RU2235425C2 - Matching method and device in flexible data transfer speed mode for data transfer system - Google Patents

Abstract

FIELD: data transfer systems; matching coded-character frames with multiplier capacitance in response to changes in data transfer speed.

SUBSTANCE: proposed method and device can be used for matching frame having coded characters flexibly detected in response to changes in data transfer speed with multiplier capacitance in data transfer system. Mentioned method involves generation of stream of LM repetitive characters by M-times by repetition of stream of L characters; calculation of first punching interval DI and first quantity PI of characters being punched; calculation of second quantity P2 of characters being punched; calculation of second punching interval D2; generation of stream of N characters by punching stream of LM repetitive characters at first punching interval D1 and second punching interval D2. Matching device for L coded characters has coder for shaping stream of L coded characters, repeater, punching pattern generator, and N-character stream shaper.

EFFECT: optimized performance characteristics without impairing them when matching coded-character frame with multiplier capacitance.

24 cl, 19 dwg

Description

FIELD OF THE INVENTION

The present invention relates generally to a data transmission system and, in particular, relates to a device and method for matching a frame having encoded symbols that are flexibly determined in accordance with a change in the data rate with an interleaver capacity.

State of the art

In a radio communication system, such as a satellite system, an ISDN system (digital integrated services network), a digital cellular system, a W-CDMA (code division multiple access) system, a UMTS system (universal mobile telecommunication system) and an IMT system -2000 (International Mobile Telecommunications-2000), the channel coding scheme mainly uses a convolutional code and a linear block code for which one decoder is used. Symbols encoded according to the indicated channel coding scheme are usually interleaved using a channel interleaver.

A typical channel interleaver was designed to perform interleaving by receiving a frame with the number of coded symbols identical to the interleaver capacity per frame.

However, in a modern channel interleaver of the FDRT type (transmission with a flexible data rate, PHSPD), interleaving is performed by receiving a frame with a number of characters different from the interleaver capacity per frame.

Figure 1 shows a channel interleaver of the non-FDRT type (that is, for the case of transmitting data at a fixed rate), which performs interleaving by receiving a frame with the number of coded symbols identical to the capacity of the interleaver.

Referring to FIG. 1, in non-FDRT mode, when the channel data rate is fixed, the number L of coded symbols per frame arriving at the channel interleaver 100 input is always equal to the interleaver capacity N. For example, the RC configuration (radio configuration), used in the IMT-2000 communication system includes various types of transmission channels, such as RC1, RC2, RC3, RC4, RC5, RC6, RC7, RC8 and RC9, having different data frame sizes, code rate, and interleaving mode. Accordingly, in non-FDRT mode, only a given fixed data rate is used.

Figure 2 shows an example of a frame format of encoded characters transmitted in non-FDRT mode.

We turn to figure 2 and suppose that in the physical channel the data rate of RC3 is set at 19.2 kilobits / s; then the capacity of the channel interleaver 100, shown in figure 1, will be N = 1536. Data transmitted at a speed of 19.2 kilobits / s in 20-millisecond periods will have a speed of 384 bps, and data after a channel encoder with R equal to 1/4 should have a speed of 1536 bps. If at this moment the user wants to transmit a frame with a data transfer rate of 20 kilobits / s, the base station and the mobile station during the initial coordination of the available data rates exceeding the required transmission speed of 20 kilobits / s, will choose a data transfer rate equal to 38.4 kilobits / s. This will happen because a data rate of 38.4 kilobits / s is the lowest data rate that exceeds 20 kilobits / s. When setting the data transfer rate to 38.4 kilobits / s, the capacity of the channel interleaver 100 will double and will be N = 3072 (= 2 × 1536).

When increasing the data transfer rate from 20 kilobits / s to 38.4 kilobits / s, as was established above, the upper level records zero data in an empty interval corresponding to an interval without a period of 20 kilobits / s × 20 ms from data symbols entered into the channel encoder (not shown). That is, a channel interleaver with a capacity of N fills (38.4-20) / 38.4 = 47.92% of its output with zero data before transmission. Thus, from the point of view of the energy Es of the received symbols, it can be considered that 47.92% of the energy is lost. The reason for the energy loss is the inability to process null data at the physical level in a non-FDRT circuit. Even if a symbolic repetition of zero data is performed before transmission, a circuit with a direct additional channel (F-SCH, P-DK) will still have the disadvantage that it is impossible to combine characters. In addition, since the null data varies depending on the input data rate, the upper layer must send null data to the base station and the mobile station in advance. In addition, it is necessary to recover the energy of the null data before the null data passes through the channel decoder, and the upper levels L1 / L2 process only the decoded information symbols after the channel decoder, which degrades the decoding performance.

To solve the above problem and improve the performance of a non-FDRT circuit, an FDRT circuit has been proposed. Intensive research has been conducted on FDRT rate matching techniques to improve channel coding efficiency and to improve the performance of a multi-channel multiple access system that uses a channel coding scheme. The principles underlying the FDRT method are based on the assumption that a convolutional code, a linear block code, or a convolutional code using a cascade code is used as the channel code. In particular, for the air interface IS-2000 (Partnership Program for 3rd Generation Projects), as a standard specification for improving the efficiency of data transmission according to the channel coding scheme and improving the performance of a multi-channel system with multiple access, the FDRT speed matching method was preliminarily determined , and studies are currently underway to implement this method.

Figure 3 shows the structure of the device matching in terms of flexible data transfer rate (FDRT), known from the prior art.

Before describing FIG. 3, we refer to the various terms used hereinafter in Table 1. Namely, c [n], d [n], f [n] and r [n] in FIG. 3 indicate data symbols defined in table 1. Here, the "symbol" is expressed with a single bit that takes the value "1" or "0". In general, a character contains one or more bits. However, here, each data bit expressed by one bit is defined as a “symbol”.

In Table 1, c [n] indicates encoded characters coming from a channel encoder (not shown), and r [n] indicates coded characters repeated by repeater 110. In addition, f [n] indicates encoded characters punctured by a puncturer 120 from among repeating coded characters, and f [n] indicates coded characters interleaved by an interleaver 100 of the punctured coded characters. The channel encoder outputs a stream (or sequence) of L encoded symbols. Repeater 110 repeats L coded symbols M times and provides LM symbols. The puncturer 120 punctures P characters from among the LM repeating coded characters and results in N characters processed according to the FDRT scheme. The channel interleaver 100 interleaves a stream of N symbols processed by the FDRT scheme.

For reference, since in the FDRT scheme L≤N, input coded symbols are always repeated. This is because the FDRT circuit is designed to ensure that the input data rate matches the IS-2000 channel interleaver capacity. Therefore, the FDRT circuit includes a puncturer used to match the capacitance of the interleaver N = LM-P after repetition, so that the number of transmitted symbols is much greater than the number L of encoded symbols.

Refer to figure 3, where with the number L of coded symbols less than the capacity N of the channel interleaver, the repeater 110 repeats the encoded symbols M times. In the case of using the IS-2000 system, since the channel interleaver capacity increases / decreases by a factor of 2 in accordance with the expansion coefficient (SF, КР), M will be at least 2. Since the number of coded symbols repeated by the repeater 110 is greater than N, the puncturer 120 performs puncturing in order to match the number of repeated coded symbols with the capacity N of the channel interleaver 100.

FIGS. 4A-4D show a frame format of encoded symbols re-formed by a repeater 110 and a puncturer 120 in a flexible data rate matching (FDRT) matching apparatus shown in FIG. 3.

In particular, FIG. 4A shows L coded symbols in one frame, and FIG. 4B shows LM coded symbols repeated M times by the repeater 110. In addition, FIG. 4C shows LM coded symbols, where N coded symbols must be interleaved with channel interleaver 100, and the LM-N encoded characters must be punctured by the puncturer 120. Here, the LM-N encoded characters are distributed so that the characters must be punctured uniformly at intervals of length D. Finally, FIG. 4D shows coded characters after alyvaniya, wherein the resulting encoded symbols are provided to a channel interleaver 100 for performing channel interleaving.

Referring to FIGS. 4A-4D, where a newly formed coded symbol frame is compared to a non-FDRT coded symbol frame shown in FIG. 2. In the FDRT scheme, there is no null data in the frame, and each character is treated as an encoded character. By using the FDRT scheme, unlike the non-FDRT scheme, the receiver can increase the energy of the encoded symbol received at the same transmit power. The energy of encoded symbols refers to the energy of encoded symbols after combining symbols. In this way, it is possible to reduce the transmit power of the base station necessary to guarantee a specified quality of service (QoS), which will lead to an increase in channel capacity.

4C, blackened blocks indicate punctured characters, and “D” indicates puncturing distance. The puncturing distance D is a parameter that determines the puncturing method, which provides the issuance of N characters from among LM characters. To specify the relationships between the parameters L, M, N, P and D, the FDRT algorithm is used.

Table 2 below shows the FDRT algorithm defined in the IS-2000 specification. In the following description of the FDRT algorithm, for convenience of explanation, the original terminology borrowed from the original document is used.

As shown in the FDRT algorithm in Table 2, the parameter D is determined from the given parameters L and N, and then each D-th coded character is punctured from the first coded character using the determined value of the D parameter, resulting in a complete puncturing P = LM-N coded characters. However, since the FDRT scheme does not take into account the features of the convolutional code described below, a problem may arise in the performance degradation.

For a channel coding scheme, a convolutional code and a linear block code are typically used using a single decoder. In this case, to increase the efficiency of data transmission according to the channel coding scheme and improve the performance of a multi-channel system with multiple access using the channel coding scheme, it is necessary to fully take into account and reflect the following conditions in the process of puncturing according to the FDRT scheme.

Condition (1). A sequence of input characters is punctured using a puncture pattern having a specific period.

Condition (2). The number of punctured bits of input characters is minimized as much as possible.

Condition (3). The encoded characters provided by the encoder are punctured using a uniform puncturing pattern.

The listed conditions are based on the assumption that the sensitivity of the encoded symbols coming from the channel encoder to errors is almost the same for each symbol in one frame (or codeword). When the data is actually transmitted in FDRT mode, you can get a positive result using the above conditions as the main limiting factors when puncturing. However, in most cases, the FDRT scheme in the IS-2000 system does not satisfy these conditions.

FIG. 5 shows how the FDRT device of FIG. 3 punctures coded symbols before transmission. In particular, FIG. 5 shows a puncturing pattern used in transmitting 15 kilobits / s characters at an RC3 data rate of 19.2 kilobits / s in FDRT mode. That is, Fig. 5 is a diagram explaining a problem that may occur if the above conditions are not met. The parameters used in figure 5 are shown below in table 3.

Turning to FIG. 5, it is noted that puncturing is actually performed only in the first 1728 bits of the encoded symbol frame and is not performed in the next 672-bit frame interval. 5, for information, blackened blocks indicate punctured characters, and dotted blocks indicate 672 characters that are repeated twice before being transmitted. The first 1728 double-repeated characters are transmitted selectively through the character. In this method, N = 1536 (= 864 + 672) characters are formed (or re-formed). A format having N = 1536 bits in a frame violates the above condition (3). Therefore, when using the FDRT circuit due to uneven puncturing, a problem may arise in the performance degradation.

Figure 6 presents a diagram revealing the problem that is characteristic of the known FDRT scheme. In particular, FIG. 6 shows the distribution of symbol energy and the number of symbols in a single frame at the terminal stage of the receiver.

Referring to FIG. 6, where the channel receiver 200 receives the characters transmitted in the FDRT mode, and supplies the received characters to the character erasing and combining unit 210. 6 shows the relative energy distribution of Es symbols for respective symbols when the symbol combining unit 210 performs combination of the received symbols. As shown in FIG. 6, if the energy Es of 864 non-repeating symbols is taken as 1.0, then for the next 672 repeating symbols subjected to combination at M = 2, Es becomes 2.0. Therefore, the final (tail) symbols have an average gain Es / No = + 3 dB under the same channel conditions. That is, the channel decoder 220 with R = 1/4 decodes the unevenly distributed 1200 characters and provides 300 information symbols. As described below with reference to FIGS. 12 and 13, the simulation results show that the performance of the known FDRT device is significantly degraded.

указывающее максимальное целое число, меньшее LM/P. В этом случае в действительности прокалывается только P×D символов, и прокалывание не выполняется на оставшемся интервале из P×(LM/P-D) битов. Например, на фиг.5, поскольку LM/P=2400/864=2,778, то D=2 и LM/P-D=0,778. Следовательно, прокалывание выполняется на интервале P×D=864×2=1728 битов, в то время как на интервале P×(LM/P-D)=864×0,778=672 бита прокалывание не выполняется. В результате возникает неравномерное прокалывание из-за разности (LM/P-D) в процессе определения значения D.Uneven puncturing occurs at a certain value of D, which defines the puncturing pattern. That is, when the LM / P ratio is not an integer, the well-known FDRT algorithm in the IS-2000 system determines the value of D as

indicating the maximum integer less than LM / P. In this case, in reality, only P × D symbols are punctured, and puncturing is not performed on the remaining interval of P × (LM / PD) bits. For example, in FIG. 5, since LM / P = 2400/864 = 2.778, then D = 2 and LM / PD = 0.778. Therefore, puncturing is performed on the interval P × D = 864 × 2 = 1728 bits, while on the interval P × (LM / PD) = 864 × 0.778 = 672 bits, puncturing is not performed. As a result, an uneven puncture occurs due to the difference (LM / PD) in the process of determining the value of D.

The known FDRT scheme has the following disadvantages.

1. For an FDRT scheme using a convolutional code or a linear block code, a uniform puncturing scheme is necessary to ensure almost the same error code sensitivity of the code symbols generated by the channel encoder for each symbol in one frame (or codeword). However, since this assumption is not satisfied in the existing ISDR 2000 FDRT scheme, it is necessary to modify the existing FDRT scheme.

2. In terms of symbol repetition, the existing FDRT scheme for the IS-2000 system considers the FDRT scheme as a repetition scheme, assuming that the puncturing pattern is not subject to major changes. However, this should be interpreted in the same context as piercing. That is, for an FDRT scheme with optimal characteristics, even in the case of repetition, a uniform repetition scheme should be used to ensure almost the same sensitivity of the encoded symbols generated by the channel encoder for errors for each symbol in one frame (or codeword). However, since this assumption is not valid in the existing ISDR 2000 FDRT scheme, it is necessary to modify the existing FDRT scheme.

SUMMARY OF THE INVENTION

Thus, it is an object of the present invention to provide an apparatus and method that guarantees optimal performance without deterioration when matching a frame with coded symbols that are flexibly defined in accordance with a change in the data rate with the interleaver capacity in the data transmission system.

Another objective of the present invention is to provide a device and method for transmitting data in a flexible data transfer rate (FDRT) mode, which provide flexible operation in accordance with the data transfer rate by simply adjusting the device structure and initial settings in a data transfer system using a convolutional code or linear block code.

To solve the above and other problems, a method for generating a stream of N symbols by piercing a stream of repeated symbols in a system comprising an encoder for generating a stream of L symbols is proposed; a repeater for repeating a stream of L symbols and a puncturer for puncturing a stream of repeated symbols and forming a stream of N symbols, where N is greater than L. The method consists in generating a stream of LM repeating symbols by repeating a stream of L symbols M times, where M is minimum integer greater than N / L; calculating a first puncture interval D1 defined as a minimum integer greater than LM / P for the number P = LM-N of punctured characters and a first number of punctured characters P1 defined as a maximum integer less than LM / D1; calculating the second number P2 of punctured characters indicating the difference between the number P of punctured characters and the first number P1 of punctured characters, and the second puncturing interval D2, defined as sD1 for one selected integer s of integers less than or equal to the maximum integer less than P1 / P2; and forming a stream of N symbols by puncturing a stream of LM repeating symbols with a first puncturing interval D1 and a second puncturing interval D2.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other objects, features and advantages of the present invention will become more apparent from the following detailed description, together with the accompanying drawings, in which:

figure 1 is a diagram illustrating a known channel interleaver type non-FDRT;

2 is a diagram illustrating a frame format of encoded symbols transmitted in accordance with a non-FDRT mode;

3 is a diagram illustrating the structure of a known matching device in a flexible data rate mode;

4A-4D are diagrams illustrating a frame format of encoded symbols re-formed by a repeater and puncturer in a matching device in a flexible data rate mode shown in FIG. 3;

FIG. 5 is a diagram showing an example in which coded symbols are punctured by the FDRT matching apparatus shown in FIG. 3;

6 is a diagram disclosing a problem encountered in the known FDRT scheme, and illustrating the distribution of symbol energy and the number of symbols in a single frame in the terminal stage of the receiver;

7 is a diagram illustrating a method for puncturing coded symbols in accordance with the puncturing pattern proposed in the present invention;

8A and 8B are diagrams illustrating the distribution of symbol energy and the number of symbols in a single frame in a terminal stage of a receiver associated with a flexible data rate matching device according to an embodiment of the present invention;

Fig. 9 is a flowchart illustrating a procedure for performing coordination operations in a flexible data and data rate mode according to a first embodiment of the present invention;

10 is a diagram illustrating a structure of a flexible data rate matching device according to a first embodiment of the present invention;

11 is a diagram illustrating another structure of a flexible data rate matching device according to a first embodiment of the present invention;

12 is a flowchart illustrating a procedure for performing coordination operations in a flexible data and data rate mode according to a second embodiment of the present invention;

13 is a diagram illustrating a structure of a flexible data rate matching device according to a second embodiment of the present invention;

14 is a diagram illustrating another structure of a flexible data rate matching device according to a second embodiment of the present invention;

FIGS. 15 and 16 are diagrams for comparing simulation results of matching operations in the flexible data and data rate mode proposed in the present invention with simulation results according to the prior art.

DETAILED DESCRIPTION OF A PREFERRED EMBODIMENT

Below, with reference to the accompanying drawings, a preferred embodiment of the invention is described. In the following description, well-known functions or constructions are not described in detail so as not to obscure the invention with unnecessary details.

The present invention provides an improved FDRT scheme capable of providing uniform puncturing or repetition, which solves the problem encountered in the known FDRT scheme. For this purpose, it is necessary to have a uniform puncture pattern or a uniform repetition pattern. Thus, the present invention provides a method for generating a new puncturing pattern for FDRT, followed by puncturing the coded symbols in accordance with the proposed new puncturing pattern.

указывающее максимальное целое число, меньшее LM/P. Следовательно, в этом случае в действительности прокалывается только P×D битов, а на остальном периоде из Р×(LM/P-D) битов прокалывание не производится. Например, поскольку LM/P=2,778, то D=2 и LM/P-D=0,778. Следовательно, прокалывание выполняется на интервале из P×D=864×2=1728 битов, в то время как на интервале P×(LM/P-D)=864×0,778=672 бита прокалывание не производится. В результате из-за наличия разности (LM/P-D) в процессе определения значения D получается неравномерное прокалывание. Для разрешения этой проблемы вводятся следующие базовые условия, а затем описывается алгоритм, основанный на этих условиях.When performing uniform puncturing or even repetition in the FDRT scheme, it is important to determine the correct puncturing distance D. In other words, the cause of uneven puncturing or repetition is parameter D, which defines the puncture pattern or repetition pattern. That is, in the well-known FDRT algorithm in the IS-2000 system, if LM / P is not an integer, then the number D is determined as the parameter

indicating the maximum integer less than LM / P. Therefore, in this case, in reality, only P × D bits are punctured, and on the rest of the period of P × (LM / PD) bits, puncturing is not performed. For example, since LM / P = 2.778, then D = 2 and LM / PD = 0.778. Therefore, puncturing is performed on the interval of P × D = 864 × 2 = 1728 bits, while on the interval P × (LM / PD) = 864 × 0.778 = 672 bits, puncturing is not performed. As a result, due to the presence of a difference (LM / PD), uneven puncturing is obtained in the process of determining the value of D. To solve this problem, the following basic conditions are introduced, and then an algorithm based on these conditions is described.

FDRT condition (1). P × D, determined from L and N, must satisfy the inequality P × D≥LM. That is, D must satisfy the inequality D≥LM / P. Here P and D are integers.

символов, определяемых исходя из значения параметра D, удовлетворяющего условию FDRT (1). Здесь определенная позиция символа не должна перекрываться с позицией, определенной параметром D, удовлетворяющим условию FDRT (1).FDRT condition (2). On LM characters, punctured or repeated as evenly as possible (or at regular intervals)

characters determined based on the value of parameter D satisfying the FDRT condition (1). Here, the specific position of the symbol should not overlap with the position defined by parameter D satisfying the FDRT condition (1).

FDRT condition (3). Uneven repetition or puncturing due to the difference (LM / P – D) in the process of determining the parameter D should be minimized.

The following describes the operation of the FDRT circuit according to an embodiment of the present invention in view of the above FDRT conditions. First, an embodiment is described to which the FDRT algorithm of the present invention is applicable. The following describes a generic FDRT algorithm.

New transmission algorithm in the mode of flexible data rate type 1.

An embodiment to which the FDRT algorithm of the present invention is applied is described below. The conditions used in this embodiment are shown below in table 4, and the algorithm itself is shown further in table 5.

Количество P прокалываемых кодированных символов определяется вычитанием значения емкости перемежителя N из количества LM повторяющихся кодированных символов. Расстояние D прокалывания определяется как

Refer to table 4, in which an embodiment of the present invention is used in RC3 in an IS-2000 system. The maximum assigned data rate is 19.2 kilobits / s, the interleaver capacity is N = 15.36 bits, and the input data rate is 15 kilobits / s. In addition, the number of encoded characters per frame is L = 1200 bits. Therefore, the number of repetitions for L = 1200 encoded characters is M = 2. The number of repetitions M is defined as the minimum integer greater than N / L (= (interleaver capacity) / number of coded symbols per frame)). That is, the number of M repetitions is defined as

The number P of punctured coded symbols is determined by subtracting the interleaver capacitance value N from the number of LM repeated coded symbols. The piercing distance D is defined as

Refer to table 5, where in the algorithm according to the present invention, “k mod (?) 3” indicates the operation modulo 3 of calculating the remainder, determined by dividing k by 3. Condition FDRT (1) is used in the calculation process D, and condition FDRT (2 ) is used in the process with the variable '36'.

7 shows a method for puncturing coded symbols in accordance with the puncturing pattern proposed in the present invention. This example is based on the conditions described in table 4 and the algorithm in table 5.

It should be noted that according to FIG. 7, puncturing is indeed performed uniformly over the entire frame interval of the encoded symbols. 7, blackened blocks indicate punctured characters. In addition, note that the characters transmitted after double repetition and the character selected from among the characters transmitted after double repetition are evenly distributed. Therefore, a frame with N = 1536 has a character format satisfying the FDRT condition (3). Therefore, in the indicated FDRT scheme, due to uniform puncturing, the performance characteristics do not deteriorate and are close to optimal.

On figa and 8B shows the distribution of the energy of the characters and the number of characters in a single frame in the terminal stage of the receiver associated with the matching device in the flexible data rate mode according to an embodiment of the present invention.

Referring to FIGS. 8A and 8B, in which the channel receiver 200 receives the characters transmitted in the FDRT mode according to the invention and provides the received characters to the character erasing and combining unit 210. The symbol combining unit 210 outputs 1200 symbols, as shown in FIG. 8A, with the output symbols having a relative energy distribution shown in FIG. As shown in this figure, if the energy Es of 864 non-repeating characters is taken as 1.0, then for the next 672 repeated characters that are combined at M = 2, we get an Es value of 2.0. On figv shows that the characters are distributed evenly throughout the interval. The uniform distribution of characters improves the performance of the channel decoder 220, which is usually used as a Viterbi decoder.

Generalized GFDRTA-I transmission algorithm in flexible data rate mode.

The following describes a generalized flexible data rate transmission algorithm according to the present invention. The FDRT algorithm and the parameters used in this algorithm are defined in table 6.

То есть количество повторений M определяется как минимальное целое число, большее N/L. Следовательно, количество P прокалываемых кодированных символов определяется как P=LM-N.In Table 6, L indicates the number of encoded symbols per frame from the encoded symbol stream at the encoder output. Further, N indicates a predetermined channel interleaver capacity, which is defined as a value greater than or equal to the number L of encoded symbols per frame. In addition, M indicates the number of repetitions of encoded characters and is defined as

That is, the number of repetitions of M is defined as the minimum integer greater than N / L. Therefore, the number P of punctured coded symbols is defined as P = LM-N.

According to the first embodiment, in the algorithm of Table 6, if P = 0, puncturing is not performed. During the puncturing process, each D1th character and each (D2 + 1) th character (where D2 is an even number) of LM coded characters are punctured until P characters are punctured per unit frame. That is, when the LM encoded characters are arranged in order from 1 to LM, the D1th, 2D1st, 3D1st, ... coded characters and (D2 + 1) -th, (2D2 + 1) -th, ( 3D2 + 1) -th, ... encoded characters (where D2 is an even number). Here (D2 + 1) -th, (2D2 + 1) -th, (3D2 + 1) -th, ... coded characters are punctured so that there is no overlap with mD1-coded characters (where m = 1, 2, 3, ...) in terms of piercing positions. Therefore, if it is necessary, then we can consider another method that prevents overlapping of the (D2 + 1) th, (2D2 + 1) th, (3D2 + 1) th, ... encoded characters with mD1 encoded characters ( where m = 1, 2, 3, ...). For example, instead of the (D2 + 1) th, (2D2 + 1) th, (3D2 + 1) th, ... encoded characters, you can also puncture the (D2-1) th, (2D2-1) th , (3D2-1) th, ... encoded characters (where D2 is an odd number). And in this case, the (D2-1) th, (2D2-1) th, (3D2-1) th, ... coded characters are punctured so that they do not overlap with mD1 coded characters (where m = 1, 2, 3, ...) from the point of view of piercing positions. In other words, D1 and D2 indicate puncture distance values for determining distances among P characters to be punctured from LM repeating coded characters. The values D1 and D2 used here are defined below by equation (1).

)-й символ из LM кодированных символов, пока не будет проколото P символов на единичный кадр. То есть, когда LM кодированных символов расположены в порядке от 1 до LM, прокалываются D1-й, 2D1-й, 3D1-й,... кодированные символы и (

)-й, (

)-й, (

)-й, ... кодированные символы. Здесь (

)-й, (

)-й, (

)-й, ... кодированные символы прокалываются таким образом, чтобы не было перекрытия с mD1-ми кодированными символами (где m=1, 2, 3, ...) с точки зрения позиций прокалывания. Используемые здесь значения D1 и D2 также определяются ниже уравнением (1).According to a second embodiment of the invention, in the algorithm of Table 6, if P = 0, puncturing is not performed. In the process of piercing, each D1-th character and each ((multiple of D2) -

) th character from LM encoded characters until P characters are punctured per unit frame. That is, when the LM encoded characters are arranged in order from 1 to LM, the D1th, 2D1st, 3D1st, ... encoded characters and (

) -th, (

) -th, (

) th ... encoded characters. Here (

) -th, (

) -th, (

) th, ... coded characters are punctured so that there is no overlap with mD1 coded characters (where m = 1, 2, 3, ...) in terms of puncture positions. The values D1 and D2 used here are also defined below by equation (1).

Equation (1)

для P>0: в противном случае прокалывание не требуется.

for P> 0: otherwise, piercing is not required.

P2 = P-P1

D2 = sD1 for P2> 0: otherwise, piercing is not required.

In equation (1), s indicates the maximum integer of integers in the range satisfying equation (2) below.

Equation (2)

We turn to equation (1) and (2), where the puncture distance (or interval) D1 is defined as the minimum integer greater than LM / P for the number P = LM-N of the remaining characters to be punctured. P1 indicates the number of punctured characters and is defined as the maximum integer less than LM / D1. P2 indicates the number of punctured characters determined by the difference between the total number P of characters to be punctured and the number P1 of punctured characters. The piercing distance D2 is defined as sD1 for an integer ‘s’ of integers less than or equal to the maximum integer less than P1 / P2.

), либо отрицательному значению для значения, определяемого путем прибавления D2 к максимальному целому числу, меньшему D1/2 (то есть смещение=-

). То есть для потока из LM повторяющихся кодированных символов сначала прокалываются P1 символов, расположенных с первым интервалом D1 прокалывания от начального символа, а затем прокалываются P2 символов, расположенных со вторым интервалом D2 прокалывания плюс смещение (D2+смещение) от начального символа. Значения первого интервала D1 прокалывания и второго интервала D2 прокалывания определяют шаблоны, используемые для прокалывания символов, равномерно распределенных в одном кадре. Следовательно, в процессе первого прокалывания выполняется относительно частое прокалывание в потоке повторяющихся кодированных символов, составляющих один кадр, а в процессе второго прокалывания выполняется относительно редкое прокалывание в потоке повторяющихся символов.In table 6 and equations (1) and (2), for matching a stream of L (<N) encoded symbols with a capacity of N interleaver, a stream of L encoded symbols is repeated M times, resulting in a stream of LM encoded symbols, and this stream of The LM of the repeated coded symbols is punctured with a first puncture interval D1 and a second puncture interval D2 in accordance with the first puncture pattern A and the second puncture pattern B. Here, the first puncture pattern A is determined as a multiple of the first puncture distance (interval) D1, and the second puncture pattern A is determined as a multiple of the second puncture distance (interval) D2 plus offset. In the first embodiment, the offset is +1 or -1 (offset = ± 1). In the second embodiment, the offset is either equal to the value determined by subtracting D2 from the maximum integer less than D1 / 2 (i.e., the offset = -

), or a negative value for the value determined by adding D2 to the maximum integer less than D1 / 2 (i.e. offset = -

) That is, for a stream of LM repeating coded symbols, P1 symbols located at the first puncturing interval D1 from the initial symbol are punctured first, and then P2 symbols located at the second puncturing interval D2 plus the offset (D2 + offset) from the initial symbol are punctured. The values of the first puncture interval D1 and the second puncture interval D2 define patterns used to puncture characters uniformly distributed in one frame. Therefore, in the first puncturing process, relatively frequent puncturing in the stream of repeated coded symbols constituting one frame is performed, and in the second puncturing process, relatively rare puncturing in the stream of repeating characters is performed.

In other words, P1 characters are punctured for a stream of LM repeating coded symbols, and when the number of coded characters remaining after puncturing is greater than the interleaver capacity N, P2 characters are punctured for a stream of (LM-P1) repeating coded characters. As described above, it is contemplated that in this embodiment of the present invention, puncturing a stream of repeated coded symbols is performed in two separate steps. This is because, although the number of coded symbols is less than the interleaver capacity, it is possible to match the number of coded symbols with the interleaver capacity by puncturing the repeated coded symbols in two separate steps. Thus, depending on the specific circumstances, it is also possible to generate coded symbols in an amount matched with the capacity N of the interleaver using only one step.

FIG. 9 shows a procedure for performing coordination operations in a flexible data and transmission rate mode shown in Table 6 according to a first embodiment of the present invention.

Turning to FIG. 9, in step 401, the initial parameters N, L, M, and P necessary to implement FDRT are initialized. The number L of encoded characters constituting the frame and the interleaver capacity N are determined in accordance with a given data rate, while the number of repetitions M and the number P of characters to be punctured are determined by the formula in table 6. In step 402, the first puncture interval D1 is calculated. and the first number P1 of punctured characters in accordance with the formula specified in the algorithm. At operation 403, a second puncture interval D2 and a second number P2 of punctured characters are calculated in accordance with the formula given in the algorithm. After all parameters are calculated at operations 402 and 403, operations 404 through 411 are performed with a sequential increment of k from 1 to LM. At each stage of the increment k, if it is determined at operations 405 and 406 that k is a multiple of D1 or D2 (where D2 is an even number), or if at operations 405 and 408 it is determined that k is a multiple of D1 or D2 (where D2 is an odd number), then, at operation 407 or 409, the corresponding kth encoded symbols are punctured. At operation 405, it is determined whether D2 is an odd or even number. If at operation 405 it is established that D2 is an even number, then at operation 406 it is determined that the number k is a multiple of k to the number D1 or D2. If at operation 40 it is established that k is a multiple of D1, then at operation 407 the kth coded symbol is punctured; otherwise, if it is determined that k is a multiple of D2, then at step 407, the (k + 1) th coded symbol is punctured. However, if it is determined at step 406 that k is not a multiple of either D1 or D2, then the procedure proceeds to step 410 to increase the value of k by +1. If it is established at operation 405 that D2 is not an even but an odd number, then at operation 408 it is determined whether the number k is a multiple of the number D1 or the number D2. If at operation 408 it is established that k is a multiple of D1, then at operation 409, the kth coded symbol is punctured; otherwise, if it turns out that k is a multiple of D2, then at step 409, the (k-1) th coded character is punctured. However, if at operation 408 it turns out that k is not a multiple of either D1 or D2, then the procedure proceeds to operation 410 to increase the value of k by +1. After operation 410, at operation 411, it is determined whether the equality k = LM + 1 is satisfied. If this equality holds, then the process ends. If not, then operations 405 through 411 are repeated until, at operation 411, it is determined that k = LM + 1. With this method, an FDRT puncturing pattern is created that is close to uniform, and puncturing is performed in the stream of repeated coded symbols in accordance with the puncturing pattern created.

When performing operations 401-407, 410 and 411, if it is determined that k is a multiple of D1 or a multiple of D2 plus 1 (where D2 is an even number), then the corresponding kth coded character is punctured. In operations 401–405 and 408–411, if k is a multiple of D1 or a multiple of D2 minus 1 (where D2 is an odd number), then the corresponding kth coded character is punctured. This is done so that the puncturing is performed in positions that do not coincide with the coded characters corresponding to the number multiple of D1. That is, coded characters corresponding to a multiple of D2 plus 1 (where D2 is an even number), or multiples of D2 minus 1 (where D2 is an odd number), are punctured at other positions that do not match the coded characters punctured at positions corresponding to a multiple of D1.

10 and 11 show structures of matching devices in a flexible data and transmission rate mode according to a first embodiment of the present invention. In particular, figure 10 shows the structure of the hardware implementing the FDRT algorithm, and figure 11 shows the structure of the software tools of the FDRT algorithm. That is, the FDRT device according to the first embodiment of the present invention can be implemented either using a programmable module, such as a digital signal processor (DSP), a central processing unit (CPU, CPU), and a microprocessor unit (MPU), as shown in Fig.11, or in the form of a hardware module, such as a specialized integrated circuit (ASIC, ASIC), as shown in Fig.10.

Referring to FIG. 10, a flexible data rate matching device according to an embodiment of the present invention includes a channel encoder 10, a repeater 110, a puncturer 350, a channel interleaver 100, a symbol index generator 310, modular operation blocks 320 and 330 and an OR gate (or logical summation block) 340.

Прокалыватель 350 выполняет операцию прокалывания в соответствии с сигналом разрешения прокалывания PUNC_EN от логического элемента ИЛИ 340. То есть сигнал разрешения прокалывания PUNC_EN является шаблоном прокалывания для определения операций прокалывания прокалывателя 350. Канальный перемежитель 100 с емкостью N выполняет перемежение N-символьного потока, выходящего из прокалывателя 350.Channel encoder 10 creates a stream of L encoded symbols. The repeater 110 repeats the stream of L coded symbols M times and provides LM repeated coded symbols. Here, M indicates the number of repetitions of a stream of L encoded symbols and is defined as a minimum integer greater than N / L. I.e

The puncturer 350 performs the puncturing operation in accordance with the puncturing enable signal PUNC_EN from the OR logic element 340. That is, the puncturing enable signal PUNC_EN is a puncturing pattern for determining piercing operations of the piercer 350. The channel interleaver 100 with a capacity of N interleaves the N-symbol stream exiting the piercer 350.

The symbol index generator 310 sequentially creates indexes indicating the symbols that form the stream of LM repeating symbols. Character index generator 310 may be implemented using a counter. The modulo operation unit 320 obtains the index k generated by the symbol index generator 310 and D1 and generates a puncture enable signal PUNC_EN of ‘1’ when the kth encoded symbol corresponds to the encoded symbol in the puncture position. For example, in the modular operation block 320, the condition “when the kth encoded symbol corresponds to the encoded symbol at the puncture position” refers to the case where the kth encoded symbol corresponds to a multiple of D1. The modulo operation unit 330 receives the index k generated in the symbol index generator 310 and D2 and generates a puncture enable signal PUNC_EN equal to '1' when the kth encoded symbol corresponds to the encoded symbol in the puncture position. For example, in operation module 330, the condition "when the kth encoded symbol corresponds to the encoded symbol in the puncture position" refers to the case when the kth encoded symbol corresponds to a multiple of (D2 + 1) (where D2 is an even number), or a multiple of (D2-1) (where D2 is an odd number). The OR gate 340 generates a puncture enable signal PUNC_EN by performing a logical OR operation on the output signals of the operation blocks 320 and 330 modulo and supplies the generated puncture enable signal PUNC_EN to the puncturer 350.

P2=P-P1, D2=sD1 и

Интервалы D1 и D2 прокалывания, а также количества P1 и P2 прокалываемых символов подаются из блока определения шаблона прокалывания (не показан). Блок определения шаблона прокалывания, блоки 320 и 330 операций по модулю и логический элемент ИЛИ 340 служат в качестве генератора шаблона прокалывания формирования сигнала разрешения прокалывания, определяющего операцию прокалывания прокалывателя 350.As described with reference to Table 6, equations (1) and (2) and FIG. 9, D1 and D2 are puncture interval values defining an interval between characters punctured in a coded symbol stream within a single frame. The first puncture interval D1 is defined as the minimum integer greater than LM / P for the number of punctured characters P = LM-N. The second piercing interval D2 is defined as sD1 for the selected integer 's' of integers less than or equal to the maximum integer less than P1 / P2. Here P1 indicates the first number of punctured characters and is defined as the maximum integer less than LM / D1. P2 indicates the second number of punctured characters, determined by the difference between the total number P of characters to be punctured, and the first number P1 of punctured characters. I.e

P2 = P-P1, D2 = sD1 and

The puncture intervals D1 and D2, as well as the number of punctured characters P1 and P2, are supplied from the puncture pattern determination unit (not shown). The puncturing pattern determination unit, the modulo operation blocks 320 and 330, and the OR gate 340 serve as a puncture pattern generator to generate a puncture permit signal defining a puncturing operation of the puncturer 350.

Referring to FIG. 11, in which, as in the flexible data rate matching device shown in FIG. 10, the flexible data rate matching device according to an embodiment of the present invention includes a channel encoder 10, a repeater 110, a puncturer 350, a channel interleaver 100, and a symbol index generator 310. The flexible data rate matching device shown in FIG. 11 is characterized in that instead of the modular operation blocks 320 and 330 and the OR gate 340 in FIG. 10, it comprises a puncture pattern generator 360. Thanks to this design, the matching device in the flexible data rate mode is implemented using software. The puncture pattern generator 360 stores a program of the address generator module, the piercing permission signal ‘1’ being generated when k satisfies a special condition according to this program. The puncturing pattern generator 360 determines the kth encoded characters corresponding to the case where k is a multiple of D1 or a multiple of D2 plus 1 (where D2 is an even number) to puncture specific encoded characters. The puncturing pattern generator 360 can also determine the kth coded symbols corresponding to the case where k is a multiple of D1, a multiple of D2 plus 1 (where D2 is an even number) or a multiple of D2 minus 1 (where D2 is an odd number) to puncture certain coded characters. Then, the matching device in the flexible data rate mode actually outputs N characters from the LM symbols, as the matching device in the flexible data rate mode in FIG. 10.

FIG. 12 shows a procedure for performing negotiation operations in a flexible data rate and transmission mode shown in Table 6 according to a second embodiment of the present invention.

), то тогда при операции 606 прокалывают соответствующие k-е кодированные символы. Если при операции 605 определяют, что k не является ни (кратное D1), ни ((кратное D2)-

), процедура переходит к операции 607 для увеличения значения k на +1. После операции 607 при операции 608 определяют, выполняется ли равенство k=LM+1. Если это равенство выполняется, то процесс заканчивается. Если нет, то операции с 605 по 608 повторяются, пока при операции 608 не будет определено, что k=LM+1. При указанном способе создается шаблон прокалывания FDRT, близкий к равномерному.Turning to FIG. 12, in step 601, the initial parameters N, L, M, and P necessary for FDRT are initialized. The number L of encoded characters constituting the frame and the interleaver capacity N are determined in accordance with a given data rate, while the number of M repetitions and the number P of characters to be punctured are determined by the formula in Table 6. In operation 602, the first puncture interval D1 is calculated. and the first number P1 of punctured characters in accordance with the formula specified in the algorithm. At operation 603, a second puncturing interval D2 and a second number P2 of punctured characters are calculated in accordance with the formula given in the algorithm. After all parameters are calculated in operations 602 and 603, operations 604 through 608 are performed with a successive increase in the value of k from 1 to LM. At each stage of the increment k of the account, if it is determined at operation 605 that k is (a multiple of D1) or ((a multiple of D2) -

), then, at operation 606, the corresponding kth encoded symbols are punctured. If at operation 605 it is determined that k is neither (a multiple of D1) nor ((a multiple of D2) -

), the procedure proceeds to operation 607 to increase the value of k by +1. After operation 607, operation 608 determines whether the equality k = LM + 1 holds. If this equality holds, then the process ends. If not, then operations 605 to 608 are repeated until, at operation 608, it is determined that k = LM + 1. With this method, an FDRT piercing pattern is created that is close to uniform.

), то тогда прокалывается соответствующий k-й кодированный символ. В альтернативном варианте, если k есть (кратное D1) или ((кратное D2)-

), то тогда прокалывают соответствующий k-й кодированный символ. Это делается для того, чтобы прокалывание выполнялось в позициях, не совпадающих с кодированными символами, соответствующими числу, кратному D1, а также для предотвращения перекрывания диапазона прокалывания и диапазона LM. Кроме того, это необходимо для того, чтобы позиция D1 прокалывания находилась как можно дальше от позиции D2 прокалывания по мере возрастания значения D1. То есть кодированные символы, соответствующие ((кратное D2)-

), прокалываются в других позициях, не совпадающих с кодированными символами, проколотыми в позициях, соответствующих числу, кратному D1.12, if it is determined that k is (a multiple of D1) or ((a multiple of D2) -

), then the corresponding kth coded character is punctured. Alternatively, if k is (a multiple of D1) or ((a multiple of D2) -

), then the corresponding kth coded symbol is punctured. This is to ensure that puncturing is performed in positions that do not match the coded characters corresponding to a multiple of D1, and also to prevent overlapping of the puncturing range and the LM range. In addition, this is necessary so that the puncture position D1 is as far as possible from the puncture position D2 as the value of D1 increases. That is, encoded characters corresponding to ((a multiple of D2) -

) are punctured in other positions that do not match the coded characters punctured in positions corresponding to a multiple of D1.

FIGS. 13 and 14 show structures of matching devices in a flexible data and transmission rate mode according to a second embodiment of the present invention. In particular, FIG. 13 shows the structure of hardware implementing the FDRT algorithm, and FIG. 14 shows the structure of software tools of the FDRT algorithm. That is, the FDRT device according to the second embodiment of the present invention can be implemented either using a programmable module, such a DSP and CPU, as shown in FIG. 11, or in the form of a hardware module, such as an ASIC, as shown in FIG. 13.

Referring to FIG. 13, a flexible data rate matching device according to an embodiment of the present invention includes a channel encoder 10, a repeater 110, a puncturer 550, a channel interleaver 100, a symbol index generator 510, modular operation units 520 and 530 and an OR gate (or logical summation unit) 540.

. Прокалыватель 550 выполняет операцию прокалывания на потоке из LM повторяющихся символов и выдает поток из N символов. В частности, прокалыватель 550 выполняет операцию прокалывания в соответствии с сигналом разрешения прокалывания PUNC_EN от логического элемента ИЛИ 540. То есть сигнал разрешения прокалывания PUNC_EN является шаблоном прокалывания для определения операций прокалывания прокалывателя 550. Канальный перемежитель 100 с емкостью N выполняет перемежение N-символьного потока, выходящего из прокалывателя 550.Channel encoder 10 creates a stream of L encoded symbols. The repeater 110 repeats the stream of L coded symbols M times and provides LM repeated coded symbols. Here, M indicates the number of repetitions of a stream of L encoded symbols and is defined as a minimum integer greater than N / L. I.e

. Punisher 550 performs a puncturing operation on a stream of LM repeating characters and produces a stream of N characters. In particular, the puncturing device 550 performs a puncturing operation in accordance with the puncturing enable signal PUNC_EN from the OR logic element 540. That is, the puncturing enable signal PUNC_EN is a puncturing pattern for determining puncturing operations of the piercer 550. The channel interleaver 100 with a capacity of N interleaves the N-symbol stream, coming out of the piercer 550.

Логический элемент ИЛИ 540 формирует сигнал разрешения прокалывания PUNC_EN путем выполнения логической операции ИЛИ над выходными сигналами блоков 520 и 530 операций по модулю и подает сформированный сигнал разрешения прокалывания PUNC_EN в прокалыватель 550.The symbol index generator 510 sequentially generates indexes indicating symbols that form a stream of LM repeating symbols. Character index generator 510 may be implemented using a counter. The operation module 520 modulo obtains the index k generated by the symbol index generator 510 and D1 and generates a puncture enable signal PUNC_EN equal to '1' when the k-th coded symbol corresponds to the coded symbol in the puncture position. For example, in operation module 520, the modulus of the condition “when the kth encoded symbol corresponds to the encoded symbol at the puncture position” refers to the case where the kth encoded symbol corresponds to a multiple of D1. The operation module 530 modulo receives the index k generated in the symbol index generator 510 and D2 and generates a puncture enable signal PUNC_EN equal to '1' when the kth encoded symbol corresponds to the encoded symbol in the puncture position. For example, in block 530 of the modulo operation, the condition "when the kth encoded symbol corresponds to the encoded symbol in the puncture position" refers to the case when the kth encoded symbol corresponds to the number ((a multiple of D2) -

The OR gate 540 generates a puncture enable signal PUNC_EN by performing a logical OR operation on the output signals of the operation units 520 and 530 modulo and supplies the generated puncture enable signal PUNC_EN to the piercer 550.

P2=P-P1, D2=sD1 и

Интервалы D1 и D2 прокалывания, а также количества P1 и P2 прокалываемых символов подаются из блока определения шаблона прокалывания (не показан). Блок определения шаблона прокалывания, блоки 520 и 530 операций по модулю и логический элемент ИЛИ 540 служат в качестве генератора шаблона прокалывания для формирования сигнала разрешения прокалывания, определяющего операцию прокалывания прокалывателя 550.As described with reference to Table 6, equations (1) and (2) and FIG. 9, D1 and D2 are puncture interval values defining an interval between characters punctured in a coded symbol stream within a single frame. The first puncture interval D1 is defined as a minimum integer greater than LM / P for the number P = LM-N of punctured characters. The second puncturing interval D2 is defined as sD1 for the selected integer 's' of integers less than or equal to the maximum integer less than P1 / P2. Here P1 indicates the first number of punctured characters and is defined as the maximum integer less than LM / D1. P2 indicates the second number of punctured characters, determined by the difference between the total number P of characters to be punctured, and the first number P1 of punctured characters. I.e

P2 = P-P1, D2 = sD1 and

The puncture intervals D1 and D2, as well as the number of punctured characters P1 and P2, are supplied from the puncture pattern determination unit (not shown). The puncture pattern determination unit, the modulo operation blocks 520 and 530, and the OR logic element 540 serve as a puncture pattern generator to generate a puncture permit signal defining a puncturing operation of the piercer 550.

) для прокалывания определенных кодированных символов. Генератор 560 шаблона прокалывания может также определить k-е кодированные символы, соответствующие случаю, когда k есть (кратное D1) или ((кратное D2)-

) для прокалывания определенных кодированных символов. Затем устройство согласования в режиме гибкой скорости передачи данных действительно выдает N символов из LM символов, как и устройство согласования в режиме гибкой скорости передачи данных на фиг.13.Referring to FIG. 14, in which, as in the flexible data rate matching device shown in FIG. 13, the flexible data rate matching device according to an embodiment of the present invention includes a channel encoder 10, a repeater 110, a puncturer 550, a channel interleaver 100, and a symbol index generator 510. The flexible data rate matching device shown in FIG. 14 is characterized in that instead of the modular operation blocks 520 and 530 and the OR gate 540 in FIG. 10, it comprises a puncture pattern generator 560. Thanks to this design, the matching device in the flexible data rate mode is implemented using software. A puncture pattern generator 560 stores a program of the address generator module, wherein a puncture enable signal '1' is generated when k satisfies a special condition according to this program. The puncturing pattern generator 560 determines the kth encoded characters corresponding to the case when k is (a multiple of D1) or ((a multiple of D2) -

) to puncture specific encoded characters. The puncturing pattern generator 560 can also determine the kth encoded symbols corresponding to the case when k is (a multiple of D1) or ((a multiple of D2) -

) to puncture specific encoded characters. Then, the matching device in the flexible data rate mode actually outputs N characters from the LM symbols, as the matching device in the flexible data rate mode in Fig. 13.

Performance analysis

In the following description, the change in performance is analyzed in accordance with the puncturing of encoded symbols encoded using a convolutional code, and the average change in the effectiveness of the convolutional code with the code rate R from the puncturing rate and the repetition rate is described. From this data, it is possible to predict the difference in performance provided by the well-known FDRT algorithm for the IS-2000 system and the new FDRT algorithm proposed in the invention, as well as to estimate the average performance value.

First, we define the notation used below:

R is the convolutional code rate, R = k / n;

Rst: (bit rate of encoded characters actually transmitted over the channel) · R, Rst = NR (bit / s); and

Rfdrt: (transmission rate of coded symbols at the output of the channel encoder in FDRT mode) · R, Rfdrt = LR (bit / s).

When using a uniform puncture or repetition pattern, the change in productivity caused by the puncture or repetition is given by equation (3) below. Here, when Rfdrt <Rst, the FDRT circuit performs symbol repetition, so that performance, i.e., coding efficiency, is improved. Conversely, when Rfdrt> Rst, the FDRT circuit punctures the characters, so that performance, i.e., coding efficiency, is degraded. However, as mentioned earlier, since N> L, the FDRT circuit typically performs symbol repetition, thereby increasing performance, i.e., coding efficiency. The objective of the study is to find ways to improve the coding efficiency in accordance with the template.

Equation (3)

Average coding efficiency =

= 10log ₁₀ (Rst / Rfdrt) dB

For example, at Rst = 19.2 kilobits / s, the coding efficiency values depending on Rfdrt are shown in table 7. Thus, if the puncture pattern or repetition pattern is determined correctly and the FDRT algorithm is used, then the coding efficiency values shown in table 7 are guaranteed.

Figures 15 and 16 compare the simulation results of the new FDRT algorithm with the simulation results of the known FDRT algorithm.

FIG. 15 graphically illustrates simulation results for a case of applying the present invention for the RC3 IS-2000 (code rate R = 1/4). These graphs were obtained under the following modeling conditions. Case (1), case (2) and case (3) correspond to the modeling conditions presented in tables 8, 9 and 10, respectively. In case (1), the data transfer rate is 15 kilobits / s, the number of encoded symbols per frame is L = 1200, and the interleaver capacity is N = 1536. Here, the curves 15k_BER_IS2000 and 15k_FER_IS200 show the simulation results according to the prior art, while the curves 15k_BER_SEC and 15k_FER_SEC show the simulation results according to the present invention. In case (2), the data transfer rate is 10 kilobits / s, the number of encoded symbols per frame is L = 800, and the interleaver capacity is N = 1536. In this case, the simulation results are shown only according to the prior art. In case (3), the data rate is 19.2 kilobits / s. In this case, puncturing / repetition of characters does not occur.

Referring to FIG. 15, the RC3 simulation results are shown, wherein the FDRT circuit (15k_BER_SEC and 15k_FER_SEC) according to the present invention provides an Eb / No gain of about 0.9 dB to 1.0 dB compared to the known FDRT scheme for IS-2000 (15k_BER_IS2000 and 15k_FER_IS2000). This almost coincides with the average coding efficiency of 1.07 dB compared to 19.2 kbit / s, as shown in table 7. These results were obtained by creating a uniform puncturing and repetition pattern, and the performance also demonstrates an almost optimal value. Thus, the FDRT condition (1) and the FDRT condition (2) of the FDRT algorithm proposed in the invention plays an important role for performance, and the new type 1 FDRT algorithm reflecting these conditions can also provide high performance. However, it should be noted that the simulation results for the well-known FDRT IS-2000 algorithm unexpectedly demonstrate coding efficiency of the order of 0.1 dB. The reason for this is an asymmetric pattern with an increase in the puncturing rate towards the end of the frame, as described above. As a result, under the same operating conditions of the channel, a difference in efficiency arises of approximately 0.9 to 1.0 dB, depending on the FDRT pattern.

Fig. 16 graphically illustrates simulation results for a case of applying the present invention to RC4 SCH (code rate R = 1/2). These graphs were obtained under the following modeling conditions. Case (1), case (2) and case (3) correspond to the modeling conditions presented in tables 11, 12 and 13, respectively. In case (1), the data transfer rate is 15 kilobits / s, the number of encoded symbols per frame is L = 600, and the capacitance N of the interleaver = 768. Here, the curves 15k_BER_IS2000 and 15k_FER_IS200 show the simulation results according to the prior art, while the curves 15k_BER_SEC and 15k_FER_SEC show the simulation results according to the present invention. In case (2), the data transfer rate is 17.5 kilobits / s, the number of encoded symbols per frame is L = 700, and the interleaver capacity is N = 768. In this case, the simulation results are shown only according to the prior art. In case (3), the data transfer rate is 10 kilobits / s, the number of encoded symbols per frame is L = 400, and the interleaver capacity is N = 768. In this case, the simulation results are shown only according to the prior art. In case (4), the data rate is 19.2 kilobits / s. In this case, puncturing / repetition of characters does not occur.

Referring to FIG. 16, in which the simulation results of RC4 also coincide with the simulation results shown in FIG. As shown in FIG. 16, the FDRT circuit (15k_BER_SEC and 15k_FER_SEC) according to the present invention provides an Eb / No gain of about 0.8 dB to 0.9 dB compared to the known FDRT IS-2000 (15k_BER_IS-2000 and 15k_FER_IS2000).

Characteristics at 10 kbps are important. In this case, the well-known FDRT algorithm almost reaches the average coding efficiency of 2.83 dB shown in Table 7. These results are obtained because the piercing distance (interval) D for the case of 10 kbit / s is set to an integer, as a result of which during the determination piercing distance (interval) D, uneven puncturing does not occur due to the LM / PD difference. Thus, this is a good example showing that the performance is directly related to the previous condition, namely that in the process of determining the piercing distance (interval) D, the difference LM / P-D should be fully taken into account. Table 14 below shows the simulation conditions for this case.

As described above, the new FDRT scheme according to the present invention matches a frame having coded symbols flexibly determined according to a change in the data rate with an interleaver capacity in the data transmission system. The FDRT circuit evenly distributes a puncture pattern or a repetition pattern in a frame by adjusting the initial settings, which allows you to flexibly transmit data at the appropriate speed without sacrificing performance.

Although the invention has been demonstrated and described with reference to a specific preferred embodiment, it will be apparent to those skilled in the art that various changes may be made in form and detail without departing from the spirit and scope of the invention as defined by the appended claims. .

Claims (28)

1. A method of generating a stream of N characters by puncturing a stream of repeating characters in a system comprising an encoder for generating a stream of L characters, a repeater for repeating a stream of L characters and a piercer for piercing a stream of repeating characters and forming a stream of N characters, where N is greater than L consisting in the fact that form a stream of LM repeating characters by repeating a stream of L characters M times, where M is the minimum integer greater than N / L, calculate the first puncture interval D1, defined as the minimum an integer greater than LM / P for the number P = LM-N of characters to be punctured, and a first number P1 of punctured characters, defined as a maximum integer less than LM / D1, a second number of P2 punctured characters is calculated indicating the difference between the number P characters to be punctured, and the first number P1 of punctured characters, and the second puncture interval D2, defined as sD1 for the selected one integer s from integers less than or equal to the maximum integer less than P1 / P2, and form otok from N symbols by puncturing flow of LM repeated symbols with a first puncturing interval D1 and second D2 puncturing interval. 2. The method according to claim 1, characterized in that the positions of the characters punctured with the first puncture interval D1 do not coincide with the positions of the characters punctured with the second puncture interval D2. 3. The method according to claim 1, characterized in that the characters punctured with the first puncture interval D1 are equivalent to characters placed at positions corresponding to a multiple of D1 from the starting characters in the stream of LM repeating characters. 4. The method according to claim 1, characterized in that the characters punctured with a second puncture interval D2 are equivalent to characters placed at positions corresponding to a multiple of D2 plus an offset from the starting characters in the stream of LM repeating characters. 5. The method according to claim 4, characterized in that the offset is 1. 6. The method according to claim 4, characterized in that the offset is -1. 7. The method according to claim 4, characterized in that the offset is equivalent to the value determined by subtracting D2 from the maximum integer less than D1 / 2. 8. The method according to claim 4, characterized in that the offset is equivalent to the value determined by multiplying -1 by the value obtained by summing D2 with a maximum integer less than D1 / 2. 9. A device for matching L encoded symbols determined in accordance with a change in the data rate with an interleaver capacity N, where N is greater than L, comprising an encoder for generating a stream of L encoded symbols, a repeater for repeating the stream of L encoded symbols M times and outputting a stream of LM repeating characters, where M is defined as a minimum integer greater than N / L, a puncturing pattern generator for (a) determining a first puncturing interval D1 defined as a minimum integer greater than LM / P for i the number of P = LM-N characters to be punctured, and the first number P1 of punctured characters, defined as the maximum integer less than LM / D1, (b) the calculation of the second number P2 of punctured characters indicating the difference between the number P of characters to be punctured, and the first number P1 of punctured characters, and the second puncture interval D2, defined as sD1 for the selected one integer s from integers less than or equal to the maximum integer less than P1 / P2, (c) formation of a puncture pattern a puncture used to puncture a stream of LM repeating characters with a first puncturing interval D1 and a second puncturing interval D2, and a puncturing device for puncturing a stream of LM repeating characters in accordance with a puncturing pattern with a first puncturing interval D1 and a second puncturing interval D2, and generating a flow from N characters. 10. The device according to claim 9, characterized in that it further comprises a symbol index generator for generating indices indicating the corresponding symbols forming a stream of LM symbols and feeding the generated indices to the puncturing pattern generator, wherein the puncturing pattern generator generates a puncturing pattern indicating symbols which correspond to the first and second puncture intervals D1 and D2 of the symbols in the stream of LM symbols. 11. The device according to 9, characterized in that it further comprises an interleaver for interleaving the output signal of the puncturer before transmission. 12. The device according to claim 9, characterized in that the positions of the characters punctured with the first puncture interval D1 do not coincide with the positions of the characters punctured with the second puncture interval D2. 13. The device according to claim 9, characterized in that the characters punctured with the first puncture interval D1 are equivalent to characters placed at positions corresponding to a multiple of D1 from the starting characters in the stream of LM repeating characters. 14. The device according to claim 9, characterized in that the characters punctured with a second puncture interval D2 are equivalent to characters placed at positions corresponding to a multiple of D2 plus an offset from the starting characters in the stream of LM repeating characters. 15. The device according to 14, characterized in that the offset is 1. 16. The device according to 14, characterized in that the offset is -1. 17. The device according to 14, characterized in that the offset is equivalent to the value determined by subtracting D2 from the maximum integer less than D1 / 2. 18. The device according to 14, characterized in that the offset is equivalent to the value determined by multiplying -1 by the value obtained by summing D2 with a maximum integer less than D1 / 2. 19. A method for matching L encoded symbols, determined in accordance with a change in the data rate, with an interleaver capacity N, where N is greater than L, comprising repeating a stream of L encoded symbols M times and outputting a stream of LM repeating symbols, wherein M is defined as a minimum integer greater than N / L, a stream of LM repeating symbols is punctured using the first number P1 of punctured characters in accordance with the first puncture pattern A, wherein P1 is defined as a maximum integer lower LM / D1, wherein the first puncturing pattern A indicates a multiple of the first puncturing interval D1 defined as a minimum integer greater than LM / P for the number P = LM-N of characters to be punctured, and the remaining characters are punctured after puncturing the stream from LM characters with a first puncturing interval D1 according to the second puncturing pattern B, and a stream of N characters is output when the second number P2 of punctured characters indicating the difference between the number P of characters to be punctured and the first by P1 punctured characters, greater than 0, and the second puncture pattern B is equivalent to the value determined by summing the offset with a multiple of the second puncture interval D2, which is defined as sD1 for the selected single integer s from integers less than or equal to the maximum integer, less P1 / P2. 20. The method according to claim 19, characterized in that the character positions defined by the first puncture pattern A do not match the character positions defined by the second puncture pattern B. 21. The method according to claim 19, characterized in that the offset is 1. 22. The method according to claim 19, characterized in that the offset is -1. 23. The method according to claim 19, characterized in that the offset is equivalent to the value determined by subtracting D2 from the maximum integer less than D1 / 2. 24. The method according to 17, characterized in that the offset is equivalent to the value determined by multiplying -1 by the value obtained by summing D2 with a maximum integer less than D1 / 2. Priority on points: 04/21/2000 according to claims 1, 3, 9, 10, 11, 13 and 19; 04.24.2000 according to claims 2, 4, 5, 6, 12, 14, 15, 16, 20, 21, 22; 04/26/2000 according to claims 7, 17, 23; 04/27/2000 according to claims 8, 18, 24.

Images