# Device and method for generating codes in a communication system

The invention relates to the transmission of data to generate codes using turbocodes in the communication system based on retransmission. The technical result is an increase in throughput. To do this, generate the original matrix of holes which receive the first subcode in the communication system containing turbocodes for generating information symbols, the first symbol and second parity symbols parity for input stream of information bits and a generator subcodes of information symbols, the first symbol parity and the second parity symbols by using a matrix of perforations, and the number of subcodes is equal to the number of matrices perforation, with select information symbols in a quantity equal to the number of columns of the original matrix perforation, their information symbols generated by turbocodes. 4 C. and 17 C.p. f-crystals, 6 ill., table 4.

The present invention relates generally to a device and method of generating codes in the data transmission system and, in particular, to a device and method of generating additional codes using turbocodes in the communication system based on retransmission.

In General, the system is westfleet soft Association to increase throughput. Methods soft associations are classified as the Association of packets passing and merging packages code. These two schemes joins are usually called soft combining packages. Although the pattern of Association of packets passing provides such a high efficiency, as the pattern of Association of packages on the codes, in the case of a small decrease in the efficiency, it wins due to the simplicity of implementation.

In the transmission of packets using the schema merge packets on the code to improve the throughput. This means that each packet transmission, the transmitter transmits the code with different encoding speed. If an error is detected in a received packet, the receiver requests retransmission and performs soft combining primary packet and the retransmitted packet. Code retransmitted packet may differ from the code of the previous package. Join scheme packages by codes is the process of combining the received N packets having the encoding rate of the R code with efficient encoding speed R/N to decode in order, therefore, to get the win due to the encoding.

As for the method of combining packets passing it, on the contrary con is if an error is detected in a received packet, the receiver requests retransmission and performs soft combining primary packet with a retransmitted packet. Code retransmitted packet is identical to the code of the previous package. In this sense we can say that the pattern of Association of packets passing averages symbols for channel with random noise. Join scheme packet spacing reduces the noise power by averaging the output signals of the soft decisions received symbols and provides gains explode like that get on the channel multipath propagation due to repeated transmission of the same code channel with fading. However, the pattern of Association of packets passing does not provide such additional gain due to the encoding, which is obtained in accordance with the structure of the code in the schema merge packets through the code.

Due to the simplicity of implementation in the majority of packet communication scheme is used to combine packets passing, which are considered in relation to a synchronous system IS-2000 and an asynchronous system UMTS. The fact that in the existing batch systems used convolutional codes and even merging packages code does not give a big win when using convolutional codes with low encoding speed. If the system with R=1/3 supports re-transferred to the hospitals in efficiency. Therefore, the pattern of Association packet spacing is chosen for reasons of simplicity of implementation. If you need codes, forward error correction (PIO) to use turbocode, you need another mechanism to merge packets, because turbocode used as codes for error correction, must have performance characteristics that are very close to “Limit the channel capacity Shannon”, due to the use of iterative decoding, and their effectiveness unlike convolutional codes depends on the coding rate. It was concluded that to achieve optimum efficiency in the system packet communication based on retransmission using turbocodes, it is useful to use a join packet code.

Thus, the present invention is a device and method of generating subcodes for optimal code associations in the system resubmit using turbocodes.

Another objective of the present invention is a device and method of generating additional codes using turbocodes in the communication system.

To solve the above and other problems, the authors propose a device and a method of generating subcodes from turbocodes in the parity and the second parity symbols for the input stream of data bits, and the generator subcodes generates subcodes of information symbols, the first symbol parity and the second parity symbols using matrices perforation. The number of subcodes is equal to the number of matrices perforation. To generate the original matrix perforation for the first subcode generator subcodes selects the information symbols in the quantity equal to the number of columns of the original matrix perforation of information symbols received from turbolader if the difference between the number Ns of the selected symbols in the matrix of the perforations and the number of columns of the original matrix perforation is greater than or equal to the number of component encoders of turbochager. Then the generator subcodes selects the first and second parity symbols in the amount equal to the difference, so that the number of selected first symbols of parity was greater than or equal to the number of selected second parity symbols.

The above and other objectives, features and advantages of the present invention made it clear from the description given in conjunction with the drawings, where

Fig.1 is a graph showing the difference in the efficiency of the packet data using turbocodes, in the case of Okolow according to a variant implementation of the present invention; and

Fig.3 is a graph showing the efficiency of the retransmission scheme without the use of subcodes, the retransmission scheme, which uses the Association for diversity using subcodes, and retransmission scheme, which is the code Association with the use of subcodes;

Fig.4 is a logical block diagram illustrating the method of generating the first subcode in a set of subcodes, representing quasitopological turbocode, according to a variant implementation of the present invention; and

Fig.5 is a logical flowchart illustrating a method of generating intermediate subcode in a set of subcodes, representing quasitopological turbocode, according to a variant implementation of the present invention; and

Fig.6 is a logical block diagram illustrating the method of generating the last subcode in a set of subcodes, representing quasitopological turbocode, according to a variant implementation of the present invention.

The following is a description of a variant of implementation of the present invention with reference to the accompanying drawings. This description will be considered in detail the well-known functions or constructions, in order not to obscure the invention is immaterial which of acetow in the system resubmit using turbocodes and system allows you to selectively apply the pattern of Association of packages code or scheme combining packets passing, depending on the coding rate of the data. It will be shown what are the advantages and efficiency gains provides the proposed system.

We will first describe the principle of operation of a system that selectively uses a scheme of combining the packet code or scheme combining packets passing, depending on the coding rate of the data.

In the system, which uses turbocode, for example, with R=1/5, merging packages code applies, until the encoding rate of the codes obtained by soft combining the retransmitted packet is less than 1/5. For subsequent retransmitted, aggregates the packets passing, and then merging packages code. If sending the first packet is used encoding speed data 1/3, then the resend request is a request to the required number of redundant symbols that will bring the total encoding speed to 1/5. Thus, when the receiver receives both packages, the overall encoding speed is equal to 1/5. The next package to duplicate transmission, preemiecare with encoding speed 1/5.

In Fig.1 depicts a graph showing the difference in efficiency provided by the Association of packages code and the Association of packets passing, in the case of turbocodes. According Fig.1, turbo code, low speed data coding 1/6 provides greater efficiency gains than the turbo code with a high bit rate 1/3, the same energy of character Es and provides a gain in efficiency, equal to 3 dB, by combining packages code. Therefore, generation of turbocodes with R=1/3 by combining packages code using subcodes with R=1/6 benefits, which provide turbocode with the encoding speed is below 1/3, as well as, gain, provide the code combining different codes.

In particular, for the same energy Es of the code symbol and the same given speed coding turbocode unlike convolutional codes provide efficiency, close to the “Limit of channel capacity Shannon” in accordance with the encoding speed, unless fully implemented iterative decoding. It is known that the turbo code with low encoding speed gives a higher efficiency gains than the turbo code with a high bit rate at one and then analiziroval change “maximum channel capacity Shannon”. The reason for is shown in Fig.1 the curves corresponding to R=1/3 and 1/6, provide the same energy of character, is that in a mixed system ASP, each repeated transmission of the same energy of character Es.

In a single repetition code with R=1/3 and the merger of the two codes by passing packets on the channel with abgs (additive white Gaussian noise) get the maximum gain of 3 dB with respect to the ratio of the energy of the symbol-to-noise ratio (Es/No). The same thing happens in the case of a code with R=1/6. Thus, the efficiency curve for turbo code R=1/3 is shifted to the left parallel to the axis at +3 dB at the expense of winning caused by combination of packages for diversity, and efficiency curve for a turbo code with R=1/6 also shifts to the left parallel to the axis at +3 dB at the same given energy symbol. In this case, efficiency curves characterize the ratio of energy to noise (Eb/No), which is measured to compare the efficiencies of codes depending on the coding rate. Hence, the difference between the curves of efficiency for turbocodes equivalent to the difference in efficiency provided by the Association of packets passing and merging packages code. The change in efficiency in zavisimosti can be obtained on the basis of the minimum required signal-to-noise ratio (SNR).

In the system, which uses turbocode with bit rate R and very large L block encoder, the minimum Eb/No required to ensure there are no errors on the channel, is expressed as

Eb/No>(4^{R}-1)/2R (1)

Table 1 shows the values of the minimum required Eb/No channel abgs each encoding speed for turbocodes according to the above value. Table 1 typical values of Eb/No correspond to the values of Eb/No required for a bit error rate (CPAB) was less than 0.00001, when the size L of the block coding for turbocodes equal to 1024.

From Table 1 it follows that for encoding speed 3/4, 2/3, 1/2, 3/8, 1/3, 1/4, 1/5 and 1/6 of the required Eb/No is 0.86, 0.57, 0.0, -0.414, -0.55, -0.82, -0.975 and -1.084 dB, respectively. For a system that uses a code with R=1/3, and systems that use code with R=1/6, the difference in efficiency is not less than 0.53 dB. This is the minimum difference in effectiveness based on “maximum channel capacity Shannon”. Taking into account the actual decoder and the system environment, the difference is even greater. The simulation shows that for a system that uses a join packet code for codes with R=2/3, and systems, ispol Table 2 shows the differences in efficiency, provided by the Association of packages code and the Association of packets passing after one retransmission in the system with velocity encoding subcode 2/3. According to Table 2, the minimum difference in efficiency is 1.12 dB, and in the system using a turbo code, join scheme package code provides greater efficiency gains.

According to the above-described merge packets code demonstrates high efficiency in the system re-transmission using turbo code. Therefore, for optimum combining of packages code in the system re-transmission using turbo code present invention provides a method of generating subcodes. Generation of subcodes for combining packages of code according to a predetermined rule provides the above benefits at the expense code combining and maximum efficiency of the system requesting the subcodes of the same size for each retransmission.

In Fig.2 shows a block diagram of a device generating subcodes using turbocodes responsible variant implementation of the present invention. According Fig.2 the device generating subcodes with the er (or the first constituent encoder) 201 encodes input information bits and outputs the first code symbols, i.e., the information symbols X and the first parity symbols Y0 and Y1. Interleaver 202 punctuates input information bits according to a predetermined rule. The second component encoder (or the second constituent encoder) 203 encodes perenesennyj stream of information bits and outputs the second code symbols, i.e., the information symbols X’ and the second parity symbols Y’0 and Y’1. Thus, turbocodes displays the first and second code symbols. Because, actually, the information symbols X’ generated by the second component encoder 203, are not transferable, the encoding speed of turbolader equal to 1/5.

The generator 204 subcodes generates the subcodes of the first and second code symbols received from the first and second component encoders 201 and 203, through the perforation and duplication under the control of the controller 205. The controller 205 stores the matrix perforation (and duplication), generated according to the algorithms shown in Fig.4, 5 and 6, and outputs the selection signals symbol for matrices perforation on the generator 204 subcodes. Then the generator 204 subcodes selects a specified number of code symbols in a given range of perforation in accordance with the selection signals of the symbol.

Used here is conditional on>0: redundant symbol issued by the upper component coder turbolader

Y1: redundant symbol issued by the upper component coder turbolader

Y’0: redundant symbol issued by the bottom component coder turbolader

Y’1: redundant symbol issued by the bottom component coder turbolader

In Fig.4, 5 and 6 is depicted a logical flowchart illustrating the procedure of generating subcode (or matrix perforation), corresponding to a variant of implementation of the present invention. In particular, in Fig.4 shows the procedure of generating the first subcode_{0}in a set of subcodes of Fig.5 shows the procedure of generating the intermediate subcodes C_{1}With_{s-2}in a set of subcodes, and Fig.6 illustrates the procedure of generation of the last subcode C_{s-1}in a set of subcodes.

Further, the signal comprising information symbols X and the first parity symbols Y0 and Y1, issued by the first component encoder 201, denote ENC1 (first code symbols), and the signal that includes the second parity symbols Y’0 and Y’1, issued by the second component encoder 203, denote ENC2 (the second code symbols).

According Fig.4, at step 401 specify the maximum encoding speed (Rmax), available for perennou encoding speed (Rmin) is set as an integer multiple of from Rmax (=k/n). Here, k is the number of input symbols, a n - number of output symbols. Although Rmin can be set arbitrarily, it is usually equal to 1/6, 1/7 or below, because the gain due to coding reaches saturation by reducing the coding rate of turbocodes to R=1/7 or below. In addition, determine the actual speed of encoding, i.e., maternal encoding speed (R) of the decoder of the receiver. R asked more than Rmin.

In a real system Rmax and Rmin ask in advance. In a sense, Rmax is the speed of encoding when generating subcodes, a Rmin is the final encoding speed after code combining subcodes. In General, Rmin is the encoding rate of the encoder of the transmitter.

At step 403 calculates the number of subcodes (S) by using the following expression, which includes Rmax and Rmin. In this case, the number of subcodes or the number of matrices perforation is the minimum integer greater than the ratio of Rmax to Rmin.

whererepresents the minimum integer greater than or equal to *.

At step 405 define the initial value 1 of the variable m, a, at step 407 To determine whether (=mk). The “s” is the number of columns of each matrix perforation, to the y value for the first transmitted subcode. In this case, for Rmax=3/4, set equal to 3.

At step 407 calculates the number Ns of symbols selected from a matrix of perforations, multiplying the variable m to a code length, i.e. the number of code symbols n, obtained from the expression Rmax=k/n. Ns is the number of selected characters or the number of selected positions in each matrix of the perforations is equal to C/Rmax.

At step 409, the value of (Ns-C) compared with the number of component encoders of turbolader transmitter. In modern turbocodes usually there are two component encoder. Thus, we assume that you are using two component encoder. At step 409 check that (Ns-C) is greater than or equal to 2, because turbocodes contains two component encoder, connected in parallel with an interleaver in between, as shown in Fig.2, in contrast to conventional encoders that use a different unit codes. In other words, to preserve the characteristics turbocodes, after the transfer of all information symbols should be submitted at least one parity symbol received from each component encoder.

If (Ns-C) is less than 2, then choose only one character from the first set of parity symbols and the second set of parity symbols. Twins is ingelow parity, are not turbocode and traditional codes with restrictive length K=4, generated by the encoder, containing only the first component encoder, and don't benefit from the interleaver provided in turbocore. On the other hand, in the second case, only the systematic character without parity symbols issued by the first component encoder gives subcodes with encoding speed 1. This is equivalent to the system without coding, which does not give any advantage at the expense of coding. Accordingly, to ensure the effectiveness of turbocodes (Ns-C) must be greater than or equal to 2.

If at step 409, it turns out that (Ns-C) is greater than or equal to 2, then at step 411 from the matrix of the perforations is selected With systematic information symbols, and other characters according to the specified type. For type 1, at step 413, the other symbols are selected from the first and second parity symbols in accordance with the expression (3). The number of selected first symbols of parity greater than or equal to the number of selected second parity symbols. For example, if the number of other characters, (Ns-C) is equal to 3, the first and second parity symbols is chosen according to the expression (3), and then choose another character from the first parity symbols.

For type 2, at step 415, the other symbols are selected from the first and second parity symbols in accordance with the expression (4). If the distribution coefficients of the characters for the first character parity and the second parity symbols a and b, respectively, of the first symbols of parity select the number of characters equal to the minimum integer greater than or equal to the ratio a(Ns-C) to (a+b), and second parity symbols select the number of characters equal to the maximum integer less than or equal to the ratio of b(Ns-C) to (a+b).

where a+b=1 and a and b are the coefficients of symbols for ENC1 and ENC2, respectively.

If the condition is checked at the step 409 is not performed, i.e., (Ns-C) is less than 2, then at step 417 the variable m is increased by 1, after which the procedure returns to step 407. Step 409 is performed in order to determine whether you are in the range of perforation (matrix size of perforation) to generate the subcodes are able to retain the character of turbocodes. If you are unable to preserve the character of turbocodes, then at step 417 the range of perforation is increased.

As described above, the original matrix perforation build to turbocodes you can choose all the information symbols and the Yu method of generating an intermediate matrix perforation with reference to Fig.5. Matrix perforation with C_{1}With_{s-2}generate, by repeating the procedure shown in Fig.5.

According Fig.5, steps 501 and 502 are carried out in compliance with the specified type. For type 1, at step 501, choose Ns symbols from the first and second sets of parity symbols, according to the expression (5). Ns is the product of m and n, obtained from Rmax (=k/n). The number of selected first symbols of parity greater than or equal to the number of selected second parity symbols. In this case, choose characters that are not selected from the previous matrices perforation.

For type 2, at step 503, choose Ns symbols from the first and second sets of parity symbols in accordance with predetermined ratios according to the expression (6). If a and b are defined distribution coefficients characters for the first character parity and the second parity symbols, respectively, of the first symbols of parity select the number of characters equal to the minimum integer greater than or equal to the ratio a(Ns) K (a+b), and second parity symbols select the number of characters equal to the maximum integer less than or equal to the ratio of b(Ns) (a+b). In this case, choose characters that are not selected from the previous matrices perforation.

According Fig.6, at step 601 selects all the remaining characters not selected from the previous matrices perforation. The number of selected symbols denote Ns2. At step 603 new set Ns (Ns-Ns2). Because when performing operations shown in Fig.4, 5 and 6, select the symbols in all positions of the matrix perforation, the new Ns is the number of characters to be re-selection. At step 605 check that Ns is greater than 0. If Ns is equal to 0, then the procedure ends. If this number is greater than 0, then from the information symbols re-select the number of characters in the new Ns. In other words, the re-transmission of the selected characters.

The following is a specific numerical calculations illustrating the above-described method of generating subcodes that meets the present invention.

For Rmax=3/4 and R=1/5, Rmin=1/6 and S=6/(4/3)=4/55. Thus, there are five matrices perforation.

{C_{0}C_{1}C_{2}C_{3}C_{4}}:Rmax=3/4.

Since the encoding speed of subcodes is 3/4 and the number of subcodes is 5, then after the merge codes, subcodes are encoding speed 3/20 ((1/S)(Rmax=(1/5)((3/4)=3/20). This means that if there are three information bits, the receiver accepts the 20 coded symbols. However, since the if">k=53=15, retransmit 5 characters of 15. Duplicate preferably information symbols. In the above example, if the information symbol X is repeated once in each subcode, the decoder takes turbocode with R=1/5, in which information symbols occur twice for each of S subcodes, when taken all S subcodes.

Subcodes obtained by the procedures shown in Fig.4, 5 and 6, are a kind of additional codes, but not in the strict sense of the term, due to the presence of duplicate characters and differences of characteristics of different subcodes. As subcodes get from turbocodes, they can be called quasitopological turbocode (CDTC). In the system of mixed ASP using CDTC apply the following scheme resubmit.

The system of mixed ASP is a basic example of using merge packets through the code. Combining packages of code is implemented in modern systems mixed ASP, namely, HARQ Type I, Type II and Type III. In these systems, methods, re-transmission can be realized using CDTC. If the transportation element (TE) is a block of information bits, which is the main the element is provided retransmission and TE primary transmission may have the same or different sizes. For the initial transmission and each retransmission using the next many CDTC.

From C_{q}CDTC with the size S of the set of codes, it is possible to reconstruct the parent code or to generate a new code Cq with a lower bit rate compared to the parent encoding rate of the Rm, by combining (or code associations) subcodes C_{i}(i=0, 1, 2,..., S-1). In this case, the mother code has a minimum bit rate that is available to the encoder. Then CDTC set as source code With encoding speed R=Rm or code With encoding speed

where S is the number of subcodes with encoding speed Ri; Rm - maternal encoding speed.

Describe the principle of operation of the system, in which at the initial transmission and each subsequent transfer using CDTC transmit TE same size. It is obvious that the present invention also supports transmission scheme using different TE. In this case, the number S of subcodes is 4 and maternal encoding speed R is equal to 1/5.

(Phase 1) Transmission is carried out on the basis of the TE, and at initial transmission and each re-transmission of broadcast subcode C_{i}CDTC.

(Step 2) When the total speed is billing purposes, exceeds 1/5, every subcode C_{i}CDTC passed in order To_{0}C_{1}With_{2},..., C_{s-1}each time a request for retransmission. This is called consolidation packages code.

(Stage 3) When the overall encoding speed for codes obtained soft Association initially transmitted and retransmitted packets is less than or equal to 1/5, every subcode C_{i}CDTC retransmit in order To_{0}C_{1}C_{2},..., C_{s-1}each time a request for retransmission. This is called the Union of the packets passing.

(Step 4) the Size of the set CDTC may be an arbitrary value, depending on the mother coding rate. For R=1/5 and the coding rate 2/3 subcodes for retransmission can be used up to four subcodes.

Table 3 lists the sets CDTC for speed packet data forward link traffic, presumed to be available to the modern system, IS-2000 1XEVDV. In this case, maternal encoding speed R=1/5, and the encoding speed of subcodes R=2/3, 1/3, or 1/6.

From Table 3 it is evident that when the coding rate of subcode 1/6, which is less than the mother coding rate 1/5, each repeated perkiest encoding 1/5, different programs use different codes_{0}and C_{1}. In this case, the size S of the set of codes is equal to 2. When the coding rate of subcode 2/3 excess maternal encoding speed 1/5, different programs use different codes_{0}With_{1}With_{2}With_{3}. In this case, the size S of the set of codes is equal to 4. After the transfer of all S subcodes, the receiver can recover maternal encoding speed R and get the maximum gain due to the coding provided by the encoder.

Table 4 shows examples of matrices perforation for each coding rate of subcodes.

According to Table 4, when the mother code using turbo code with rate 1/5 and generate subcode with rate 2/3 code symbols which take 4 information bits, the 4 information bits to generate code 20 characters. Subcode with rate 2/3 generate through perforations 14 characters 20 characters. In the process of merging these subcodes by combining packet spacing at each request retransmission transmit the same With_{0}obtained from the above matrix perforation. On the other hand, in the process of merging pactour subcodes_{0}C_{1}C_{2}With_{3}, combine packages to explode. For HARQ Type III using merge packets code all code symbols of the parent code decode after making four passes.

Pay attention to the fact that “1” in the matrix of perforation are shown in Table 4, indicate that the characters in these positions are subject to selection or transmission, a “0” indicates that the characters in these positions are subject to perforation. “2” indicates that the character in this position is duplicated. Matrix perforation (and duplication) must satisfy the following conditions.

(Condition 1) Information symbol X is repeated in subcode CDTC, in the case of duplication.

(Condition 2) If the information symbol X is repeated in subcode CDTC using duplication, repetition period set equal to the minimum constant value in CDTC uniting all subcodes.

(Condition 3) If you use a perforation, the subcodes CDTC redundant symbols except the information symbol X, if possible, undergo perforation.

(Condition 4) If you use a perforation, in the subcodes CDTC, unnecessary characters, except for the information symbol X, if possible, podoben. Decoding is performed after soft combining of the symbols X, which occur twice, and because the actual encoding rate of the decoder is equal to 1/5. A code with a rate 1/5, at higher energy information symbol X, provides increased efficiency compared to the source with a speed of 1/5, in which the energy of the symbols are distributed uniformly. In other words, it is most preferable to duplicate the information symbol. We can say that matrix perforation and duplication are shown in Table 4, is constructed in such a way that the energy information symbol is increased at a uniform repetition of information symbols.

In Table 4, for R=1/6, the sequence of transmission code symbols is as follows:

With_{0}: X, X, Y0, Y1, Y’0, Y’1, X, X, Y0, Y1, Y’0, Y’1,...

Since the input of one information symbol is generated six code symbols, the encoding speed of subcode is 1/6.

For R=1/3 transfer sequence of code symbols is as follows:

With_{0}: X, Y0, Y’0, X, Y0, Y’0, X, Y0, Y’0, X, Y0, Y’0,...

C_{1}: X, Y1, Y’1, X, Y1, Y’1, X, Y1, Y’1, X, Y1, Y’1,...

Since the input of one information symbol is generated three code symbols, the encoding speed of subcode with After soft combining With_{0}and C_{1}X appears twice, and each of Y0, Y1, Y’0 and Y’1 appears once. Thus, in this case, you can use the decoder with the encoding speed 1/5, and matrix perforation satisfy the above-mentioned conditions, providing maximum efficiency.

In the first case, R=2/3, specified in Table 4, the sequence of transmission code symbols is as follows:

With_{0}: Y0, X, Y’0, Y0, X, Y’0, Y0, X, Y’0, Y0, X, Y0, ...

C_{1}: X, Y0, Y0, X, Y’0, Y0, X, Y’0, Y0, X, Y’0, Y’0, ...

C_{2}: Y1, X, Y’1, Y1, X, Y’1, Y1, X, Y’1, Y1, X, Y’1, ...

With_{3}; X, X’1, Y1, X, Y’1, Y1, X, Y’1, Y1, X, X’1, Y1, ...

Because when you enter two information symbols generated three code symbol encoding speed of subcode is 2/3. Each transmission is assigned a different code, due to the use of other matrix perforation. After soft combining With_{0}C_{1}C_{2}With_{3}X appears twice, and each of Y0, Y1, Y’0 and Y’1 appears once. Thus, in this case, as in the case of R=1/6, you can use the decoder with the encoding speed 1/5 and matrix perforation satisfy the above-mentioned conditions, providing maximum efficiency.

In the second case, R=2/3, specified in Table 4, the sequence of transmission code symbols looks sleduu, ...

C_{2}: Y1, Y1, X’1, Y’1, Y1, Y’1, Y1, Y1, Y’1, Y’1, Y1, Y’1, ...

With_{3}: X, Y’1, X, X, Y’1, X, X, Y’1, X, X, Y’1, X, ...

Because when you enter the four information symbols generated six code symbols encoding speed of subcode is 2/3. Each transmission is assigned a different code, due to the use of other matrix perforation. After soft combining With_{0}C_{1}C_{2}With_{3}X appears twice, and each of Y0, Y1, Y’0 and Y’1 appears once. Thus, in this case, as in the case of R=1/6, you can use the decoder with the encoding speed 1/5, and matrix perforation satisfy the above-mentioned conditions, providing maximum efficiency.

In Fig.3 depicts a chart in order to compare the performance of the HARQ system, which uses the join packet code, with the efficiency of the HARQ system, which uses the join packet diversity in terms of its bandwidth for CDTC with R=2/3 and 3=4, according to the present invention. According Fig.3, the HARQ system 301, in which the applied consolidation package code CDTC, and the HARQ system 302, which is used for combining packets passing for CDTC show higher efficiency than the system HARQ 303, which COI is 0.25), the value of Es/No in the HARQ system 301 should be about -4 dB, in the HARQ system 302 is about -1.3 dB, and in the HARQ system 303 is about 1 dB. Consequently, the use CDTC according to the present invention ensures higher throughput with less energy symbol.

According to the above subcodes generated for optimum combining of packages code according to a predetermined rule, increase system throughput retransmission using turbocodes that meets the present invention.

Although the invention has been shown and described with reference to preferred variants of its implementation specialists can offer various changes regarding the form and detail, without going beyond the nature and scope of the invention defined in the claims.

Claims

1. The method of generating the initial matrix of holes which receive the first subcode, in the communication system containing turbocodes for generating information symbols, the first symbol parity and the second parity symbols for encoding in response to the input stream of data bits with a given encoding speed and the generator subcodes for g matrices perforation, moreover, the number of subcodes is equal to the number of matrices perforation containing phases in which select information symbols in a quantity equal to the number of columns of the original matrix perforation of information symbols generated by turbocodes if the difference between the number Ns of the selected symbols in the matrix of the perforations and the number of columns of the original matrix perforation is greater than or equal to the number of component encoders in turbocodes, and select the first and second parity symbols in the amount equal to the difference, and the number of selected first symbols of parity greater than or equal to the number of selected second parity symbols.

2. The method according to p. 1, which further comprises a step, which increases the number of columns in the matrix perforation at integer multiples if the difference is less than the number of component encoders.

3. The method according to p. 1, which further comprises a step where, when generating the second matrix perforation, choose Ns symbols of the first and second parity symbols, not selected in the original matrix of the perforations, the number of the selected first symbols of parity greater than or equal to the number of selected second parity symbols.

4. The method according to p. 3, which Supplement the matrices perforation, except for the last matrix perforation, and re-choose (Ns-Ns2) information symbols, where Ns2 is the number of other characters parity, not the other matrices perforation.

5. The method according to p. 1, in which the communication system uses subcode for mixed ASP (automatic repeat request).

6. The method according to p. 1, in which the number Ns of the selected symbols is determined on the basis of C/Rmax, where C is the number of columns, and Rmax is the encoding speed of subcode.

7. The method of generating the first subcode subject to transfer by selecting a specified number of characters from the information symbols and first and second parity symbols in a given range of perforation in the communication system containing turbocodes for generating information symbols, the first symbol parity and the second parity symbols for encoding in response to the input stream of data bits with a given encoding speed, and the specified number of characters is Ns symbols containing phases in which purchased all of the information symbols in the range of perforation, if the difference between the Ns and the number of information symbols is greater than or equal to the number of component encoders of turbolader, and choose the first and second parity symbols in ku selected second parity symbols.

8. The method according to p. 7, which further comprises a step, which increases the range of perforation at integer multiples if the difference is less than the number of component encoders.

9. The method according to p. 7, which further comprises a step where, when generating the second subcode, choose Ns symbols of the first and second parity symbols, not selected in the first subcode, the number of the selected first symbols of parity greater than or equal to the number of selected second parity symbols.

10. The method according to p. 9, which further comprises a step where, when generating the last subcode, choose the remaining first and second parity symbols, not the other subcodes, except the last subcode, and re-choose (Ns-Ns2) information symbols, where Ns2 is the number of other unselected symbols parity.

11. The method according to p. 7, whereby the communication system uses the subcodes for mixed ASP (automatic repeat request).

12. The method according to p. 7, in which the number Ns of the selected symbols is determined on the basis of C/Rmax, where C is the number of columns of the matrix perforation, and Rmax is the encoding speed of subcode.

13. Method of generating matrices perforation through which perforined information symbols and parfievich characters parity and the second parity symbols for encoding in response to the input stream of data bits with a given encoding speed and the generator subcodes for generating subcodes of information symbols, the first symbols of parity and the second parity symbols through the perforation containing phases, which determine the number of S matrices perforation, calculating the minimum integer greater than or equal to the ratio of Rmax to Rmin, where Rmax is given the maximum encoding speed, and Rmin is this minimum speed encoding turbolader, determine the number of Ns symbols, subject to the choice of each matrix perforation, on the basis of C/Rmax, where C is the number of columns of the matrix perforation, and generate a first matrix of holes to choose what information symbols, choose the first character parity in number, equal to the minimum integer greater than or equal to the ratio a(Ns-C) to (a+b), the second parity symbols in the quantity equal to the maximum integer less than or equal to the ratio of b(Ns-C) to (a+b), if the difference between the Ns (Ns-C), greater than or equal to the number of component encoders in turbocore, where a and b are the coefficients of symbols for the first and second parity symbols.

14. The method according to p. 13, which further comprises a step, which increases the number of columns in the matrix perforation at integer multiples if the difference is less than the number of component encoders.

15. The method according to p. 13, kouroi and last matrices perforation, when choosing the first parity symbols in the amount equal to the minimum integer greater than or equal to the ratio of aNs to (a+b), and the second parity symbols in the quantity equal to the maximum integer less than or equal to the ratio of bNs to (a+b) without selecting any of the information symbols, so that the first and second parity symbols in the first and other matrices perforation differed from each other.

16. The method according to p. 15, which further comprises a stage on which to generate the last matrix perforation, choosing the rest of the first and second parity symbols, not the other matrices perforation, except the last matrix perforation, and re-choose (Ns-Ns2) information symbols, where Ns2 is the number of other characters parity, not the other matrices perforation.

17. The method according to p. 13, whereby the communication system uses the subcodes for mixed ASP (automatic repeat request).

18. The device generating subcode in the communication system containing turbocodes for encoding an input stream of data bits with this encoding speed and generating information symbols, the first symbol and second parity symbols parity and generator subcodes for generating Pervy and selection of first and second parity symbols in number, equal to the difference between a specified number of Ns symbols, subject choice, and the number of information symbols in a given range of perforation, and the number of selected first symbols of parity greater than or equal to the number of second parity symbols, if the difference is greater than or equal to the number of component encoders of turbochager.

19. The device according to p. 18, in which the generator subcodes generates the second subcode, choosing Ns symbols of the first and second parity symbols, not choosing information symbols, and the number of selected first symbols of parity greater than or equal to the number of selected second parity symbols.

20. The device according to p. 19, in which the generator subcodes generates the last subcode, choosing the rest of the first and second parity symbols, not the other subcodes, except the last subcode, and re-choose (Ns-Ns2) information symbols, where Ns2 is the number of other characters parity, not the other subcodes.

21. The device according to p. 18, in which the generator subcodes generates subcodes, increasing the range of perforation at integer multiples if the difference is less than the number of component encoders.

**Same patents:**

FIELD: Witterby algorithm applications.

SUBSTANCE: system has first memory element for storing metrics of basic states, multiplexer, capable of selection between first and second operating routes on basis of even and odd time step, adding/comparing/selecting mechanism, which calculates metrics of end states for each state metric. Second memory element, connected to adding/comparing/selecting mechanism and multiplexer is used for temporary storage of end states metrics. Multiplexer selects first operating route during even time steps and provides basic states metrics, extracted from first memory element, to said mechanism to form end state metrics. During odd cycles multiplexer picks second operating route for access to second memory element and use of previously calculated end state metrics as metrics of intermediate source states.

EFFECT: higher efficiency.

2 cl, 9 dwg