CN102394722A - Soft decoding method of visual block product turbo code (VBPTC) in data management routine (DMR) under 4-frequency shift key (4FSK) modulation mode - Google Patents

Abstract

The invention discloses a soft decoding method of a visual block product turbo code (VBPTC) in a data management routine (DMR) under a 4-frequency shift key (4FSK) modulation mode. The soft decoding method comprises the following steps of: S1, calculating external information w of a parity check bit; S2, correcting quantified soft bit information by using the calculated external information of the parity check bit; and S3, determining a symbol. Compared with the prior art, the soft decoding method of the invention has the biggest advantage of increasing the performance of an error code, so the error detection function can be fully realized.

Description

A kind of soft decoding method of VBPTC sign indicating number among the DMR under the 4FSK modulation system

Technical field

The present invention relates to the communications field of wireless communication system, design a kind of soft decoding method of VBPTC sign indicating number under the 4FSK modulation system among the DMR especially.

Background technology

Product code is that Elias proposed in 1954, and it is that first can realize not having the error correction coding of error code transmission after the Shannon information theory proposed when the non-zero code check.VBPTC sign indicating number among the DMR is actually a kind of Turbo product code, and its component code is a Hamming code, has comprised row or row parity check.Three kinds of logical channel type: SLC_CH, RC_CH, EMS_CH are arranged, wherein:

The capable subcode of SLC_CH is used Hamming (17,12,3), and the row subcode is used parity check, is designated as SLC (17,12) (4,3).

The capable subcode of RC_CH is used expansion Hamming (16,11,4), and the row subcode is used parity check, is designated as RC (16,11) (2,1).

The capable subcode of EMS_CH is used expansion Hamming (16,11,4), and the row subcode is used parity check, is designated as EMS (16,11) (8,7).

People such as Pyndiah in 1994 have proposed a kind of soft input/soft output decoding algorithm of linear block codes on the basis of Chase decoding algorithm, be applied to the decoding of product code subcode, and the decoding of this and Turbo code is similar, is called Turbo product code (TPC).Its component code mainly contains Hamming code, parity check code etc.VBPTC sign indicating number among the DMR promptly is that a kind of component code is a Hamming code, comprises the Turbo product code of row or row parity check bit simultaneously.Can adopt the decoding algorithm of soft input/soft output that it is deciphered.

Introduce following knowledge below about log-likelihood ratio.If the transmission bit is d, the symbol mapped relation is: 0 →-1,1 → 1, and its log-likelihood ratio does

$L\left(d\right)=\ln \left(\frac{P(d=+1)}{P(d=-1)}\right)---\left(1\right)$

For stable state independent random variable d ₁, d ₂, then have

$P({\mathrm{<figure-callout\; id="1"\; label="d"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">d</figure-callout>}}_1\&CirclePlus;{\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HSA00000600856000011.png,HSA00000600856000021.png"\; state="\{\{state\}\}">d</figure-callout>}}_2=+1)=P({\mathrm{<figure-callout\; id="1"\; label="d"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">d</figure-callout>}}_1=+1)\&CenterDot;(1-P({\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HSA00000600856000011.png,HSA00000600856000021.png"\; state="\{\{state\}\}">d</figure-callout>}}_2=+1))$

$+P({\mathrm{<figure-callout\; id="2"\; label="d"\; filenames="HSA00000600856000011.png,HSA00000600856000021.png"\; state="\{\{state\}\}">d</figure-callout>}}_2=+1)\&CenterDot;(1-P({\mathrm{<figure-callout\; id="1"\; label="d"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">d</figure-callout>}}_1=+1))$

(2)

Can get by formula (1)

$P(d_i=+1)=\frac{e^{L\left(d_i\right)}}{1+e^{L\left(d_i\right)}},i=1,2---\left(3\right)$

Obtain being reduced to of corresponding log-likelihood ratio by formula (2) and (3)

$L({\mathrm{<figure-callout\; id="1"\; label="d"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">d</figure-callout>}}_1\&CirclePlus;d_2)=\ln \frac{e^{L\left(d_1\right)}+e^{L\left(d_2\right)}}{1+e^{L\left(d_1\right)}\&CenterDot;e^{L\left(d_2\right)}}$

$\&ap;-\mathrm{sign}\left(L\left(d_1\right)\right)\&CenterDot;\mathrm{sign}\left(L\left(d_2\right)\right)\&CenterDot;\min \left(\right|L\left(d_1\right)|,|L\left(d_2\right)\left|\right)$

(4)

If expand to m stochastic variable, then have

The difference of each yard in soft input soft output decode is external information L just _e(d) calculating.

Summary of the invention

In order to overcome the defective of prior art, the present invention proposes a kind of soft decoding method of VBPTC sign indicating number under the 4FSK modulation system among the DMR, its maximum advantage just is to improve error performance, thereby can better bring into play its error detection function.

Technical scheme of the present invention is following:

A kind of soft decoding method of VBPTC sign indicating number among the DMR under the 4FSK modulation system may further comprise the steps:

S1: the external information w of computation of parity bits;

S2: the external information with the parity check bit of aforementioned calculation is proofreaied and correct the soft bit information after quantizing;

S3: carry out symbol judgement.

Preferably, said step S1 further comprises:

For parity check code, the bit of establishing in the code word is u _j(j=1 ..., m), code length is m, satisfies following relation: Then arbitrary uj can be expressed as

Can obtain its external information does $L_e\left(u_j\right)=L(\underset{i=1,i\&NotEqual;j}{\overset{m}{\mathrm{\&Sigma;}}}\&CirclePlus;u_i).$

Preferably, said step S2 further comprises:

It is received soft bit r proofread and correct, r is arranged ₁=r+w;

Data will be encoded before transmission, after reception, will decipher, and carrying out coded data all is 01 this bit sequence; But resulting data all are some decimals in the time of decoding, and we are referred to as soft bit, at first need revise these soft bits; Adjudicate according to certain rule then, obtain 01 such bit sequence at last, r is exactly received soft bit in the literary composition; With external information it is revised then, revised result is r1.

Preferably, said step S3 further comprises:

Carrying out hard decision does $d=\left\{\begin{array}{cc}1& {\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">r</figure-callout>}}_1\&GreaterEqual;0\\ 0& {\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">r</figure-callout>}}_1<0\end{array}\right.;$

Data will be encoded before transmission, after reception, will decipher, and carrying out coded data all is 01 this bit sequence; But resulting data all are some decimals in the time of decoding, and we are referred to as soft bit, at first need revise these soft bits; Adjudicate according to certain rule then; Obtain 01 such bit sequence at last, r is exactly received soft bit in the literary composition, with external information it is revised then; Revised result is r1, and the d that finally obtains is exactly 01 this bit sequence.

Preferably, said method has provided the external information computational methods that parity check bit quantizes.

Preferably, the capable subcode of said VBPTC sign indicating number is a Hamming code.

Preferably, said method adopts the decoding algorithm of soft inputting and soft output, and the check digit external information in this algorithm is to try to achieve through the algorithm of parity check bit.

Preferably, said method is to receiving the level quantification that soft bit information carries out 8bit, and whole decoding all is that integer is handled.

Compared with prior art, the advantage of maximum of the present invention just is to improve error performance, thereby can better bring into play its error detection function.

Description of drawings

Fig. 1 logic channel SLC_CH parity check bit is soft, the hard decision error performance compares;

Fig. 2 logic channel SLC_CH is soft, the hard decision error performance compares;

Fig. 3 logic channel SLC_CH is soft, hard decision mistake letter rate performance compares;

Fig. 4 logic channel RC_CH is soft, the hard decision error performance compares;

Fig. 5 logic channel RC_CH is soft, hard decision mistake letter rate performance compares;

Fig. 6 logic channel EMS_CH is soft, the hard decision error performance compares;

Fig. 7 logic channel EMS_CH is soft, hard decision mistake letter rate performance compares.

Embodiment

The below combines accompanying drawing and specific embodiment that the present invention is done further description.

Embodiment

A kind of soft decoding method of VBPTC sign indicating number among the DMR under the 4FSK modulation system may further comprise the steps:

S1: the external information of computation of parity bits;

S2: the external information with the parity check bit of aforementioned calculation is proofreaied and correct the soft bit information after quantizing.

S3: carry out symbol judgement.

In the present embodiment, said step S1 further comprises:

For parity check code, the bit of establishing in the code word is u _j(j=1 ..., m), code length is m, satisfies following relation:

Then arbitrary u _jCan be expressed as Can obtain its external information does $L_e\left(u_j\right)=L(\underset{i=1,i\&NotEqual;j}{\overset{m}{\mathrm{\&Sigma;}}}\&CirclePlus;u_i).$

Here just takeed an example, for the correlation of external information is described.Code length is m, representes that this sequence has m bit, available u1, and u2 ... um representes that their value only possibly be 0 or 1.Each bit all has its external information, so just to m external information should be arranged.Since u1, u2 ... satisfy certain relation between the um, so also there is certain relation between their external information.

Like this, the external information of current bit is uncorrelated with the log-likelihood ratio of itself.Therefore in the VBPTC sign indicating number, obtain the external information of Hamming code after, and then can try to achieve the external information w of parity check bit.

In the present embodiment, said step S2 further comprises:

It is received soft bit r proofread and correct, r is arranged ₁=r+w;

In the present embodiment, said step S3 further comprises:

Carrying out hard decision does $d=\left\{\begin{array}{cc}1& {\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">r</figure-callout>}}_1\&GreaterEqual;0\\ 0& {\mathrm{<figure-callout\; id="1"\; label="r"\; filenames="HSA00000600856000011.png"\; state="\{\{state\}\}">r</figure-callout>}}_1<0\end{array}\right.;$

For step S2 and involved r and the r of step S3 ₁Data will be encoded before transmission, after reception, will decipher, and carrying out coded data all is 01 this bit sequence; But resulting data all are some decimals in the time of decoding, and we are referred to as soft bit, at first need revise these soft bits; Adjudicate according to certain rule then, obtain 01 such bit sequence at last, r is exactly received soft bit in the literary composition; With external information it is revised then, revised result is r ₁, at last according to r ₁Positive negativity carry out symbol judgement, the d that finally obtains is exactly 01 this bit sequence.

The u of front _jBe hard bit, its value possibly be 0 or 1 only, and r is soft bit, and its value is indefinite, can be any number (comprising decimal, negative).

In this soft inputting and soft output soft decoding process, to receiving the level quantification that soft bit information carries out 8bit, span is [127,128].Therefore whole decoding all is that integer is handled, and greatly reduces operand, but does not almost have performance loss.

Because soft bit information all is some random data, all is decimal basically, the operand between the data is just very big like this.The principle that quantizes is to change into integer to these decimals through certain mapping, thereby reduces operand.The meaning that 8bit quantizes is exactly to represent data with 01 combination of 8 bit.Combined result has 256 kinds of possibilities, with the 1st 01 represent positive and negative, back 7 bit representation numerical value, obviously span is [127,128].Make up with 8 bit and to represent level value, precision is enough.

The method of present embodiment has provided the external information computational methods that parity check bit quantizes, and the capable subcode of stating the VBPTC sign indicating number that it is related is a Hamming code.And said method adopts the decoding algorithm of soft inputting and soft output, and the check digit external information in this algorithm is to try to achieve through the algorithm of parity check bit.And said method is to receiving the level quantification that soft bit information carries out 8bit, and whole decoding all is that integer is handled.

Be an emulation of carrying out with above-mentioned algorithm below,, under the white Gaussian noise channel, use the soft inputting and soft output algorithm that three kinds of logic channels are carried out soft decoding in the 4FSK modulation.Error performance before and after the above algorithm compares Fig. 1 for parity check bit uses, and can find out that performance is significantly increased.Soft, hard error rate of translation, mistake letter rate performance that Fig. 2, Fig. 3 are respectively SLC_CH compare; Soft, hard error rate of translation, mistake letter rate performance that Fig. 4, Fig. 5 are respectively RC_CH compare, and soft, hard error rate of translation, mistake letter rate performance that Fig. 6, Fig. 7 are respectively EMS_CH compare.

In Fig. 1, Fig. 2, Fig. 4 and Fig. 6, abscissa is represented signal to noise ratio, and ordinate is represented the error rate.Usually all use this two qualities of describing a kind of coded system.Popular says, is exactly that abscissa is used for describing background, and ordinate is used for describing interference free performance.The value of abscissa is turned right more, just representes that signal is strong more in the background, and noise is more little, and the value of ordinate is big more, just representes that the error rate is big more, and we hope that the value of ordinate is more little good more.The interval of abscissa, it is strong to very from very poor (almost not having) to have contained signal strength signal intensity, and ordinate is that corresponding calculating comes according to abscissa.

Among Fig. 3, Fig. 5 and Fig. 7, abscissa is a signal to noise ratio, i.e. the ratio of information and noise, reflection be the power of ambient signal.Ordinate is represented mistake letter rate, and mistake letter rate just is meant because the influence of channel noise causes signal in the process of transmission, can make a mistake.Under identical signal to noise ratio, mistake letter rate is more little good more.

Be relatively each time complexity soft, hard decoding, table 1 has provided clock cycle (clock) and the operation time (ms) of its c code under CCSStudio v3.3 environment, and frequency is 48MHz.It has shown that logic channel SLC_CH, RC_CH, EMS_CH compile, the time complexity of decoding.

Table 1

Be not difficult to find that compare with corresponding hard decision, the method improvement in performance of present embodiment is remarkable.

The preferred embodiment of the present invention just is used for helping to set forth the present invention.Preferred embodiment does not have all details of detailed descriptionthe, does not limit this invention yet and is merely described embodiment.Obviously, according to the content of this specification, can do a lot of modifications and variation.These embodiment are chosen and specifically described to this specification, is in order to explain principle of the present invention and practical application better, thereby person skilled can be utilized the present invention well under making.The present invention only receives the restriction of claims and four corner and equivalent.

Claims (8)

1.DMR in a kind of soft decoding method of VBPTC sign indicating number under the 4FSK modulation system, it is characterized in that, may further comprise the steps:

S1: the external information w of computation of parity bits;

S2: the external information with the parity check bit of aforementioned calculation is proofreaied and correct the soft bit information after quantizing;

S3: carry out symbol judgement.

2. a kind of soft decoding method of the VBPTC sign indicating number among the DMR according to claim 1 under the 4FSK modulation system is characterized in that said step S1 further comprises:

For parity check code, the bit of establishing in the code word is u _j(j=1 ..., m), code length is m, satisfies following relation: Then arbitrary u _jCan be expressed as

Can obtain its external information does $L_e\left(u_j\right)=L(\underset{i=1,i\&NotEqual;j}{\overset{m}{\mathrm{\&Sigma;}}}\&CirclePlus;u_i).$

3. a kind of soft decoding method of the VBPTC sign indicating number among the DMR according to claim 1 under the 4FSK modulation system is characterized in that said step S2 further comprises:

It is received soft bit r proofread and correct, r is arranged ₁=r+w;

Data will be encoded before transmission, after reception, will decipher, and carrying out coded data all is 01 this bit sequence; But resulting data all are some decimals in the time of decoding, and we are referred to as soft bit, at first need revise these soft bits; Adjudicate according to certain rule then, obtain 01 such bit sequence at last, r is exactly received soft bit in the literary composition; With external information it is revised then, revised result is r1.

4. a kind of soft decoding method of the VBPTC sign indicating number among the DMR according to claim 1 under the 4FSK modulation system is characterized in that said step S3 further comprises:

Carrying out hard decision does $d=\left\{\begin{array}{cc}1& r_1\&GreaterEqual;0\\ 0& r_1<0\end{array}\right.;$

Data will be encoded before transmission, after reception, will decipher, and carrying out coded data all is 01 this bit sequence; But resulting data all are some decimals in the time of decoding, and we are referred to as soft bit, at first need revise these soft bits; Adjudicate according to certain rule then; Obtain 01 such bit sequence at last, r is exactly received soft bit in the literary composition, with external information it is revised then; Revised result is r1, and the d that finally obtains is exactly 01 this bit sequence.

5. a kind of soft decoding method of the VBPTC sign indicating number among the DMR according to claim 1 under the 4FSK modulation system is characterized in that, said method has provided the external information computational methods that parity check bit quantizes.

6. according to a kind of soft decoding algorithm of the VBPTC sign indicating number among the said DMR of claim 1 under the 4FSK modulation system, it is characterized in that the capable subcode of said VBPTC sign indicating number is a Hamming code.

7. according to a kind of soft decoding algorithm of the VBPTC sign indicating number among the said DMR of claim 1 under the 4FSK modulation system; It is characterized in that; Said method adopts the decoding algorithm of soft inputting and soft output, and the check digit external information in this algorithm is to try to achieve through the algorithm of parity check bit.

8. a kind of soft decoding method of the VBPTC sign indicating number among the DMR according to claim 1 under the 4FSK modulation system is characterized in that, said method is to receiving the level quantification that soft bit information carries out 8bit, and whole decoding all is that integer is handled.

Images