CN1372394A - Method for transmitting bit sequence through radio channel - Google Patents

Abstract

The present invention relates to a method for performing channel coding by using a 8 PSK modulating technology and special RM codes in a 1.28Mcps TDD mode. But RM codes can only correct bundle errors or symbol errors. The identifying and correcting ability of errors can be improved greatly by using b-phase adjacent codes when the channel coding is performed. Thus, the TFCI bit sequence to be sent is grouped into b-bit data symbols (b=3), then,the data symbols (b=3) are encoded systematically. Each generated code word consists of two data symbols (d1, d2) representing TFCI data with 3 to 5 bit length and six redundance symbols. 3 adjacent codes generate the first redundance symbol group from the first data symbol (d1), and generate the second redundance symbol group from the second data symbol (d2), and generate the third redundance symbol group from the first and the second data symbol.

Description

The method of transmitting bit sequence through radio channel

Background technology

The current version that is called as the UMTS standard of Release (issue) 99 (grade 06/2000) has two kinds of Radio Transmission Technologys at object: fdd mode and 3.84Mcps tdd mode.For next version that is called as Release 2000 of UMTS standard, stipulate that now the 1.28McpsTDD pattern is as the third Radio Transmission Technology.1.28Mcps tdd mode relates to the mobile radio system TD-SCDMA of China, it also resembles the technology that has made up TDMA and CDMA the 3.84Mcps tdd mode.Under the 1.28Mcps tdd mode, between up link and down link, transmit data with a frequency by time multiplexing.

Under the 1.28Mcps tdd mode, data in time slot with fixing structure, also promptly transmit by mobile radio channel with so-called pulse train.Fig. 1 shows the burst structure of standard time slot.Described pulse train is made up of two data blocks, an intermediate sequence part and a guard time (so-called guard time GP) that is used for channel estimating.This pulse train comprises 864 chips altogether, and wherein said two data blocks comprise 352 chips respectively, and described intermediate sequence partly comprises 144 chips, and described GP comprises 16 chips.In the chip frequency is that (corresponding to the chip time delay is T to 1.28Mcps _cUnder=781.25ns) the situation, described burst length is 675 μ s.According to selected spreading factor, each pulse train can transmit different data volumes.

Except useful data originally, need data block transmitted also to have so-called TFCI bit, this bit comprises the important information about the composition of useful data.Described data (useful data and TFCI) are coding in advance in transmitter usually, so that discern and proofread and correct the bit error that may occur in the wireless transmission process by coding/decoding method in receiver.Because the Special Significance of TFCI bit will adopt error-preventing method very reliably when these data types of coding, can cause useful data to produce wrong reconstruct under the stable condition because the mistake of these bits is received in.

Under the 1.28Mcps tdd mode, defined two kinds of digital methods at present and modulated the described interior data of data block that are included in the standard time slot: be used for the QPSK and the 8PSK that be used for data rate 2Mbps of data rate up to 384kbps.At this, modulation is meant that the data according to the need transmission change the carrier signal of high frequency.In QPSK, always 2 coded data bit combinations are become a data symbol.This data bit utilizes 0 ° in phase place, 90 °, 180 ° or 270 ° to come modulated carrier signal then.In 8PSK, always constitute a data symbol by 3 coded data bits, this data symbol utilizes 22.5 ° in phase place, 67.5 °, 112.5 °, 157.5 °, 202.5 °, 247.5 °, 292.5 ° or 337.5 ° to come modulated carrier signal more then.

Under the QPSK of 1.28Mcps tdd mode modulation case, stipulated the chnnel coding of TFCI bit, and also not open under the situation of 8PSK modulation technique.Under the situation that adopts the 8PSK modulation technique, be that the method for implementing Samsung (Samsung) comes the TFCI bit is carried out chnnel coding at present.According to the quantity of TFCI bit, can use different Reeds-Miu Le sign indicating number (RM) herein.For length is the TFCI of 6-10 bit, can use so-called " subcode of (64,10) second order Reed-Miu Le sign indicating number ", and 16 bits of " perforation " (removing) subsequently.After coding, can obtain length is the TFCI code word of 48 bits.This encoding scheme as shown in Figure 2.

For length is the TFCI of 3-5 bit, can use so-called " (32,5) biorthogonal (sign indicating number of or single order Reed-Miu Le) ", and 8 bits of " perforation " (removing) subsequently, makes described TFCI code word have the length of 24 bits (referring to Fig. 3).

The single error of the random distribution of Reed-Miu Le sign indicating number in can the correction data symbol.Yet when mobile radio channel transmits, it is that form with single error is disturbed that data have more than.Specifically, also can occur so-called bundle mistake or symbol error continually by the utmost point, can bring problem and proofread and correct them by Reed-Miu Le sign indicating number.

Summary of the invention

The objective of the invention is to create a kind of by being disturbed wireless channel, especially come the method for transmitted bit pattern by mobile radio channel, utilize this method both can discern and proofread and correct single error clearly, can discern again and corrective beam mistake and symbol error.Method with the described feature of claim 1 possesses this performance.This method especially can be used in 3-G (Generation Three mobile communication system), for example under the 1.28Mcps of UMTS tdd mode the TFCI bit is carried out chnnel coding and transmission.Dependent claims relates to the advantageous embodiment and the expansion scheme of the described method of claim 1.

For the RM sign indicating number that is used for chnnel coding so far, the method that the present invention is based on the b adjacent code also has following advantage:

The realization of aspect hardware and software is simpler.

The better error correction of.

The better wrong identification ability of.

Faster and the simpler decoding of.

This realizes in the following way, promptly

especially selects and constructs the b adjacent code useful data (TFCI) of encoding, and this useful data divides into groups by the b bit and modulates by the b bit,

especially is grouped into a plurality of code subframes with different abilities to the data bit that is encoded, and

adopts special decoding agreement to carry out error correction.

Description of drawings

Tell about the present invention by accompanying drawing below.Wherein:

Fig. 1 shows the burst structure of the standard time slot under the 1.28Mcps tdd mode;

Fig. 2 shows the coding (8PSK) of 6-10 TFCI bit so far;

Fig. 3 shows the coding (8PSK) of 3-5 TFCI bit so far;

Fig. 4 shows the structure of the code word that the b adjacent code by 1 symbol error-correcting produced;

Fig. 5 shows the embodiment of the encoder that is used for single 3 adjacent codes;

Fig. 6 shows the embodiment of the decoder that is used for single 3 adjacent codes;

Fig. 7 shows the 24 bits of encoded schemes that length is the TFCI data of 3-5 bit;

Fig. 8 shows and is used for the decoding algorithm that length is the code word of 24 bits;

Fig. 9 shows the 48 bits of encoded schemes that length is the TFCI data of 6-10 bit;

Figure 10 shows and is used for the decoding algorithm that length is the code word of 48 bits;

Figure 11 shows and is used for 3 adjacent code C1/C2 one or two data symbol, that channel is optimized;

Figure 12 shows the generator matrix P1 (coding of a data symbol) that is used for code C1;

Figure 13 shows the generator matrix P2 (codings of two data symbols) that is used for code C2;

Figure 14 shows the channel optimization coding of the TFCI data block of being made up of the 3-5 bit (b=3);

Figure 15 shows the channel optimization coding of the TFCI data block of being made up of the 6-10 bit (b=3);

Figure 16 show to or be that the data symbol of the TFCI data block of 3-5 bit distributes and produces redundant symbol from length;

Figure 17 show to or be that the data symbol of the TFCI data block of 6-10 bit distributes and produces redundant symbol from length;

Figure 18 shows the embodiment of the encoder of 3 adjacent codes that are used for channel optimization;

Figure 19 shows the embodiment of the decoder of 3 adjacent codes that are used for channel optimization;

Figure 20 shows the embodiment that is used for drawing from a data symbol and two redundant symbols the device of symptom;

Figure 21 shows the embodiment that is used for drawing from two data symbols and two redundant symbols the device of symptom;

Figure 22 shows the first symptom mistake allocation table;

Figure 23 shows the second symptom mistake allocation table.

Embodiment

Method of the present invention has been utilized the special characteristics of b adjacent code, wherein, needs the bit of transmission to be combined into the group that length is the b bit (symbol), and presses encoding symbols [1,2].Can when chnnel coding, consider modulating/demodulating thus, or in turn tuned channel is encoded and selected modulation system.Under QPSK or 8PSK modulation case, resemble article beginning described the data-bit-group Synthetic 2 (QPSK) of need transmission or the group of 3 (8PSK) bits, and be mapped in the corresponding data symbol.The coded system of this consideration modulation especially is very effective when being applied to mobile radio system, is the data symbol of b bit because the b adjacent code allows gauge length, and irrelevant with the quantity and the type of bit error in each symbol.The method of this place suggestion can be applied to the various mobile radio systems by b bit transfer data.The application of the b adjacent encoder method the during value of telling about b=3 below, and be assumed to the 8PSK modulation of high-frequency carrier signal.But this does not limit the b value that correspondingly is applicable to other.

B adjacent code by a kind of 1 symbol error-correcting produces code word shown in Figure 4.It is by maximum N=(2 ^b-1) individual data symbol and two redundant symbols are formed.Opposite with Reed-Miu Le sign indicating number, this code is a system, and it is constant to that is to say that data bit is rendered as in code word.Under 3 adjacent code situations of 1 symbol error-correcting, described code word comprises 7 data symbols and two redundant symbol R1, R2 thus at most, and wherein, the length of data symbol and redundant symbol is 3 bits.The mismark that goes out that any position occurred in code word can be discerned and proofread and correct to the code of this system.

Fig. 5 shows the possible hardware configuration that encoder adopted (referring to the Fig. 3 of [1]) of 3 adjacent codes of 1 symbol error-correcting.By a kind of wide by four adders, three words be that the wide encoder that is the feedback shift register of 3 bits is formed of the shift register of 1 bit and word comes data symbol d0, d1, the d2 from the input that imposes on this encoder ... middle redundant symbol R1, the R2 that generates 3 bit long, wherein d0:=(d0 ₁, d0 ₂, d0 ₃), and the rest may be inferred.Prime polynomial g (x) :=1+x+x ³The functional multinomial that is used as R2.Described encoder includes only above-mentioned hardware components, and its function and acting in conjunction are open by [1].The very clock pulse of two-forty is provided for described encoder.The structure diagram of the decoder of Fig. 6 under the showing Fig. 3 of (equally referring to [1]), wherein, from the data that are applied to input side and redundant symbol R1/R2 ... the logical circuit that draws two symptom S1 and S2 among d2, d1, the d0 is structurally corresponding to encoder shown in Figure 6.

Carry out error correction by combinational network M1 single, that tell about in [1] equally in the shortest time, described network M1 locatees mistake by analyzing two symptom S1 and S2.At this, symptom S1 represents error-pattern.If one of described two symptoms equal 0, then must in redundant symbol, there be mistake.Obtain bit-error locations according to S1 and S2, and only relate to described data symbol.This means, when the bit-error locations of obtaining during greater than the data symbol number that exists, generated error identification signal always then.How to determine described bit-error locations from two symptoms for network M1, [1,2] is seen in its explanation.

Set forth the function of Fig. 5 and unit (encoder or decoder) shown in Figure 6 below once more by embodiment.Be in the reason of simplification, only have d2:=(1,0,1); Three data symbols of d1:=(0,1,0) and d0:=(0,0,0) are so that reduce the number of calculation procedure.

Utilize

(MSB…LSB)

d2＝101≌1+x ²

d1＝010≌x

d0＝000≌0

Can be calculated as follows described two redundant symbol R1 and R2: $\mathrm{<figure-callout\; id="1"\; label="R"\; filenames="A02106540.300171.png,A02106540.300311.png"\; state="\{\{state\}\}">R</figure-callout>}1=\underset{i=0}{\overset{2}{\mathrm{\&Sigma;}}}{d}_i$ (symbol is made xor logic mutually and is connected) $\mathrm{<figure-callout\; id="2"\; label="R"\; filenames="A02106540.300311.png,A02106540.300331.png"\; state="\{\{state\}\}">R</figure-callout>}2=\left(x\underset{i=0}{\overset{2}{\mathrm{\&Sigma;}}}{x}^i{d}_i\right)\mathrm{mod}g\left(x\right)$ , g (x) :=1+x+x wherein ³

Described second formula of bit value substitution is correspondingly drawn:

x d0＝ 0000

x ²?d1＝ 010

x ³?d2＝ 101

R2＝x ⁵＝100000?mod(1+x+x ³)＝

＝1+1x+1x ²≌111

In view of the above, draw bit-patterns R1=111 or R2=111 at R1 and R2, thereby, the code word CW=(d0) that needs transmission (d1) (d2) (R1) (R2) be expressed as: CW:=000 010 101,111 111.The judgement of decoder is based on determines described two symptom S1 and S2, and they can be determined by following formula: $\mathrm{<figure-callout\; id="1"\; label="S"\; filenames="A02106540.300171.png,A02106540.300311.png"\; state="\{\{state\}\}">S</figure-callout>}1=R1+\underset{i=0}{\overset{2}{\mathrm{\&Sigma;}}}{\mathrm{<figure-callout\; id="2"\; label="d\; i\; S"\; filenames="A02106540.300311.png,A02106540.300331.png"\; state="\{\{state\}\}">d</figure-callout>}}_i$ $S$ $2=x(R2+x\underset{i=0}{\overset{2}{\mathrm{\&Sigma;}}}{x}^i{d}_i)\mathrm{mod}g\left(x\right).$

Under the situation of error free transmission, calculating symptom is S1=000 and S2=000.

If symbol d1=010 for example is received (error appears in bit middle and the right) with d1*=011, an error-pattern e=011 just occurs, calculates symptom S1 and S2 in this case:

S1=011 ≌ error-pattern e=011 ≌ 1+x or

S2＝x ³(1+x)mod(1+x+x ³)＝1+x ²≌101。

Draw bit-error locations L=001 (that is to say mistake is arranged) by combination S 1 and S2 and according to the arithmetic that [3] are told about in d1.This result is stored in network M1 as mapping table (having 6 inputs and 5 outputs).

4.1 length is 24 bits of encoded of the TFCI frame of 3-5 bit

4.1.1 code format:

For length is the TFCI data conditions of 3-5 bit, and producing length from the useful data piece is the code word of 24 bits, that is to say to data bit to insert 19 redundant bits other, that be used as data protection altogether.The characteristic of analyzing mobile radio channel draws following result, and promptly channel error demonstrates characteristic at random, but also can observe the mistake of a plurality of bunchys.Therefore, in order to utilize the property of described b signaling code, the symbol error more than two should not appear in a code word as much as possible.This condition is to satisfy at least approx by the number of symbols that reduces each code word.It is very favorable that following code division has shown to mobile radio channel:

As shown in Figure 7, the code word of 24 bit long is divided into three code subframes, and wherein, d1 and d2 are called data symbol.The useful data of expression TFCI is positioned at preceding two bits (the 3rd bit of d2 is changed to 0) of d1 and d2.

First code subframe SF1 forms by data symbol d1 with from b adjacent code redundant symbol R11, R12 that d1 produces.

Second code subframe SF2 forms (LSB of d2 is changed to 0 the Code And Decode process) by data symbol d2 with from b adjacent code redundant symbol R21, R22 that d2 produces.

Third generation numeral frame SF3 is made up of the b adjacent code redundant symbol R31, the R32 that produce from data symbol d1, d2.

All redundant symbols are to produce by encoder shown in Figure 5.

4.1.2 decoding agreement:

The code word that is received and 24 bit long of error may occur is divided into subframe SF1, SF2 and SF3, and these subframes are input to the input of decoder shown in Figure 6, wherein, moves decoding and error correction according to algorithm shown in Figure 8.In this flow chart, letter " K " expression signal " is successfully carried out error correction ".

Come to tell about briefly this flow chart below:

The agreement flow process:

1. subframe SF1 decodes:

2. subframe SF2 decodes:

Produce a signal (error correction is/not successfully execution) by corresponding decoder for each subframe.Type according to this message is further carried out following method:

3. if SF1=K and SF2=K: replace data symbol d1 and d2 among the SF3 with data symbol d1 among the SF1 and the data symbol d2 among the SF2.Proceed to LABEL1*.

4. SF1=K else if: replace data symbol d1 among the subframe SF3 with the data symbol d1 among the subframe SF1.Proceed to LABEL1.

5. SF2=K else if: replace data symbol d2 among the subframe SF3 with the data symbol d2 among the subframe SF1.Proceed to LABEL1.

* LABEL1: decoding SF3

6. if SF3=K: code word is correct

7. otherwise (SF3=D): code word is made mistakes

Therefore, utilize this method can proofread and correct up to two symbol errors or up to 6 bits.

1.1. length is 48 bits of encoded of the TFCI frame of 6-10 bit:

1.2. 4.2.1 code format:

If described TFCI data are made up of 6-10 bit, then from the useful data piece, produce the code word that length is 48 bits, that is to say to this data bit and insert 38 redundant bits other, that be used as data protection altogether.Following code division is proved to be favourable:

Code word is divided into 5 subframe SF1-SF5 (referring to Fig. 9), and wherein d1, d2, d3 and d4 are also referred to as data symbol.

First code subframe SF1 is made up of data symbol d1 and b adjacent code redundant symbol R11, R12.

Second code subframe SF2 is made up of data symbol d2 and b adjacent code redundant symbol R21, R22.

Third generation numeral frame SF3 is made up of data symbol d1, d2 and b adjacent code redundant symbol R31, R32.

SF1, SF2 and SF3 produce according to 24 bits of encoded methods.

The 4th code subframe SF4 is made up of data symbol d3 and redundant symbol R41, R42, R43, R44 and R45.It is the duplicated code of R41=R42=R43=R44=R45=d3 that all symbols of this subframe have constituted form.Therefore, data symbol d3 repeats altogether in SF4 6 times.

The 5th code subframe SF5 is made up of data symbol d4 and redundant symbol R51.Also set R51=d4 at this, make the 10th TFCI bit in this subframe, repeat altogether 6 times.

4.2.2 decoding agreement:

The code word that is received and 48 bit long of error may occur is divided into 5 subframe SF1, SF2, SF3, SF4 and SF5, wherein, also can be respectively by decoder shown in Figure 6 decode 3 adjacent code subframe SF1, SF2 and SF3.

The decoding agreement (referring to Figure 10) that is used for 48 TFCI bits is made up of two parts basically.At this, the protocol section that relates to data symbol d1 and d2 is corresponding to the decode flow chart of 24 TFCI bits of being used to of being told about in chapters and sections 4.1.2.Second protocol section that relates to data symbol d3 and d4 is based on most resolution principles (referring to following).The decoding of being carried out is as follows in detail:

Protocol section 1:

1. subframe SF1 decodes:

2. subframe SF2 decodes:

Produce a signal (error correction is/not successfully execution) by corresponding decoder for each subframe.Type according to this message is further carried out following method:

3. if SF1=K and SF2=K: replace data symbol d1 and d2 among the SF3 with data symbol d1 among the SF1 and the data symbol d2 among the SF2.Proceed to LABEL1*.

4. SF1=K else if: replace data symbol d1 among the subframe SF3 with the data symbol d1 among the subframe SF1.Proceed to LABEL1.

5. SF2=K else if: replace data symbol d2 among the subframe SF3 with the data symbol d2 among the subframe SF1.Proceed to LABEL1.

* LABEL1: decoding SF3

6. if SF3=K: code word is correct

7. otherwise (SF3=D): code word is made mistakes

Protocol section 2:

The 1st step: SF4 is according to most resolution principle decoding symbols d3.

The 2nd step: SF5 is according to most resolution principle decoding symbols d4 (only estimating first bit of d4 herein) with the form of useful data bit.

If protocol section 1 and protocol section 2 successfully finish, then evaluate entire frame d1, d2, d3 and d4 for correct.

This scheme allows to proofread and correct altogether up to 4 symbol errors, or under certain situation, as long as mistake only appears in the data bit, then allows to proofread and correct all 10 data bits.

The elaboration of most resolution principles

This principle is based on the repetition of data bit.In this case, three bits of described data symbol d3 or the 10th TFCI bit among the d4 are repeated respectively 5 times, so that draw the code word that whole length is 6 bits from a data bit.In addition, data are not transmitted in this case with encoding.Receiver side to code word in 1 quantity count.If amount of bits is an even number, then provide following decoding to judge:

1. as long as 1 and 0 number is inequality, then do to proofread and correct to judge.

1 with situation that 0 number equates under, also carry out fault identification in addition.

, this introduces following embodiment for being illustrated:

In emitting side:

Data bit is code word=000000 that 0  produces

Data bit is code word=111111 that 1  produces

At receiver side:

Situation 1: the code word of reception is: the judgement of 111100  decoders helps bit value " 1 ".

Situation 2: receiving code word is: 111000  decoders produce error signal.

4.2 b (=3) adjacent code that channel is optimized:

Different with above-mentioned 3 adjacent codes is that the channel of telling about is below optimized code C1 and C2 has used special generator matrix P1 and P2, distributes to data symbol d with generation _i(i=1,2) or two data symbol d ₁/ d ₂Redundant symbol R _iOr R _IjFigure 11 illustrates the principle of 3 adjacent codes of described mobile channel optimization, wherein, (6 * 3) generator matrix P1 of expression code C1 has form as shown in figure 12, and (6 * 6) generator matrix P2 of expression code C2 has form as shown in figure 13.At this, so select described generator matrix P1 and P2, make the single b symbol error of described code C1 in mainly can the correction data symbol, and code C2 can proofread and correct the described redundant symbol R that derives from two data symbols _IjThe two mistakes of the residue of the maximum quantity in (subframe 3 vide infra).

At this, the TFCI data block that comprises 3-5 bit also is mapped in the code word that length is 24 bits (referring to Figure 14).At this, code C1 had both produced the described subframe SF1 that is made up of data symbol d1 and redundant symbol R11, R22, also produce the described subframe SF2 that is made up of data symbol d2 and redundant symbol R21, R22, code C2 then produces described two redundant symbol R31 and the R32 that is assigned to subframe SF3 from two data symbol d1 and d2.Correspondingly, when shining upon the TFCI data block of forming by 6-10 bit, the code word that length is 48 bits (referring to Figure 15) will be processed into.Subframe SF4 and SF5 relate to above-mentioned duplicated code, also promptly set d3=R41=...=R45 and d4=R51.In Figure 16 and 17, show once more not only be used for length be the TFCI data block of 24 bits but also be used for length be 48 bits the TFCI data block, to or distribute and produce redundant symbol R from data symbol di _Ij

4.2.1 encoder structure:

Figure 18 illustrates a kind of possible hardware configuration, be used to realize that length is the encoder of the TFCI data block of 3-5 bit.It mainly comprises: distribute to unit each data symbol di and redundant symbol Rij, that always store 3 bits and correspondingly indicate with P1 and P2; The XOR gate circuit of representing corresponding generator matrix; And the continuous selection circuit (demultiplexer) of the outlet side of AND circuit P1.The decoder as shown in figure 19 that is positioned at receiver side has a kind of simpler structure.It is made up of three subelements (along separate routes) of distributing to each subframe SF1/SF2/SF3 altogether, and wherein each subelement all has decoder DEC1 or DEC2,6 * 1 symptom memory S1, a S2 or S3 and shown in Figure 20 or 21 symptom error list 1 shown in Figure 22 and 23 or 2 64 * 4 or 64 * 8 memory matrix are housed.Interior, described be used for the encoder of TFCI data block that length is the 6-10 bit or the structure of decoder is corresponding together with five times of duplicated codes of data symbol d3 and d4.Be used for length and be 3-5 or 6-10 bit the TFCI data block the decoding agreement with above be identical (referring to chapters and sections 4.1.2 and 4.2.2).

5. prior art

[1] W.Adi, H.G.Kolloge, Double error miscorrectionevaluation for a byte error correction scheme (mistake that is used for two mistakes of byte correcting scheme is proofreaied and correct assessment), (1983) the 11st of ntz-Archiv the 5th volumes, 305-311 page or leaf

[2] people such as Hodges, Method and apparatus for generating errorlocating and parity check bytes (being used to produce the method and apparatus of error location and parity byte), european patent application 79101717.1,1979, the 1-38 page or leaf (:=US 4,185,269)

6. the abbreviation of Cai Yonging

The CDMA code division multiple access

FDD Frequency Division Duplexing (FDD) GP guard time kbps kilobits per second Mbps MBPS Mcps million chip per second PSK phase shift keying QPSK four phase PSKRM Reeds-Miu Le TDD time division duplex TDMA time division multiple access TD-SCDMA time-division synchronization CDMA TFCI transport format combination indicator UMTS Universal Mobile Telecommunications System

Claims (10)

1. by carrying out the method for following steps transmitting bit sequence through radio channel:

A) bit sequence that respectively need is transmitted at the emission pusher side is grouped into the data symbol (d of b bit long _i),

B) by the systematic code described data symbol (d that encodes _i), wherein, the code word that is produced is up to 2 ^b-1 data symbol (d _i), and give described each data symbol (d _i) to distribute at least one length be the redundant symbol (R of b bit _Ij),

C) utilize described code word by b ratio specially modulated carrier signal, and

D) by the described carrier signal of demodulation and the signal that received of decoding in the described data of receiver side reconstruct.

2. the method for claim 1 is characterized in that:

Produce by the b adjacent code and to distribute to described data symbol (d _i) single or all redundant symbol (R _Ij).

3. method as claimed in claim 1 or 2 is characterized in that:

Described carrier signal is carried out numeral 2 ^bHeavy phase modulation.

4. as the described method of one of claim 1～3, it is characterized in that:

Distribute one by at least two symbol (d for described bit-patterns _i) group of data symbols formed; And give described data symbol (d _i) distribute one by described data symbol (d _i) and at least three redundant symbol group (R _Ij) code word that constitutes, wherein, the first redundant symbol group (R11, R12) produce from the first data symbol subgroup (d1) by first systematic code (C1), (R21 R22) produces from the second data symbol subgroup (d2) by first systematic code (C1) the second redundant symbol group, and triple redundance symbols (R31, R32) by data symbol (d1, d2) middle produce of second systematic code (C2) from first and second subgroups.

5. method as claimed in claim 4 is characterized in that:

Distribute one by at least four symbol (d for described bit-patterns _i) group of data symbols formed; And described code word has the 4th and the 5th redundant symbol group, wherein, the 4th redundant symbol group (R41, R45) produce from the 3rd data symbol subgroup (d3) by first duplicated code, and the 5th redundant symbol group (R45) produces from the 4th data symbol subgroup (d4) by second duplicated code.

6. as claim 4 or 5 described methods, it is characterized in that:

Described data symbol subgroup is that element is unusual, and always only comprises a data symbol.

7. as the described method of one of claim 1～6, it is characterized in that:

Described code word is separated into the subgroup at receiver side, and wherein, the subgroup of the code word that is received is by at least one data symbol (d _i) and the redundant symbol of distributing to this reception data symbol constitute.

8. method as claimed in claim 7 is characterized in that:

Decode in the subgroup of the reception code word that is constituted with being separated.

9. method as claimed in claim 4 is characterized in that:

Described bit-patterns comprises 3-5 bit, and described code word is made up of two data symbols and six redundant symbols, and wherein, described data symbol and redundant symbol all have the length of 3 bits.

10. as claim 4 or 5 described methods, it is characterized in that:

Described bit-patterns comprises 6-10 bit, and described code word is made up of four data symbols and 12 redundant symbols, and wherein, described data symbol and redundant symbol all have the length of 3 bits.

Images