The apparatus and method of generating and decoding codes in a communication system

 

The proposed device, generating quasicomplete turbocode in the communication system. The device includes turbocodes, interleaver to interleave the symbols provided by turbolader, according to a given speed and a code generator for generating quasicomplete of turbocodes through thinning and repetition perenesennyj symbols from the interleaver. Further, the device decodes quasicomplete turbocode in the communication system. The decoding device includes a decoder code to generate code according to the code rate, transmitted through deponirovanie of subcodes quasicomplete of turbocodes transmitted from the transmitter, and soft combining of subcodes, departmental to deteremine characters issued from the decoder code, and turbodecoding for decoding the output of deteremines. The technical result is to create a device and method of decoding complementary turbocodes, given the characteristics of turbocodes in the communication system. 4 N. and 22 C.p. f-crystals, 7 Il.

The present invention relates in General to a device generating codes in the data transmission system, in particular to a device and method genetic repeat request (ASP) (ARQ) or in a conventional communication system, support scheme ASP.

In the General case, the system supporting hybrid scheme ASP (hereinafter referred to as "system GASP" (HARQ), the method uses soft combining to improve their throughput and methods soft combining are divided into batches and combine with the division and method of the packet code combining. These two methods combined are commonly referred to as "soft packet combining". Compared with the method of the packet code combining batch method combined with diversity is suboptimal in terms of performance, but is often used because of the simplicity of his incarnation, when the loss of quality is not a major factor in the design of the system.

Method of packet code combining is used in the packet communication, in order to increase its throughput. Method of packet code combining transmits changing codes with code rate R for each transmitted packet. Upon detection by the receiver of the error in a received packet after decoding, the receiver keeps faulting package instead of discarding it, and then gently combines the saved package with the package, re-transmitted transmit the billing purposes, having a code rate R, the method of combining batch code converts the code speed in the effective code rate R/N using packet before decoding, thereby improving coding efficiency.

On the other hand, the method of packet combining with the division sends the same codes with code rate R for each transmitted packet. When the receiver of the error in a received packet after decoding, the receiver keeps faulting package instead of discarding it, and then gently combines the stored packet with the packet retransmitted from the transmitter. In all cases, the same codes are used for the re-transmitted packet. Therefore, the method of batch combined with diversity can be considered a process of averaging the energy of the symbols in a random channel and uses only the effect of reducing the interference power, achieved by averaging the soft output of the received symbols, and the efficiency of separation provided in the channel passing through the transfer of many of the characters in the fading channel. In contrast, the method of batch code combination has the additional coding efficiency based on the code patterns in addition to e is RA with R=1/5 this turbocodes generates information symbols X, the first characters of Y0, Y0' and second parity symbols Y1, Y1' parity by encoding information symbols. Turbocodes consists of two partial encoders and one interleaver. The first characters of Y0and Y0' parity represent the outputs of the first private encoder by encoding input information symbols and the second symbols Y1and Y1' parity are the outputs from the second private encoder by encoding information symbols, perenesennyj in the interleaver. More, Yo represents the row of the first parity symbols generated from the first private coder and Y0' this is a line of second parity symbols generated from the first private coder.

Currently, due to the simplicity of the embodiment of most of the packet communication using the batch scheme combined with diversity and, in particular, a synchronous system IS-2000 and asynchronous UMTS system used in batches and combine with the division for the same reasons. However, most existing systems for packet communications using convolutional codes, and when using convolutional codes with low code rate R, these systems can not about the group. That is, when a system using a convolutional code with R=1/3, supports the scheme ASP, the difference between the schema of the packet code combining scheme and packet combining with the division insignificant, but because the system uses a batch method combined with diversity. However, when using turbocodes as code forward error correction (PIO) (FEC) requires a different method. This is because turbocode designed to be qualitative characteristics approaching lennonesque limit bandwidth, and their characteristics vary obviously depending on the speed of encoding unlike convolutional codes. It is therefore desirable to use a method of packet code combining in the packet communication using turbocopy to improve the system performance.

Therefore, the aim of the present invention is to provide a device and method for generating and decoding complementary turbocodes, given the characteristics of turbocodes in the communication system.

Another objective of the present invention is to provide a device and method for decoding quasicomplete of turbocodes in the communication system.

Thu is wascompletely of turbocodes in the communication system. The decoding device includes a code decoder for generating coded symbols according to a code rate transmitted through depreciating of subcodes are transmitted from the transmitter quasicomplete of turbocodes, and soft combining these subcodes, departmental to deteremine characters issued from the code decoder, and turbodecoding for decoding the output of deteremines.

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

Fig.1 illustrates the structure of the device is generating CCD according to a variant implementation of the present invention.

Fig.2 illustrates a procedure of generating quasicomplete of turbocodes according to a variant implementation of the present invention.

Fig.3 illustrates a method for selecting subcodes quasicomplete code according to a variant implementation of the present invention.

Fig.4 illustrates the structure of a receiver for receiving data transmitted by the transmitter of Fig.1, according to a variant implementation of the present invention.

Fig.5 is a functional block diagram illustrating the procedure of data processing in the receiver according to a variant implementation of the present invention.

Fig.7 illustrates a procedure of dividing the received data, store the divided data and the decoding of these data according to a variant implementation of the present invention.

The preferred embodiment of the present invention will be described below with reference to accompanying drawings. In the following description, well-known functions or constructions are not described in detail because they were cluttering the invention with unnecessary detail.

The present invention provides a method of using quasicomplete of turbocodes (CCTC) regardless of changes to the code length in the system that supports channel interleaving, using quasicomplete turbocode, or in a system that requires quasicomplete codes with different code rates. CCD is defined as the complementary code is generated by using a turbo code. CCTC is not perfect complementary code, as follows from the term "quasi" because the sub code includes repeated characters and is different from other sub-code characteristics, such as the ability to correct errors.

Fig.1 illustrates the structure of the device is generating CCD according to a variant of execution of this izopet the ary turbocodes (CCTC), performed after channel interleave.

In Fig.1, the encoder 101 encodes the input packet of the encoder, the coded symbols. Here, the encoder 101 can be used convolutional encoder or turbocode. In the example used here assumes that the code rate of the encoder 101 is R=1/5. Therefore, the encoder 101 receives 3072 information bits and generates 15360 coded symbols. Channel interleaver 102 punctuates encoded characters issued from the encoder 101, according to a predetermined rule. Here, if the encoder 101 is turbocodes, interleaver 102 punctuates each of the X code word and the symbols Y0, Y1, Y0' and Y1' parity on the basis of the properties of turbochager. The generator 103 CCD (or block weed/repetition CCTC) generates quasicomplete turbocode through thinning and repetition perenesennyj symbols provided from the channel interleaver 102. As shown above, the channel interleaver 102 and the generator 103 CCTC perform the process of generating CCTC.

As illustrated, if the number perenesennyj code is 15360, and the transmission speed of data (or code rate) of the subcodes specified as 307,2 kbit/s, the generator 103 CCD genereration perenesennyj code. If the data transfer speed is 614,4 kbit/s, the generator 103 CCD generates the first sub code by taking the first 10752 code symbols from the first half of perenesennyj code. And if the data rate is 1228,8 kbps or 2457,6 kbit/s, the generator 103 CCD generates the first sub code by taking 5376 code of perenesennyj code.

Here, the channel interleaver 102 should be designed in a special way to generate quasicomplete turbocode (or subcodes). This is because 5 characters, i.e., the X code word and the symbols Y0, Y1, Y0' and Y1' parity issued from the channel encoder 101 are distributed after you are exposed to the channel interleaving. It is difficult to realize a system for using distributed symbols as inputs in the unit thinning and repetition to generate quasicomplete of turbocodes, and mixed symbols X, Y0, Y1, Y0' and Y1' it is not easy to generate subcodes that meets the characteristics CCTC. To resolve this problem, the present invention provides a method of generating quasicomplete of turbocodes through a special method bastnasite is Arbatov according to a variant implementation of the present invention. In Fig.2, the encoder 201 encodes the input packet encoder in the characters. The encoder 201 uses the source codes with different code rates. Source codes are defined using their system. In this case, as the example used turbocode with R=1/5 as source code. Then, the encoder 201 generates information symbols X, the first characters of Y0and Y0' and second parity symbols Y1and Y1' parity by encoding input information symbols. The first characters of Y0and Y0' parity are issued from the first private encoder, and the second symbols Y1and Y1' parity from the second private coder. The first and second private encoders (not shown) contained in the encoder 201. Primary symbols Y0and Y1the parity of the first and second private coders have a higher priority of transmission than secondary characters Y0' and Y1' parity.

The demultiplexer (DEMUX) 202 demuxes 5 characters, including symbol X code words and symbols Y0, Y1, Y0' and Y1' parity issued from the channel encoder 201, in 5 groups. That is, the symbols X code words and symbols Y0the symbols Y1the symbols Y0' and the symbols Y1' parity consistently demultiplexers and is to randomly change the order of the sequences, issued from the demultiplexer 202, by alternation of sub-blocks. Suitable different ways alternation of sub-blocks, if the following condition is satisfied.

(Condition) Peremerzanie code symbols were subjected to thinning out partially so that the combination of thinned code symbols before interleaving has the same length thinning.

The reason satisfying the above conditions is that when a predetermined number of characters is weeded from the respective sets of symbols X, Y0, Y0', Y1and Y1' code words, the distance between thinned code symbols in the symbols of the code words before the alternation of sub-blocks must be equal in order to achieve optimal performance of the turbo code. In other words, when the decimation is applied to turbocodes, uniformity is a significant factor that determines the characteristics of turbocodes. In accordance with the present invention to code the symbols X, Y0, Y0', Y1and Y1' applies independent interleaving of sub-blocks. Uniform thinning in the output signal of each of the interleaver supports equal distance between thinned code symbols at the output of the encoder. Poetomu the characters could maintain uniform distribution of thinning at the output of the channel encoder.

Such methods channel interleave include alternation with the reverse order of bits (OPB) (BRO) and interleaving with partial reverse the order of bits (COPB) (PBRO). The alternation OPB practically applicable only if the number of input information bits in the encoder and the number of characters in the character sets X, Y0, Y0', Y1and Y1' code words generated from the source code that are powers of 2, i.e., 2mwhere m is a parameter to make the block size of the interleaver subblock same as the block size N=2mJ. Interleaving COBB was developed to satisfy the above condition, even when the number of characters in the corresponding character sets X, Y0, Y0', Y1and Y1' code words are not powers of 2 to overcome the limitations alternation OPB. A detailed description of this channel interleave sub-blocks is not given here, it should be noted that the present invention can be realized any way channel interleave, if satisfied the above condition.

The symbols of the code word, the randomized using alternation of sub-blocks, served on the appropriate blocks. Here re is input to the multiplexer 207 characters (or schemer characters). Peremerzanie symbols Y0and Y1parity respectively from the second and third premaritally 214 and 224 are input to the first multiplexer (MUX) 205, and peremerzanie symbols Y0' and Y1' parity, respectively, from the fourth and fifth premaritally 234 and 244 are served during the second multiplexer 215. The first multiplexer 205 multiplexes peremerzanie symbols Y0and Y1parity and delivers its output signal to the combiner 207 characters. The second multiplexer 215 multiplexes peremerzanie symbols Y0' and Y1' parity and delivers its output signal to the combiner 207 characters. Peremerzanie symbols of the code word is issued from premaritally, reordered, and then divided into three subgroups 206, 216 and 226.

The above described process, which is essential for generating CCD according to the present invention will be described in detail. As shown in Fig.2, the information symbols X form an independent subgroup without passing through the multiplexing after alternation of sub-blocks. Let peremerzanie the subblocks characters will be denoted as Sbi_X that can be written as

Sbi_X(l), Sbi_X(2), Sbi_X(3), Sbi_X(4)... (1)

where SbiX(1) the decree shall e outputs perenesennyj code symbols Y0and Y1from the second and third premaritally 214 and 224 are grouped into one subgroup. If the code symbols Y0represent Sbi_Y0then Sbi_Y0can be expressed in the form

Sbi_Y0(l), Sbi_Y0(2), Sbi_Y0(3), Sbi_Y0(4)... (2)

where Sbi_Y0(l) indicates the first code symbol output from the second interleaver 214. If the code symbols Y1represent Sbi_Y1then Sbi_Y1can be expressed in the form

Sbi_Y1(l), Sbi_Y1(2), Sbi_Y1(3), Sbi_Y1(4)... (3)

where Sbi_Y1(l) and Sbi_Y1(2) indicate, respectively, the first and second code symbols outputted from the third interleaver 224. After multiplexing code symbols Y0and Y1in the first multiplexer 205 to the output of the first multiplexer 205

Sbi_Y0(1), Sbi_Y1(1), Sbi_Y0(2), Sbi_Y1(2), Sbi_Y0(3), Sbi_Y1(3)... (4)

These multiplexed symbols are called sequence Century.

The reason for multiplexing perenesennyj code Sbi_Y0and Sbi_Y1is that when M consecutive SATELNODE, the number of thinned characters in Sbi_Y0equal to the number of thinned characters in Sbi_Y1,only if M is an even number. If M is odd, the difference between the number of thinned characters in Sbi_Y0and Sbi_Y1equal to only 1. Multiplexing is always satisfies the characteristic CCTK that the number of thinned characters Y0parity is equal to the number of thinned characters Y1the parity.

Similarly peremerzanie code symbols Y0' and Y1'that is output from the fourth and fifth premaritally 234 and 244, are grouped into one subgroup. If the code symbols Y0' and Y1' are respectively Sbi__Y0' and Sbi_Y1'then Sbi_Y0' and Sbi_Y1' can be expressed in the form

Sbi_Y0'(1), Sbi_Y0'(2), Sbi_Y0'(3), Sbi_Y0'(4)... (5)

and

Sbi_Y1'(1), Sbi_Y1'(2), Sbi_Y1'(3), Sbi_Y1'(4)... (6)

Then the output of the second multiplexer 215 can be written as

Sbi_Y0'(1), Sbi_Y1'(l), Sbi_Y0'(2), Sbi_Y1'(2), Sbi_Y0'(3), Sbi_Y1' (3)... (7)

These multiplexed symbols are called sequence C.

Pricina when M consecutive symbols were subjected to thinning out in sequence without regard to the first half or the second half of the sequence, the number of thinned characters in Sbi_Y0' is equal to the number of thinned characters in Sbi_Y1'only if M is an even number. If M is odd, the difference between the number of thinned characters in Sbi_Y0' Yves Sbi_Y1' is equal to only 1. Multiplexing is always satisfies the characteristic CCTK that the number of thinned characters Y0' parity is equal to the number of thinned characters Y1' parity.

A combiner 207 characters subsequently binds the sequence a, b and C of the first, second and third subgroups and generates a sequence [A:B:C] characters:

[A:B:C]=[Sbi_X(1), Sbi_X(2), Sbi_X(3),...] [Sbi_Y0(1), Sbi_Y1(1), Sbi_Y0(2), Sbi_Y1(2),...] [Sbi_Y0'(1), Sbi_Y1'(1), Sbi_Y0'(2), Sbi_Y1'(2),...]... (8)

As can be seen from this formula, in the sequence [A:B:C] the information symbols are arranged first, followed by alternating symbols Y0and Y1parity, followed by alternating symbols Y0' and Y1' parity. This placement of the characters is very essential when generating CCD that will be described below.

Thinning should be done in order to GE the blowing characteristics.

(1) Information symbols precede all other characters during transmission. This feature becomes especially important when the code rate of the subcodes are close to 1.

(2) the Combination of thinning is formed so that the number of parity symbols that are output from each of the private coder (the first private coder and the second private coder), or equal, their difference is minimal.

(3) the Number of thinned characters in symbols Y0and Y0' parity is defined so that the code rate of the first private coder was always less than 1. That is, the characteristic of turbocodes achieved when there is at least one character Y0or Y0' parity.

(4) the Distance between thinned characters in KKTC obtained by thinning the same.

(5) turbo code obtained by combining subcodes CCD, assumes the characteristics of quasicomplete code.

CCD with a code rate of subcodes, which is generated by thinning or reducing as many characters as you want, from one end of the sequence [A:B:C] characters, satisfies the above five characteristics. In other words, assigned to the sub code of CCD is generated by repetition and reduction of many symbological 209 characters. Repeater 208 sequences of symbols repeats in advance as specified in the sequence of characters adopted from the unifier of the characters. Repeat method is determined according to the code rate of the sub-code. Thinner 209 characters thins or reduces a predetermined number of characters, starting with the last character in the sequence of symbols adopted from the repeater 208 symbol sequences, so as to create sub code CCTC. The number of thinned characters depends on the code rate of the sub-code. So for repeater 208 sequences and thinner 209 characters shall be provided with a code rate of the sub-code to perform the repeat sequence and the decimation of the characters. Alternatively, the controller (not shown) can calculate the number of repeated symbols and the number of thinned characters according to the source code rate and speed of the sub-code and submit this information to the repeater 208 sequences and thinner 209 characters.

In other words, the thinner 209 characters selects a predetermined number of characters counted from the specified character position in the sequence of symbols adopted from repeat what imbalu, following the last character that was selected for the previous transmission. Therefore, the thinner 209 characters can be called the "selector symbols.

Premarital 203, 213, 223, 233 and 243, the multiplexers 205 and 215, and a combiner 207 symbols in Fig.2 correspond to the channel interleaver 102 in Fig.1 and the repeater 208 sequences and thinner 209 characters both correspond to the generator 103 CCTC.

In Fig.1, assuming the original code rate R=1/5 and 3072 input information bits, the channel encoder 101 generates 15360 code. Below is a description of the generation CCD with different code speed (or speeds), for example, the first CCD0jon 307,2 kbps, the second CCD C1jon 614,4 kbps and third CCD C2jon 1288,8 kbit/s, of the code.

As described previously, 15360 code symbols are classified into five groups, interspersed, and then again placed as a sequence of characters according to equation (8). Then these 15360 code characters are repeating according to a predetermined rule, and were subjected to thinning out (or reduced) according to a predetermined code rate of the sub-code. Thus generated appointed sub code.

For data rate 3 the om first 21504 characters from peremeshennoi and repeated sequences of characters. The second sub code C01generated by selecting 21504 characters from repeated sequences of characters starting with the character following the first sub-code With a00. The third sub code With02is generated by selecting the following 21504 characters.

Similarly, for data transfer speeds 614,4 kbps, if the subcodes of the second CCD C1jhave a length 10752 bits, the first sub code With10is generated by selecting the first 10752 characters from repeated sequences of characters. In other words, the first sub code With10is generated by the reduction of all other characters after the first 10752 characters in repeated sequences of characters. This reduction is performed in the thinner 209 characters, as described above. The second sub code11generated by selecting 10752 characters from repeated sequences of characters starting with the character following the first sub-code With a10. The third sub code C12generated by selecting 10752 characters.

Similarly, for data transfer speeds 1228,8 kbps, if subcodes third CCD2jhave a length 5376 bits, the first sub code With20is generated by selecting the first 5376 characters from peremeshennoi sequence of characters. The second sub code C21generated by selecting 5376 characters from peracod22is generated by selecting the following 5376 characters. This generates the subcodes CCTK on 1288,8 kbit/s

The system stores information about the position of the last character in the previous transmitted sub-code for each CCD. When the determined data transmission rate (or code rate) for the second transmission, the system selects CCD corresponding to the data rate, and generates sub code by selecting a predetermined number of characters following the last stored character for the selected CCD according to the data rate. If the selected symbols exceed one unit perenesennyj characters, the remaining characters are selected from the next block. In this method, the subcodes are generated by the repeating unit perenesennyj characters. To do this, the required area, remember to save repeated blocks.

Alternative peremerzanie characters can be stored in the ring buffer, and the sub code is generated by a recursive selection of characters. That is, if peremerzanie characters selects all, then a predetermined number of symbols is selected from perenesennyj characters, starting with the first character. Then the repeater 208 sequences of characters can be omitted, because the ring is t perform the present invention describes a two-dimensional CCD. In the two-dimensional scheme CCD quasihereditary turbo code corresponding to each code rate, is generated independently, and subcodes this CCTK passed each other. However, two-dimensional CCD not optimal for the reasons described below.

As shown in Fig.2, it is assumed that the first sub code With00first CCD C0jused for initial transmission, the first sub code With10second CCD C1jused for the next transmission, and the first sub code With20third CCTC C2jused for third gear. Then the receiver decodes the data, combining these three sub-code (C00With10With20). In this case, however, the combined code does not restore the original code with a code rate of 1/5, only increasing symbolic energy information symbols and thereby optimizing the characteristics of the decoding. This leads to the problem with the transfer order under the code, i.e. the choice of subcodes. To overcome this problem are proposed adaptive CCTC. In the adaptive scheme CCTK the number of code symbols that are subject to selection, is determined according to the code rate of the sub-code and sub code is generated by selecting a specified number of characters starting with the character following Ianto execution device is generating CCTC. The design shown in Fig.3 is the same as that shown in Fig.2, except repeater sequences and thinner characters, working otherwise. Therefore, the following description is made primarily for repeater 308 sequences and thinner 309 characters.

Repeater 308 sequences of symbols repeats in advance as specified in the sequence of characters adopted from the combiner 307 characters. This repetition can be performed according to the specified parameter in the repeater 308 sequences under control of a controller (not shown), or on request from the combiner 307 characters. The above processes are embodied in the same manner as that described with reference to Fig.2. Further, the thinner 309 characters thins characters received from the repeater 308 sequences according to a rule different from the rules applied to Fig.2, to generate a sub-code. This rule thinning following.

It is assumed that the transmission starts at time k was at the time (k+h) sub code is expressed as Cij(k+h), a code characters of the source code with R=1/5 arem(0), Cthe L_INF denotes the size of the interleaver subblock or the number of information symbols.

Step 1: determine the length of the initial sub-code.

For the initial transmission selects the one With thei0of the first under code00With10With20available CCD according to a given code rate, and the length of the selected sub-code With ai0saved as a variable L_SC. Code speed or length L_SC of the sub-code is set in advance in the system according to channel environment, including conditions channel transmission and the input data rate. The description is done in the context of three CCD shown in Fig.3, for a better understanding of the present invention, but it should be borne in mind that the number of subcodes is not limited to this number.

Step 2: select and transmitted sub code for the initial transmission.

Once you have identified the length to transfer the sub-code, among the characters of the source code are selected Withm(0)m(1),..., Cm(L_SC-l). If L_SC exceeds N, thenm(0)m(1),..., Cm(N) transferred to P times, and then passed Withm(0)m(1),..., Cm(q-l). Here P and q represent the quotient and the remainder from L SC/N, respectively, and P and q are calculated as L_SC mod N. Then the variable q is saved for the next transmission for use in determining the position of the last character of the previous transmitted sub-code for otina of the sub-code.

For the next transmission code rate R_SC new sub-code to be transmitted, is determined according to the channel environment, and the length L_SC of the sub-code is determined according to the found code rate. Length L_SC and code rate R_SC are in the ratio of

L_SC=L_INF(1/R_SC) (9)

The system passes the length of the sub-code and code rate R_SC of the sub-code on the thinner 309 symbols for each transmission.

Step 4: select and transmitted sub code for the next transmission.

Once you have identified the length L_SC to transfer the sub-code, among the characters of the source code chosen Cm(q), Cm(q+1),..., Cm(q+L_SC-1). In other words, from the characters of the source code is selected as many characters, what is the length of the sub-code, starting with the character following the last character that was selected for the previous transmission. If q+L_SC exceeds N, the string consisting of N code symbols, starting with Cm(q), is selected recursively and transmitted P times, then the other q' code symbols are transmitted sequentially. Here P and q' represent the quotient and the remainder from (L_SC)/N, respectively, and q' is calculated as (q+L_SC) mod N. Then the value of the next character position to the position of the last selected character for Leo symbol among the symbols formed last transferred to the sub code. Once generated sub code transmitted, the procedure returns to step 3.

Transmission adaptive CCTC will be clearer for the case shown in Fig.3. In Fig.3 low-speed sub code with a code rate of 1/7 is transmitted first, in Case 1, and a high-speed sub code with a code rate of 4/7 is transmitted first in Case 2. As can be seen from these Cases, N(=15360) successive characters of the source code again, and so much code, what is the size corresponding to the length you want to transfer sub-code (or code rate of the sub-code) are selected sequentially from the repeated characters of the source code at each transfer.

In a real embodiment for storing repeated (R-1) times the source code does not use the buffer, and applies only ring buffer to store N code symbols and recursively select code symbols, thereby generating a sub code assigned length. That is, the use of the ring buffer eliminates the need to repeat the sequence. Any receive buffer suitable for the receiver, if it can save N soft metrics for combining codes.

Next will be described a method of receiving data, re is s Fig.3.

Fig.4 illustrates the structure of a receiver for receiving data transmitted by the transmitter of Fig.1. In Fig.4 signals 401, 402 and 403 of the data transmitted by the transmitter to its associated data rates, are subcodes of CijKKTK. Received signals 401, 402 and 403 of the data fed to the CPU 411 KKTK (deponirovanie/combining KKTK). The processor 411 KKTK performs deponirovanie on the received signals to convert the received signals to turbocode with their original code rate R from turbocodes based on the characteristics of the KKTK, and performs soft combining on the adopted subcodes. "Deponirovanie" is a process for introducing a character wipe at the position of the thinned character. For soft combining of the adopted subcodes you can use Chase combining. As mentioned with reference to the characteristics of the KKTK, the receiver generates code symbols with a code rate R by soft combining under code Cijtransmitted from the transmitter.

Now will be described the receiver taking into account the fact that the code rate R=1/5, as in the transmitter, where the code rate is R=1/5. Adopted under codes Cijsimilarly pererazmeschaet and gently combine with the receiver soglasna has different properties metrics in accordance with the number of bits per received symbol, the present invention considers even hard combination as soft combination with one bit resolution. In fact, since the performance does not improve significantly when the hard combination, there will be described a soft combining. A detailed description will be given together with a description of the respective functional blocks in the receiver.

The processor 411 CCD generates N gently combined symbols of the code words, initially generated by the encoder by deponirovanie and soft combining, and transmits them to the channel depechemiami block 421. Gently combined symbols of the code words represent values report generated from the received symbols of the code words expressed by many bits or real numbers, and N represents the number of symbols of the code words issued from the encoder used in the transmitter. Channel depechemiami block 421 impeller departmeat gently combined symbols of the code words received from the CPU 411 CCTC, and transmits N deprimerende gently combined symbols of the code words in the channel decoder 431. Channel decoder 431 decodes N deprimerende gently combined character code used in the transmitter.

Fig.5 illustrates a functional block diagram of the procedures for processing received signals in the receiver according to a variant implementation of the present invention. The structure and operation of the respective functional blocks will be described in detail with reference to Fig.5.

In Fig.5 buffer 501 subcodes reception (bothersome adopted subcodesij) saves the subcodes Cijtaken from a transmitter. The size of this buffer depends on the number N of symbols of the code words, and if one character is used Q bits for the buffer 501 subcodes of used QN-bit memory. Further information about the type of sub-code, adopted in every moment, and the moment of transfer of subcodes of a code word with R=1/5 can be recognized by the receiver via a control channel or a control message transmitted together with subcodes. The received symbols are stored in connection with the subcodes are due to the previously transferred data in the buffer 501 subcodes reception. The way to save the received symbols in the buffer with N cells will be described in detail with reference to Fig.7.

Schemer sequence Combinator symbols 502 (combining sequence - combining characters of the adopted subcodes with thinning) performs the inverse process to repeat what molov transmitter, at the same time, when the admission process. That is, the buffer 501 under codes accepting gently combines previously received symbols with symbols taken from the current channel. If the number of received code symbols corresponding subcodes are less than N, this means that the transmitter used decimation. In this case, the buffer 501 subcodes admission introduces the characters erase in the corresponding character position code words before their soft combining. When combining sequences used a combination of Chase buffer 501 subcodes admission applies different weights to the code word of the previously received symbols and the code word symbols taken from the current channel before combining. If it is assumed that the weight of the code words of the previously received symbols stored in the buffer 501 subcodes reception, is represented as w1, and the weight of the code words of the characters taken from the current channel, represented as w2, the weight w1 and w2 are applied to the respective soft metrics for combining. Weight set with block channel estimation in the receiver. The Chase algorithm for soft combining sequences (or code words) are well known, so that its detailed description will be omitted for Prost code words and passes them to the separator 503 code words (the separation or splitting of the sequence of code words (A:B:C)). In Fig.5 soft metrics for the N symbols of the code words issued by the schemer sequences - Combinator symbols 502, for the sake of convenience, is represented as "D".

The switch 503 code words separates the soft metrics for the N symbols of the code words in part X information symbols, the first part of Y0, Y1character parity and the second part of Y0', Y1' symbols of parity. Demultiplexes M1 (515) and M2 (525) demultiplexer accordingly, the first part of Y0, Y1character parity and the second part of Y0', Y1' symbols of parity, thereby pererazmeschaet them in part X information symbols, the first part of Y0, Y1character parity and the second part of Y0', Y1' symbols of parity. This operation can be performed either sequentially or simultaneously. Next part X information symbols, demultiplexing the first part of Y0, Y1character parity and demultiplexing the second part of Y0', Y1' character parity is again separated into a stream of five symbols X, Y0, Y0', Y1and Y1' symbols of the code word and transmitted to podlakovye deprimida blocks 506, 516, 526, 536 and 546, respectively.

These podlakovye deprimida blocks is=5 (where R=1/5) threads X, Y0, Y0', Y1and Y1' symbols of the code words. Odblokowanie deteremine either possible so that 5 characters of the code word can depresiates by deteremine one subblock, or so that the symbols of the code words can depresiates independently through so many of detereminately as there are symbols of the code words. The description of the present invention is not limited to the implementation of specific odblokowac deteremine, but given the assumption that the reverse process odblokowany the alternation, in the General case used in the transmitter, is odblokovani depechemiami blocks in the receiver.

5 combined - deprimerende characters (507, 517, 527, 537, and 547) are subjected to multiplexing 508, where they recombinants in the process inverse to that performed by the demultiplexer 302 in the transmitter. In the end, the decoder 509 decodes the soft metrics and displays the transmitted information symbols, i.e., the encoded packet.

The procedure, performed relevant Funktsionalnyi blocks in the receiver shown in Fig.6. As is illustrated in Fig.6, it is assumed that the subcodes are transmitted by the transmitter to the present, is C00With10that is the sub-code, having 10752 symbols of the code words, and With20and C21are subcodes having 5376 symbols of the code words. Therefore, to date, the receiver has received a total of 4 sub-code, all of which were passed as subcodes having different code rate of subcodes, through one 3072-bit encoded packet information unit. Thus, the receiver would have to generate soft metrics for the N code words by soft combining of subcodes of the above-described manner. Therefore, as shown in Fig.6, the receiver gently combines 4 sub-code so that the positions of 15360 (=30725) character code word for the code word with R=1/5 would have to be identical to the positions of the symbols of the code word of each sub-code. In addition, as illustrated in Fig.6, since00with the length of the sub-code 21504 longer than N (i.e., 15360), the receiver sequentially places 6144 (=21504-15360) symbols of the code words remaining after the allocation 15360 characters, as in the way of repeat sequences, starting again from the beginning, and then gently combines placed the symbols of the code words. Similarly, as With10transmitted by the transmitter after00in visheukazannogo, since20and C21transmitted by the transmitter after10in the above transmission method, they are added after the end of the10in the receiver and gently combine.

This work will be described in detail with reference to Fig.7. As shown in Fig.7, the receiver can use the N - or NxQ-bit buffer to implement a circular buffer, or use a buffer memory of a fixed size and design of the generator buffer address to generate the ring address. In addition, as shown in Fig.7, p00stores N characters, starting with the start address addr:00, a then stores 6144 (=21504-15360) characters in the buffer. Because this step is to preserve these characters after saving N characters, symbols, gently combined with the previously stored characters in the above way. Let the address, which is part of the soft combining, will addr_. Then, if the trace is taken With10received characters are stored in the buffer by advancing on 10752 bits from addr_A. Since this is also the step of storing characters after saving N characters, symbols, gently combined with the previously stored characters in the above way. May address, in which vypolnyayutsya in the buffer by advancing on 5376 bits from addr_. Let the address, which is part of the soft combining, will addr_. Then, if the trace is taken With21received characters are stored in the buffer by advancing on 5376 bits from addr_. Let the address, which is part of the soft combining, will addr_D. Then the receiver continuously performs soft combining on the subcodes are transferred one coded packet in the above method, and generates the soft metrics for all N symbols of the code word after the completion of this process. This method can be viewed as a way of implementing the method of generating of subcodes for CCD in the transmitter. Summarizing, the method includes step 1 determine the length of the original sub-code, step 2 identify and transfer sub-code to be transmitted, step 3 determine the position of the code symbol to transfer the following, and determine the length of the transmission, and the step 4 of the definition and transmission of the sub-code to transfer the following. Accordingly, the receiver can gently combine the subcodes in connection with code words with R=1/5 on the basis of information about the type of subcodes, transmitted by the transmitter in the way of the ring buffer.

Returning to Fig.6, the separator code words separates the soft metrics for N makingyou parity and the second part of Y0', Y1' parity symbols.

Group separation. Then demultiplexes Ml (515) and M2 (525) demultiplexing the first part of Y0, Y1character parity and the second part of Y0', Y1' symbols of parity, respectively, to thereby pererazmeschaet them in part X information symbols, demultiplexing the first part of Y0, Y1character parity and demultiplexing the second part of Y0', Y1' symbols of parity. This operation can be performed either sequentially or simultaneously. Following this, part X information symbols, demultiplexing the first part of Y0, Y1character parity and demultiplexing the second part of Y0', Y1' symbols of parity again divided into the flow of the 5 symbols X, Y0, Y0', Y1and Y1'and then passed on to podlakovye deprimida blocks 506, 516, 526, 536 and 546, respectively.

Podlakovye deprimida units perform the reverse process odblokowany the interleaving performed at the transmitter, NR=5 streams of symbols X, Y0, Y0', Y1and Y1' code words. Odblokowanie deteremine either possible so that the 5 character code Siemiatycze independently through so many of detereminately, as there are symbols of the code words. The description of the present invention is not limited to the implementation of concrete block deteremine, but given the assumption that the reverse process odblokowany the alternation, in the General case used in the transmitter, is odblokovani depechemiami blocks in the receiver.

Finally, the decoder decodes the soft metrics for the N symbols of the code words received from odblokowac deprimida blocks, and outputs the transmitted information symbols, i.e., the encoded packet.

As described above, the communication system according to the present invention can generate complementary turbocode and quasicomplete turbocode. This communication system can significantly improve its throughput by applying complementary turbocodes to the scheme ASP.

Although the invention is shown and described with reference to some preferred ways of its implementation, it should be understood that it can be made various changes in form and detail without departing from the essence and scope of the invention as defined by the attached claims.

Claims

1. Device for receiving quasicomplete turpene code rate and to perform soft combining on the adopted subcodes by sequence conservation defloriani characters; channel departmental gently to separate the combined characters issued from depreciates in the stream of information symbols and at least one stream of parity symbols, for demuxing streams of parity symbols in a pair of streams of parity symbols, and then for the separate removal of streams of symbols, parity and flow of information symbols; and turbodecoding to multiplex streams of characters parity and flow of information symbols, issued from the channel deteremines, for decoding multiplexed output according to the code rate and to output a stream of information symbols.

2. The device under item 1, in which the channel departmental contains a separator for separating the output depreciates CCTK on the stream of information symbols and at least one stream of parity symbols; at least one demultiplexer for demuxing streams of characters parity issued from the separator, a pair of streams of parity symbols; and at least one departmental to deteremine outputs and demultiplexes the stream of information symbols.

3. Device for receiving quasicomplete of turbo code (CCTC) and decoding CCTC, the content is selenia combined symbols in the stream of information symbols and parity symbols, to pair separated by respective streams of characters parity at least one other stream of parity symbols, for demuxing paired streams of characters parity of at least one stream of parity symbols according to a given code rate and independent deteremine and output a stream of information symbols and demultiplexing streams of parity symbols; and a decoder quasicomplete of turbo code for multiplexing independently perenesennyj streams of characters parity and flow of information symbols and after decoding according to a predetermined speed decoding excretion stream of information symbols.

4. The device according to p. 3, in which the channel departmental contains separator characters for separating and removing a stream of information symbols and flow of parity symbols from the combined symbols; a demultiplexer for demuxing the streams of the parity symbols associated with another thread of parity symbols, and to separate the respective threads of parity symbols; and departmental for independent deteremine demultiplexing streams of parity symbols and the corresponding thread is tiplexer to derive independently deprimerende streams of characters parity and flow of information symbols; and turbodecoding to output a stream of information symbols after decoding the output symbols of the multiplexer according to a predetermined speed decoding.

6. The device according to p. 3, in which the Configurator contains the ring buffer memory to save the received symbols; and the schemer sequences for selection and removal of a predetermined number of characters from the characters stored in the ring buffer, according to the code rate, starting from its initial position.

7. The device according to p. 6, in which the initial position of the ring buffer is the character following the last character transmitted by the latter character, whatever was passed a stream of characters.

8. The device according to p. 6, in which the smooth sequences of symbols combines with the previously transmitted symbols and outputs on channel departmental, if there is a retransmitted symbol of the received symbol.

9. The device under item 8, in which the combination is a soft combination.

10. The device under item 8, in which the combination is a hard combination.

11. The method quasicomplete of turbo code (CCTC) and decoding CCD, containing the following steps: (a) posledovatelyam and streams of the parity symbols, the pair separated streams of characters parity at least one other stream of characters parity, demultiplexing paired streams of parity symbols in at least one stream of parity symbols according to a given code rate and independent deteremine and excretion of the stream of information symbols and demultiplexing streams of parity symbols; and (C) multiplexing independently perenesennyj streams of characters parity and flow of information symbols and, after decoding it according to a predetermined speed decoding, outputting the stream of information symbols.

12. The method according to p. 11, in which step (b) includes the following steps: (a) the separation and removal of the stream of information symbols and flow of parity symbols from the combined symbols; (b) the demux streams of parity symbols associated with another thread of parity symbols, and the separation of the respective streams of parity symbols; (C) independent deteremine demultiplexing streams of parity symbols and the corresponding stream of information symbols.

13. The method according to p. 11, in which step (C) includes the following steps: (a) a multiplexer for multiplexing and excretion independently researched the origin of the stream of information symbols after decoding the output symbols of the multiplexer according to a predetermined speed decoding.

14. The method according to p. 11, in which step (a) includes the following steps: ring the preservation of the received symbols; and selecting and removing a predetermined number of characters from the characters from the retaining ring according to the code rate, starting from its initial position.

15. The method according to p. 14, in which in step (a) initial position of the ring buffer is the character following the last character transmitted by the latter character, whatever was passed a stream of characters.

16. The method according to p. 14, in which in step (b) schemer sequences of symbols combines with the previously transmitted symbols and outputs on channel departmental, if there is a retransmitted symbol of the received symbol.

17. The method according to p. 16, in which the combination is a soft combination.

18. The method according to p. 16, in which the combination is a hard combination.

19. The method according to p. 11, in which in step (b) information symbols and the many streams of characters parity deprimida respectively independently by way of deteremine partially reverse the order of bits (COPB).

20. The pickup device quasicomplete of turbo code (CCTC) and decoding CCD containing Combinator for sequential combinetransformation characters and threads of parity symbols from the combined symbols; a demultiplexer for demuxing the streams of the parity symbols associated with another thread of parity symbols, and separation of the respective streams of parity symbols; departmental for independent deteremine demultiplexing streams of parity symbols and the corresponding stream of information symbols; and a decoder quasicomplete of turbo code for multiplexing independently perenesennyj streams of characters parity and flow of information symbols after decoding according to a predetermined speed decoding excretion stream of information symbols.

21. The device according to p. 20, in which the interleaver punctuates the flow of information symbols and the many streams of parity symbols respectively and independently by way of deteremine partially reverse the order of bits (COPB).

22. The device according to p. 20, in which the Configurator contains the ring buffer memory to save the received symbols; and the schemer sequences for selection and removal of a predetermined number of characters from the characters stored in the ring buffer, according to the code rate, starting from its initial position.

23. The device according to p. 22, which began the R the last character, when the matter was passed a stream of characters.

24. The device according to p. 22, in which the smooth sequences of symbols combines with the previously transmitted symbols and outputs on channel departmental, if there is a retransmitted symbol of the received symbol.

25. The device according to p. 24, in which the combination is a soft combination.

26. The device according to p. 24, in which the combination is a hard combination.

 

Same patents:

The invention relates to systems and networks, spread spectrum and multiple access, and more particularly to the elimination of the (initial) uncertainty synchronization in the transmission taken on the access channels in the communications system spread spectrum

The invention relates to a device and method for providing data services in a mobile communication system

The invention relates to the field of communications, in particular intended for the formation of a pseudo-random noise sequence with the ability to quickly summarize from one shift to another

The invention relates to the provision of reliable data transfer

The invention relates to the field of radio communications

The invention relates to a wireless communication system

The invention relates to the field of railway transport

The invention relates to digital data transmission

The invention relates to a mobile communication system to transmit information over the channel, the pilot signal

The invention relates to the field of encoding for communications systems and can be applied, in particular, to premaritally for turbocodes

The invention relates to mobile communications systems

The invention relates to communication systems

The invention relates to turbocodes used in radio systems, digital cellular systems, digital networks integration services

The invention relates to a device data transmitted from the transmitter to the receiver

The invention relates to methods and devices for adaptive channel coding for communication systems

FIELD: coding in communication systems.

SUBSTANCE: proposed partial reverse bit-order interleaver (P-RBO) functions to sequentially column-by-column configure input data stream of size N in matrix that has 2m lines and (J - 1) columns, as well as R lines in J column, to interleave configured data, and to read out interleaved data from lines.

EFFECT: optimized interleaving parameters complying with interleaver size.

4 cl, 7 dwg, 3 tbl

Up!