CN101771644B - Joint detection and soft decision decoding-based signal receiving method - Google Patents

Abstract

The invention provides a joint detection and soft decision decoding-based signal receiving method. The method comprises the following steps: a, jointly detecting a received signal; b, calculating reliable information of the received signal based on the result of the joint detection; and c, carrying out soft decision decoding on the received signal based on the reliable information. The method effectively combines the joint detection technology and the soft decision decoding technology, and can further improve the performance of a communication system.

Description

A kind of signal acceptance method based on joint-detection and Soft decision decoding

Technical field

The present invention relates to wireless communication field, relate in particular to the signal acceptance method in a kind of wireless communication system based on joint-detection and Soft decision decoding.

Background technology

Code division multiple access (CDMA) is a kind of spread spectrum cut-in method.Its principle is: the information data with certain signal bandwidth that will transmit, modulate much larger than the high speed pseudo noise code of signal bandwidth with a bandwidth, and the bandwidth of former data-signal is expanded, again through carrier modulation and send.Receiving terminal uses identical pseudo noise code, makes relevant treatment with the bandwidth signal that receives, and the narrow band signal that converts broadband signal to the prime information data is despreading, to realize information communication.

OFDM (OFDMA) be a kind of with OFDM be the basis the multicarrier cut-in method.Its principle is that frequency spectrum is divided into many orthogonal sub-carriers, chooses a part of subcarrier and forms subchannel, and utilize OFDM to transmit.Because each subcarrier is a parallel transmission, the interference between the erasure signal waveform greatly.OFDM can also adopt sign indicating number expansion technology, and the frequency domain data that will send is assigned on each number of sub-carrier of a subchannel after expanding through sign indicating number.

Above-mentioned ofdm system and cdma system all relate to the problem of a joint-detection.In addition, Soft decision decoding is a kind of interpretation method with respect to Hard decision decoding, Soft decision decoding coding gain higher 2 decibels than Hard decision decoding under additive white Gaussian noise channel.As if both ability Combined application, then can further improve the performance of communication system.The present invention proposes a kind of communication means that combines joint-detection and Soft decision decoding.

Summary of the invention

The present invention proposes a kind of signal acceptance method based on joint-detection and Soft decision decoding, this method comprises:

A. carry out joint-detection to received signal;

B. based on the result of joint-detection, calculate the authentic communication that receives signal;

C. carry out Soft decision decoding to received signal based on said authentic communication.

In this method, said received signal can be the CDMA signal, uses channel estimating and spreading code to carry out joint-detection to received signal.Said reception signal also can be ofdm signal, uses channel estimating, carries out joint-detection to received signal.Said received signal can also be the ofdm signal that expands through sign indicating number, uses channel estimating and sign indicating number to expand the aggregate channel matrix that matrix is formed, and carries out joint-detection to received signal.Above-mentioned said channel estimating can be to utilize pilot tone or training sequence to estimate channel, and carries out the channel match, and said channel match can be LPF.

Preferably, among the step b, the authentic communication that calculate to receive each bit of signal be the likelihood ratio of each bit, and the absolute value of employing authentic communication weighs the reliability of each bit, and said absolute value is more little, and the reliability of said bit is low more; Among the step c; The minimum bit of several reliabilitys of a code element to received signal makes up; Again the code element of each combination is carried out Hard decision decoding; And the Euclidean distance of the authentic communication of calculating Hard decision decoding result and said code element, choose the minimum grouping of bits of Euclidean distance as decoding output.

Preferably, among the step b, the authentic communication that calculate to receive each symbol of signal be and this symbol time nearest Euclidean distance, and the employing authentic communication weighs the reliability of each symbol, and the value of said authentic communication is more little, and the reliability of said symbol is low more; Among the step c; The minimum symbol of several reliabilitys of a code element to received signal makes up; Again the code element of each combination is carried out Hard decision decoding; The result is modulated into symbol sebolic addressing with Hard decision decoding, calculates said symbol sebolic addressing and the Euclidean distance that receives signal, chooses the corresponding grouping of bits of the minimum symbol sebolic addressing of Euclidean distance as decoding output.

The present invention has effectively combined joint-detection and two kinds of technology of Soft decision decoding, can further improve the performance of communication system.Especially for ofdm system, at first adopt sign indicating number to expand and to obtain frequency diversity, effectively reduce the decline degree of channel; Adopt channel fitting technique and Soft decision decoding can further improve the ofdm system performance then.

Description of drawings

Fig. 1 is the structural representation of the ofdm system of the embodiment of the invention;

Fig. 2 is the constellation sketch map that after chnnel coding, carries out the 16QAM modulation of the embodiment of the invention;

Fig. 3 is the sketch map of an OFDM subchannel of the embodiment of the invention.

Embodiment

Specify practical implementation method of the present invention below in conjunction with accompanying drawing.

With an ofdm system as shown in Figure 1 is example, at transmitting terminal, after information source process source encoding and the chnnel coding, goes out through the OFDM modulate emission; At receiving terminal, through carrying out joint-detection behind the OFDM demodulated received signal, calculate the authentic communication that receives signal based on the result of joint-detection, carry out Soft decision decoding to received signal according to said authentic communication then, carry out source decoding at last.The signal that specifies present embodiment below transmits and receives the committed step in the process.

1, chnnel coding

Present embodiment adopts (26,24) RS sign indicating number to do channel code, and it is by GF (2 ⁵) (31,29) sign indicating number brachymemma gained on the territory.Contain 96 bits in each encoding block.If the bit in each encoding block is (b ₀, b ₁..., b ₉₅), according to 29 symbol (S of coded-bit structure ₀, S ₁..., S ₂₈), each symbol is made up of 5 bits, and method is following: S ₀=(0b ₀b ₁b ₂b ₃), S ₁=(0b ₄b ₅b ₆b ₇) ..., S ₂₃=(0b ₉₂b ₉₃b ₉₄b ₉₅), S ₂₄=(00000), S ₂₅=(00000), S ₂₆=(00000), S ₂₇=(00000), S ₂₈=(00000); Add 2 check codes, 31 * 5=155 bit altogether after the coding, S ₀, S ₁..., S ₂₃In the 1st bit do not launch S ₂₄, S ₂₅..., S ₂₈Do not launch bit one total 155-24-5 * 5=106 bit of then emission after the coding.

2, OFDM modulate emission

Binary message after the chnnel coding at first through 16QAM modulation as shown in Figure 2, is carried out the OFDM modulation with the data after the 16QAM modulation again.

The ofdm system of present embodiment adopts 128 number of sub-carrier, shares the bandwidth of 1M.Per 8 number of sub-carrier are formed a subchannel, and the i subchannel is used i, i+16, i+32, i+48, i+64, i+80, i+96, i+112 work song carrier wave.Be illustrated in figure 3 as the sketch map of a subchannel, each subchannel is one 8 * 8 a time-frequency two-dimensional grid.Adopt sign indicating number to expand technology to each subchannel, the frequency domain data that promptly will send is assigned on each number of sub-carrier of a subchannel after expanding through sign indicating number.

Sign indicating number expansion technology is specially: expand conversion x=W s through sign indicating number, l sign reversing become N _sIndividual data.Wherein, s is the symbol before sign indicating number expands, and x is the data after sign indicating number expands, and W is big or small N _sThe sign indicating number of * l expands matrix.In the present embodiment, the spread spectrum matrix adopts W ₈Preceding l row, l data are expanded to 8 data.

$W_8=\frac{1}{\sqrt{8}}\left(\begin{array}{cccccccc}1& 1& 1& 1& 1& 1& 1& 1\\ 1& -1& 1& -1& 1& -1& 1& -1\\ 1& 1& -1& -1& 1& 1& -1& -1\\ 1& -1& -1& 1& 1& -1& -1& 1\\ 1& 1& 1& 1& -1& -1& -1& -1\\ 1& -1& 1& -1& -1& 1& -1& 1\\ 1& 1& -1& -1& -1& -1& 1& 1\\ 1& -1& -1& 1& -1& 1& 1& -1\end{array}\right)$

Sign indicating number is expanded on the subchannel of later data map to 8 * 8, and adopt the pilot symbol as channel estimating.At last with the time data after the frequency map through the IFFT conversion, add Cyclic Prefix, up-conversion is launched.

3, OFDM demodulate reception

Received signal is accomplished after the reception synchronously by down coversion, each OFDM symbol is carried out the FFT shift conversion become frequency domain data.

4, joint-detection

In order to make joint-detection more accurate, at first subchannel is carried out channel estimating, and the subchannel of estimating is carried out match in frequency domain.Said channel fitting technique can for: earlier channel estimating is carried out interpolation in frequency domain, the subcarrier that not have to utilize carries out LPF with all subcarrier channel estimation then through its channel estimation value of channel interpolation acquisition.Adopt length to carry out filtering for the low pass window function of L, the preceding edge effect for fear of filter of filtering can be with channel estimating to each equivalent continuation L-1 point length of height/low frequency.In the present embodiment, all already used subcarriers are carried out the LS channel estimating, it is the channel of 128 number of sub-carrier that the channel after estimating is inserted earlier; With 6 points of each equivalent continuation about 128 channels, promptly 140 some channel responses use [1 at last then; 1,1,1; 1,1,1] weights/7 carry out filtering to 140 point data.

Next step, the channel estimating that from filtered channel estimating, extracts needed subcarrier is carried out the joint-detection computing.If in the present embodiment after the match of certain each subcarrier of subchannel channel estimating is respectively

then receives signal and be: y=Hs+n; H is that the channel estimating after the match expands the aggregate channel matrix that matrix is formed with sign indicating number; Wherein, Y is that frequency domain receives signal, and H calculates as follows:

Method through joint-detection is recovered S, and joint-detection adopts MMSE to detect:

r＝(H ^*H+σ ²I) ^-1H ^*y

Data r after then MMSE being detected carries out the 16QAM demodulation.

5, calculate authentic communication

Based on the result of joint-detection, calculate the authentic communication that receives signal.In the present embodiment, the authentic communication of calculating each bit that receives signal is the likelihood ratio of this bit, is:

$L=\log \frac{\underset{a_i\∈{\mathrm{\Ω}}_0}{\mathrm{\Σ}}\exp (-\frac{{(r-a_i)}^2}{2{\mathrm{\σ}}^2})}{\underset{a_i\∈{\mathrm{\Ω}}_1}{\mathrm{\Σ}}\exp (-\frac{{(r-a_i)}^2}{2{\mathrm{\σ}}^2})}$

Ω wherein ₀Be that the corresponding bit position is the set that 0 constellation point is formed, Ω ₁Be that the corresponding bit position is the set that 1 constellation point is formed, σ ²Be noise power.

The absolute value of employing likelihood ratio is weighed the reliability of this bit, and the absolute value of likelihood ratio is more little, and the reliability of this bit is low more.

6, soft decoding

Utilize the authentic communication of each bit of above-mentioned reception signal, each channel symbol to received signal adopts Soft decision decoding.Find out the minimum m of a reliability bit from n bit authentic communication of a code element, i.e. m minimum bit of the absolute value of reliable information.After this code element makes up this m bit, to having 2 ^mIndividual sequence.Each sequence is carried out Hard decision decoding, and the Euclidean distance of the authentic communication of code element behind the calculating Hard decision decoding and reception signal.Choose 2 ^mThe code element of Euclidean distance minimum is as decoded output symbol in the inferior decoding.

In the present embodiment, remember that the every bit corresponding reliable of a code element information is L=L ₁, L ₂..., L ₁₀₆, in these 106 authentic communications, seek out 3 minimum bits of absolute value of authentic communication, its corresponding sequence number might as well be assumed to be 1,2,3.These 3 bits are made up, obtain 8 code elements:

${\hat{x}}_0={\hat{s}}_1,{\hat{s}}_2,{\hat{s}}_3,...,{\hat{s}}_{106},$ ${\hat{x}}_1={\hat{s}}_1,{\hat{s}}_2,{\stackrel{\‾}{\hat{s}}}_3,...,{\hat{s}}_{106},$ ${\hat{x}}_2{={\stackrel{\‾}{\hat{s}}}_1,{\hat{s}}_2,\stackrel{\‾}{\hat{s}}}_3,...,{\hat{s}}_{106},$

${\hat{x}}_3={\stackrel{\‾}{\hat{s}}}_1,{\hat{s}}_2,{\hat{s}}_3,...,{\hat{s}}_{106},$ ${\hat{x}}_4={\stackrel{\‾}{\hat{s}}}_1,{\stackrel{\‾}{\hat{s}}}_2,{\hat{s}}_3,...,{\hat{s}}_{106},$ ${\hat{x}}_5={\stackrel{\‾}{\hat{s}}}_1,{\stackrel{\‾}{\hat{s}}}_2,{\stackrel{\‾}{\hat{s}}}_3,...,{\hat{s}}_{106},$

${{\hat{x}}_6={\hat{s}}_1,\stackrel{\‾}{\hat{s}}}_2,{\stackrel{\‾}{\hat{s}}}_3,...,{\hat{s}}_{106},$ ${\hat{x}}_7={\hat{s}}_1,{\stackrel{\‾}{\hat{s}}}_2,{\hat{s}}_3,...,{\hat{s}}_{106},$ Wherein $\stackrel{\‾}{\hat{s}}=\hat{s}\⊕1.$

After above 8 code elements are taken turns doing RS decoding, the code element after obtaining deciphering: $\hat{b}={\hat{b}}_1,{\hat{b}}_2,...,{\hat{b}}_{106}.$ Because the code element after the decoding is 0,1 sequence, the code element after therefore needing will decipher earlier converts-1 to ,+1 bipolar sequence: $\hat{c}={\hat{c}}_1,{\hat{c}}_2,...,{\hat{c}}_{106},$ ${\hat{c}}_k=2{\hat{b}}_k-1.$ The Euclidean distance d of the bipolarity code element after the calculating conversion and the authentic communication of this code element _i=|| c _i-L||.Therefrom find out the result of the minimum grouping of bits of Euclidean distance as soft decoding.

In an alternative embodiment of the invention, based on the result of joint-detection, can directly not calculate the authentic communication of each bit that receives signal, and change the authentic communication of calculating each symbol that receives signal into.On the 16QAM planisphere; Signal r after MMSE detects is designated as

and

nearest constellation point respectively with time nearest constellation point recently and is demodulation symbol, remembers that then the authentic communication of each demodulation symbol is:

$L={|r-\hat{\hat{a}}|}^2$

The L value is big more, and the reliability of demodulation symbol

is low more.

Each symbol corresponding reliable information of a code element of note is L ₁, L ₂..., L _N, in this N authentic communication, seek out 3 minimum symbols of absolute value of authentic communication, its corresponding sequence number might as well be assumed to be 1,2,3.These 3 symbols are made up, obtain 8 code elements:

${\hat{x}}_0={\hat{a}}_1,{\hat{a}}_2,{\hat{a}}_3,...,{\hat{a}}_N,$ ${\hat{x}}_1={\hat{a}}_1,{\hat{a}}_2,{\hat{\hat{a}}}_3,...,{\hat{a}}_N,$ ${\hat{x}}_2={\hat{a}}_1,{\hat{\hat{a}}}_2,{\hat{a}}_3,...,{\hat{a}}_N,$

${\hat{x}}_3={\hat{a}}_1,{\hat{\hat{a}}}_2,{\hat{\hat{a}}}_3,...,{\hat{a}}_N,$ ${\hat{x}}_4={\hat{\hat{a}}}_1,{\hat{a}}_2,{\hat{a}}_3,...,{\hat{a}}_N,$ ${\hat{x}}_5={\hat{\hat{a}}}_1,{\hat{a}}_2,{\hat{\hat{a}}}_3,...,{\hat{a}}_N,$

${\hat{x}}_6={\hat{\hat{a}}}_1{,\hat{\hat{a}}}_2,{\hat{a}}_3,...,{\hat{a}}_N,$ ${\hat{x}}_7={\hat{\hat{a}}}_1{,\hat{\hat{a}}}_2,{\hat{\hat{a}}}_3,...,{\hat{a}}_N.$

After above 8 code elements are taken turns doing RS decoding; Obtain 8 bit sequences; Again these 8 bit sequences are modulated into 8 symbol sebolic addressings, and then calculate these 8 symbol sebolic addressings and the Euclidean distance that receives signal, the corresponding bit sequence of Euclidean distance minimum value is the result of soft decoding.

The above is merely preferred embodiment of the present invention, not in order to restriction the present invention, all any modifications of within spirit of the present invention and principle, being done, is equal to and replaces and improvement etc., all should be included within protection scope of the present invention.

Claims (12)

1. signal acceptance method based on joint-detection and Soft decision decoding said method comprising the steps of:

A. carry out joint-detection to received signal;

B. based on the result of joint-detection, calculate the authentic communication that receives signal;

C. carry out Soft decision decoding to received signal based on said authentic communication;

It is characterized in that:

Among the step b, the authentic communication that calculate to receive each bit of signal be the likelihood ratio of each bit, and the absolute value of employing authentic communication weighs the reliability of each bit, and said absolute value is more little, and the reliability of said bit is low more;

Among the step c; The minimum bit of several reliabilitys of a code element to received signal makes up; Again the code element of each combination is carried out Hard decision decoding; And the Euclidean distance of the authentic communication of calculating Hard decision decoding result and said code element, choose the minimum grouping of bits of Euclidean distance as decoding output.

2. the method for claim 1 is characterized in that step a, and said reception signal is the CDMA signal, uses channel estimating and spreading code to carry out joint-detection to received signal.

3. the method for claim 1 is characterized in that step a, and said reception signal is an ofdm signal, uses channel estimating, carries out joint-detection to received signal.

4. the method for claim 1 is characterized in that step a, and the ofdm signal of said reception signal for expanding through sign indicating number uses channel estimating and sign indicating number to expand the aggregate channel matrix that matrix is formed, and carries out joint-detection to received signal.

5. like claim 2,3 or 4 described methods, it is characterized in that said channel estimating is to utilize pilot tone or training sequence to estimate channel, and carries out the channel match.

6. method as claimed in claim 5 is characterized in that, said channel match is meant carries out LPF.

7. signal acceptance method based on joint-detection and Soft decision decoding said method comprising the steps of:

A. carry out joint-detection to received signal;

B. based on the result of joint-detection, calculate the authentic communication that receives signal;

C. carry out Soft decision decoding to received signal based on said authentic communication;

It is characterized in that:

Among the step b, the authentic communication that calculate to receive each symbol of signal be and this symbol time nearest Euclidean distance, and the employing authentic communication weighs the reliability of each symbol, and the value of said authentic communication is more little, and the reliability of said symbol is low more;

Among the step c; The minimum symbol of several reliabilitys of a code element to received signal makes up; Again the code element of each combination is carried out Hard decision decoding; The result is modulated into symbol sebolic addressing with Hard decision decoding, calculates said symbol sebolic addressing and the Euclidean distance that receives signal, chooses the corresponding grouping of bits of the minimum symbol sebolic addressing of Euclidean distance as decoding output.

8. method as claimed in claim 7 is characterized in that step a, and said reception signal is the CDMA signal, uses channel estimating and spreading code to carry out joint-detection to received signal.

9. method as claimed in claim 7 is characterized in that step a, and said reception signal is an ofdm signal, uses channel estimating, carries out joint-detection to received signal.

10. method as claimed in claim 7 is characterized in that step a, and the ofdm signal of said reception signal for expanding through sign indicating number uses channel estimating and sign indicating number to expand the aggregate channel matrix that matrix is formed, and carries out joint-detection to received signal.

11., it is characterized in that said channel estimating is to utilize pilot tone or training sequence to estimate channel, and carries out the channel match like claim 8,9 or 10 described methods.

12. method as claimed in claim 11 is characterized in that, said channel match is meant carries out LPF.

Images