit, when puncturing only one code bit in one code symbol, the effect on the total decoding performance is different. There is a good code bit compared to the performance when puncturing the bit. Let us denote a code symbol of x, y, and z in the third rate convolution code currently used in the 3GPP specification. Where x bits are the generated polynomial Is the bit output by y, and the y bit is the generating polynomial , And the z bit Is a bit generated by Among these, puncturing on x bits has less impact on overall decoding performance than puncturing on other y or z bits. A puncturing process in units of code symbols of steps 1 and 2 may be performed, and a method of limiting punctured bits to x bits may be used. For this purpose, in the present invention, the output order of convolutionally coded bit strings is modified to be MIL. When entering the interleaver, use the following order. That is, 'yxzyxzyxzy…' In order. The reason for modifying the output sequence of the convolutional code is that the rate matching process of step 2 is to be used in combination with the turbo code and the convolutional code. We need to calculate the shifting parameters that will be used for each output column of the interleaver. Here, the difference from the step 1 used in the turbo code is the configuration method of the symbol interleaver and the mapping rule of each column of the MIL interleaver. In the case of the turbo code, the parity of the upper RSC and the lower RSC when Because puncturing is performed, the interleaver for odd-numbered symbols and the interleaver for even-numbered symbols are independently configured. However, you do not have to do this for convolutional code.

FIG. 8 is a diagram illustrating a pattern in code symbol units for puncturing only a specific bit (x) in a convolutional code to realize uniform puncturing in code symbol units according to the present invention. Among the constituent bits, only the x bits are desired to be punctured. Under these assumptions, comparing the symbol interleaver's symbol index with the MIL interleaver's index, we can create a mapping rule to determine the serial number of the MIL interleaver that will use the shifting parameter S for each column obtained from the symbol interleaver. This is represented by the shading in the top row of FIG. That is, the shifting parameter obtained from the leftmost column (column 0) of the symbol interleaver of FIG. 8 indicates that the shifting parameter obtained from column 1 of the MIL interleaver is used by the column 4 of the MIL interleaver. Indicates that This rule performs a rate matching process on a code symbol basis for the output string of the MIL interleaver, and is a mapping rule designed to puncture only x bits among selected code symbols. This mapping rule is simply expressed by the following equation. Q [k] = (3k + 1) mod K: k = 0,1,2, ..., K In Equation 3, K denotes the number of columns of the MIL interleaver, k denotes the column number of the configured symbol interleaver, and the resulting Q [k] value corresponds to the shifting parameter obtained from the k-th symbol interleaver. By applying the above mapping rule to the shifting parameter calculation process, the following shifting parameter calculation algorithm can be obtained.

Step 1b: Calculation of the Shifting Parameters for the Order of the Output String of the MIL (for Convolutional Code)

N = number of code symbols per column of MIL interleaver (Nc: number of code bits per column of MIL interleaver)

y: desired puncturing number

q: Average puncturing distance

if (q <= 2) {

Where K is the number of columns in the MIL interleaver

if (k = even)

if q is even

As described above, the integrated puncturing algorithm for the channelization code in the uplink is summarized below.

Prior to starting the puncturing algorithm, which code is used as the channelization code in the turbo code or the convolutional code? Is the number of columns (K) of the MIL interleaver used more than 2? According to the channel coding after the output pattern is determined.

If K is a value greater than 2 and the channel code to be used is a turbo code, the output order of the channel code block is x, y, z, x, z, y, x, y, z, x, z, y, .. In order of convolution code, and in the order of y, x, z, y, x, z, .. The bit stream output after the channel coding is filled in the MIL interleaver memory, and then the interleaved bit strings are applied to the following integrated puncturing algorithm.

'if (number of columns in the interleaver K = 1)

{Apply puncturing algorithm on existing downlink}

else {

if (convolution code)

Calculate the shifting parameter (S) using step 1b for the output strings of each MIL interleaver

else if (turbo code)

Calculate the shifting parameter (S) using step 1a for the output strings of each MIL interleaver

Perform the puncturing algorithm of step 2 for channelization code in the uplink}

As described above, in the rate matching method for the channelization code in the uplink according to the present invention, the puncturing condition for the turbo code in the uplink has a uniform code symbol distance for each parity bit of the RSC coders. Puncturing is realized. In addition, an equal amount of puncturing conditions for each parity bit can be satisfied.

In addition, integrated puncturing algorithms for channelization codes such as turbo codes and convolutional codes can be implemented to achieve optimal rate matching in the uplink.

In addition, puncturing for convolutional codes in the uplink can be realized, and puncturing in code symbol units is performed unlike the conventional puncturing algorithm for convolutional codes. In addition, in the integrated puncturing algorithm according to the present invention, since puncturing occurs for each code symbol, it is possible to realize uniform puncturing to significantly improve overall decoding performance.

Claims (9)

In the uplink of the mobile communication system using the channelization code, Outputting a plurality of bit strings in code symbol units by performing channel coding on a transmission channel; Grouping and interleaving the bit strings in code symbol units so that uniform puncturing occurs in units of code symbols at regular intervals with respect to the outputted plurality of bit strings; Performing puncturing on the interleaved bit streams in units of code symbols at predetermined intervals. The method of claim 1, wherein the puncturing step, Except for the specific bit string output by the channel coding, the rate matching method characterized in that the remaining bit strings used for decoding are performed in units of code symbols at regular intervals. 3. The method of claim 2, wherein the puncturing excludes puncturing for the specific bit string having a relatively high importance in consideration of the importance of each bit string output by the channel coding. The method of claim 1, wherein the puncturing step includes applying a uniform puncturing algorithm at a predetermined interval in each column to the interleaved bit streams. In the uplink of the mobile communication system using the channelization code, Outputting the systematic bits and one or more parity bits in symbol units as a result of channel coding for the transmission channel; Interleaving the parity bits with different priorities for each symbol unit output; Performing puncturing on the interleaved parity bits at a predetermined interval of each symbol unit. The method of claim 5, wherein the puncturing step, And a uniform number of puncturing occurs for each column bit stream after the interleaving. Outputting the systematic bits and one or more parity bits in symbol units as a result of channel coding on a transmission channel of the mobile communication system; Parity bits are grouped and interleaved at different priorities for each output symbol unit; Performing puncturing on the parity bits of each interleaved group at predetermined intervals in each symbol unit. The method of claim 7, wherein the puncturing step, And a uniform number of puncturings occur at regular intervals for each column bit string of each interleaving group. The rate matching method of claim 1, wherein the priority of the bits in the code symbol unit is changed.