CN102655589B - Based on the combined signal source channel decoding method of variable length code and arithmetic code - Google Patents

Abstract

The invention discloses the combined signal source channel decoding method based on variable length code and arithmetic code, be applicable to the video encoding standard H.264 transmission of video be widely used at present, also be applicable to compression standard HEVC transmission of video of new generation simultaneously, and the image transmitting of JPEG2000, consider H.264, the entropy code characteristic of HEVC and JPEG2000, the present invention mainly comprises two modules, combined signal source channel arithmetic decoder and combined signal source channel decoding length changeable code device.The soft decoded information that combined signal source channel arithmetic code decoder exports as the soft input information of combined signal source channel decoding length changeable code device, can obtain best decoded symbol sequence through further joint variable-length code trellis structure search.The codeword structure information of variable length code can be utilized to delete invalid searching route in combined signal source channel arithmetic code decoder section, improve decoding performance, the method computation complexity is low, and time delay is little simultaneously, is applicable to actual video and image delivering system.

Description

Based on the combined signal source channel decoding method of variable length code and arithmetic code

Technical field

The present invention relates to a kind of video, the method in image communication technology field, particularly relate to the combined signal source channel decoding method based on variable length code and arithmetic code.

Background technology

In traditional communication system, source coding and channel coding carries out all respectively.This design principle is the message source and channel separation theorem based on Shannon.But in systems in practice, especially in multi-media transmission system, owing to postponing and the restriction of complexity, piece-rate system can not ensure the optimum of system.So in order to improve systematic function, people have focused on Joint Source Channel encoding and decoding sight.

H.264, video compression standard is widely used in recent years, and in the middle of the standard of video compression technology HEVC of new generation is also being formulated, their entropy code is all based on contextual arithmetic code CABAC.Due to the compression performance that CABAC is good, for H.264 and H.265 adding much beauty to.But CABAC due to its compression performance higher, contrary redundant information just retains less, very sensitive to error code.Decoding, once occur that error bit will spread by error code soon, causes the degradation of reconstruction video quality.Therefore, under error code environment, the error code resistivity how improving CABAC is exactly a problem demanding prompt solution.In order to provide good performance under error code environment, H.264 standard it is also proposed the method much improving video decode robustness at coding layer.Although these error codes opposing method can coding layer by error code diffusion-restricted in less time domain or spatial domain, but they cannot correct error of transmission, particularly when interchannel noise is very large, decoder may produce a lot of mistake, the serious quality reducing reconstruction video.

CABAC is compressed with two steps, the first step is first for symbol to be compressed binaryzation, the mode that binaryzation adopts has a primitive encoding, one primitive encoding of brachymemma, the modes such as k rank Exp-Golomb coding, symbol sebolic addressing is just sent into based on contextual two-value arithmetic code (CABAC) encoder after binaryzation, bit sequence after two-value arithmetic code coding is served noisy communication channel through modulation, at receiving terminal, the people such as Salma Ben are in " IEEEtrans on communication " (ieee communication transactions), 2009, 57 volume 7 phases, pp.2014-2023 delivers and has been entitled as " improved sequential MAP estimation of CABAC encoded data withobjective adjustment of the complexity/efficiency tradeoff ", this article devises the combined signal source channel decoding device based on maximum a posteriori probability (MAP) of the code part that to count based on CABAC two-value, but she does not consider the application of policies of combined decoding to anti-binary conversion treatment device, consequently, once wrong from the data that the arithmetic code joint decoder of CABAC is come, will spread by error code with during common anti-binary conversion treatment device decoded symbol sequence, cause gross error.In fact at present comparatively conventional video standard H.264, HEVC, and image compression standard JPEG 2000 all first converted symbol sebolic addressing to be compressed to binary sequence and then does two-value arithmetic code coding before the coding that counts, and the conversion of this binary sequence is exactly the process of a variable length coding, current to combine the code decoding that counts be all the part that only considered yard combined decoding that counts, and usually also all will not relate to through this process of decoding length changeable code after the code decoding that counts in video and image procossing.Thus, we expect designing the combined signal source channel decoding device based on variable length code and the code that counts being applicable to video and image transmitting, the combined decoding part of variable length code is also taken into account simultaneously, improve the performance of combined decoding with this further, also can be applied in actual transmission system simultaneously and go.

Summary of the invention

For above-mentioned technological deficiency, the present invention proposes the combined signal source channel decoding method based on variable length code and arithmetic code.

In order to solve the problems of the technologies described above, technical scheme of the present invention is as follows:

Based on the combined signal source channel decoding method of variable length code and arithmetic code, comprising length is the source symbol sequence a={a of N ₁, a ₂... a _i... a _nbinaryzation coding is done through the binaryzation encoder of CABAC, export the binary sequence b={b that length is S ₁, b ₂... b _s; Through the two-value arithmetic code encoder of CABAC, two-value arithmetic code coding is carried out to described binary sequence, exports the binary sequence x={x that length is M ₁, x ₂... x _m, become sequence y={y through binary phase shift keying BPSK modulation and chnnel coding ₁, y ₂... y _nsend into noisy communication channel afterwards;

Sequence is received at receiving terminal after obtain output information through channel-decoding

Described sequence for sequences y={ y ₁, y ₂... y _nthe sequence that obtains through noisy communication channel;

Described for the sequence that x sequence obtains through noisy communication channel;

Comprise the steps:

11) the likelihood ratio LLR (x of output sequence is calculated at combined signal source channel arithmetic code decoder by formula (a), formula (b) and formula (c) _i), and then according to x _iwith b _ibe the relation mapped one by one, obtain

$\left\{\begin{array}{c}{\mathrm{\α}}_i={\mathrm{\α}}_{i-1}+{\mathrm{\γ}}_i\\ {\mathrm{\α}}_0\left(0\right)=0\end{array}\right.i=\mathrm{1,2},...M---\left(a\right)$

$\left\{\begin{array}{c}{\mathrm{\β}}_{i-1}=\ln \underset{{\mathrm{\σ}}_i}{\mathrm{\Σ}}\exp ({\mathrm{\γ}}_i+{\mathrm{\β}}_i),\\ B_S={\mathrm{\Π}}_{i=0}^{M-1}\ln P\left(x_i\right)\end{array}\right.i=\mathrm{1,2},...M---\left(b\right)$

$\mathrm{LLR}\left(x_i\right)=\ln \left[\frac{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=1}\exp ({\mathrm{\α}}_{i-1}+{\mathrm{\λ}}_i+{\mathrm{\β}}_i)}{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=0}\exp ({\mathrm{\α}}_{i-1}+{\mathrm{\λ}}_i+{\mathrm{\β}}_i)}\right]---\left(c\right)$

Described α _i=ln A _i, β _i=ln B _i, described σ _irepresent the decoded state of each state node in binary tree; Described a _i-1for forward recursion; $G_i({\mathrm{\σ}}_{i-1},{\mathrm{\σ}}_i)=P\{{\mathrm{\σ}}_i,{\hat{x}}_i/{\mathrm{\σ}}_{i-1}\},$ G _ifor σ _i-1→ σ _ibranch transition probability; $B_i\left({\mathrm{\σ}}_i\right)=P\{{\hat{x}}_{i+1}^{\left(M\right)}/{\mathrm{\σ}}_i\},$ B _ifor backward recursion

12) arithmetic decoding of combined signal source channel adopts the M stack algorithm of breadth First to search for optimal path, at each search depth i, N number of state node with maximum metric is only retained for subsequent expansion according to formula (a), and the expansion of each state node has two paths, one is 0, and another is 1; The variable length code code word whether output sequence on current decoding paths is legal follow-up decoding length changeable code device is detected when extensions path, if not, then delete current path, when search depth reaches the length of binary sequence, search terminates, and the maximum backward progressively recursion in path of selectance value just can obtain optimal estimation sequence;

13) by step 11) and 12) after Output rusults be input to combined signal source channel decoding length changeable code device, in described combined signal source channel decoding length changeable code device, carry out path metrics value by formula (d) and formula (e);

$P(\widetilde{b_k}=l)=\left\{\begin{array}{ccc}\frac{\exp \left(\mathrm{LLR}\left(\widetilde{b_k}\right)\right)}{1+\exp \left(L\left(\widetilde{b_k}\right)\right)}& \mathrm{if}& l=0\\ \frac{1}{1+\exp \left(\mathrm{LLR}\left(\widetilde{b_k}\right)\right)}& \mathrm{if}& l=1\end{array}\right.---\left(d\right)$

$\log P\left(B\right|\tilde{B})=\underset{i=0}{\overset{S-1}{\mathrm{\Σ}}}[\left(\underset{j=1}{\overset{i_U}{\mathrm{\Σ}}}\log P\left(\widetilde{b_i^j}\right|b_i^j)\right)+\log P\left(b_i\right|b_{i-1})+V\left(\widetilde{b_i}\right)\log 2]---\left(e\right)$

14) in combined signal source channel decoding length changeable code device, adopt trellis structure way of search to calculate the optimal path of decoding, in each bit decoding moment in trellis structure, all active paths are sorted by measure formulas (e) value, only retain the maximum K paths of metric, when depth of decode equals the length of the bit sequence received, more current all paths retained, there is maximum path metric value and that paths that its decoding symbols number exported meets coding symbol number just can think optimal path, this path can obtain output decoded symbol sequence by backward recursion.

Further, in combined signal source channel decoding length changeable code device part, if do not have a paths to have correct symbolic number in the backward derivation of trellis structure search, the combined signal source channel decoding length changeable code device path of selecting cumulative metric value maximum is as the Output rusults of combined signal source channel decoding length changeable code.

Beneficial effect of the present invention is: to encode based on CABAC H.264 and HEVC transmission of video data acquisition is decoded with the decoding process of the combined signal source channel having merged arithmetic code and variable length code and be corrected in and transmit in noisy communication channel time some error bits of causing, thus obtain the performance of better error code opposing, prevent and reduce error code diffusion.The feature that make use of variable length code codeword structure in the path search process of combined signal source channel arithmetic code decoder well, to delete some Invalid paths, reaches the object improving decoding performance further.

Accompanying drawing explanation

Fig. 1 is the structured flowchart of combined signal source channel decoding device of the present invention;

Fig. 2 is arithmetic code solution code tree;

Fig. 3 is the bit search block diagram of combined signal source channel decoding length changeable code device.

Embodiment

Below in conjunction with the drawings and specific embodiments, the present invention is described further.

The present invention proposes a kind of combined signal source channel decoding scheme based on arithmetic code and variable length code, actual video compression standard can be applied to H.264, in the transmission system of HEVC and image compression standard JPEG 2000.As shown in Figure 1, at transmitting terminal, multimedia symbols sequence containing redundancy is first through CABAC encoder, this encoder comprises two parts, binary conversion treatment device and two-value arithmetic code encoder, the output of CABAC encoder is a binary bit sequence, then changes sequence and is sent to the chnnel coding of next stage and modulation module by chnnel coding and is modulated into the waveform being adapted at transmitting in wireless channel, transmit.After noisy communication channel, at receiving terminal, reception antenna first carries out demodulation to the information sequence received, and channel-decoding, obtain the bit soft information sequence based on likelihood ratio, just can adopt this Soft Inform ation sequence afterwards and decode based on the combined signal source channel decoding method of arithmetic code and variable length code, the final estimate symbol sequence exporting the best.

1, the realization of combined signal source channel arithmetic code decoder.

(1) based on the combined signal source channel arithmetic decoding of bcjr algorithm

The solution code tree schematic diagram of arithmetic code as shown in Figure 2.This On Binary Tree Representation length is all possible codeword sequence x of M _i, each state node in tree represents decoded state σ _i, G _i(σ _i-1, σ _i) represent by x _ithe branch transition probability caused.Decoder is according to code word bits x _iconstantly from state σ _i-1be updated to state σ _i, and decoding obtains the source symbol subsequence of different length.Thus, the derivation of equation of combining the maximum a posteriori probability of arithmetic decoding is as follows:

Suppose that length is the source symbol sequence a={a of N ₁, a ₂... a _i... a _nthe binary sequence b={b that length is S is become after binaryzation variable length coding ₁, b ₂... b _s, and then enter two-value arithmetic code encoder (variable length coding), export the binary sequence x={x that length is M ₁, x ₂... x _m, after through BPSK modulation and chnnel coding become sequence y={y ₁, y ₂... y _n, serve noisy communication channel, the sequence received at receiving terminal is channel decoding obtains Soft output information can think that x sequence is through noisy communication channel. obtain after conversion:

$P({\hat{x}}_i/x_i)=\left\{\begin{array}{ccc}\frac{\exp \left(\mathrm{LLR}\left({\hat{x}}_i\right)\right)}{1+\exp \left(\mathrm{LLR}\left({\hat{x}}_i\right)\right)}& \mathrm{if}& x_i=0\\ \frac{1}{1+\exp \left(\mathrm{LLR}\left({\hat{x}}_i\right)\right)}& \mathrm{if}& x_i=1\end{array}\right.---\left(1\right)$

The soft decoding of combined signal source channel of arithmetic code will estimate posterior probability (APP) exactly because for given receiving sequence, be a constant, therefore only can calculate joint probability according to bayesian criterion, λ _im () can be expressed as in code tree by x _ithe state transitions σ that=m causes _i-1→ σ _iall possible paths average, namely due to λ _ithe value of (m) not only with and with with relevant, namely receiving sequence can be decomposed into thus have

${\mathrm{\λ}}_i\left(m\right)=\underset{({\mathrm{\σ}}_{i-1},{\mathrm{\σ}}_i):x_i=m}{\mathrm{\Σ}}{A}_{i-1}\left({\mathrm{\σ}}_{i-1}\right)G_i({\mathrm{\σ}}_{i-1},{\mathrm{\σ}}_i)B_i\left({\mathrm{\σ}}_i\right)---\left(2\right)$

Wherein $A_{i-1}\left({\mathrm{\σ}}_{i-1}\right)=P\{{\mathrm{\σ}}_{i-1},{\hat{x}}_1^{i-1}\},A_{i-1}$ For forward recursion; $G_i({\mathrm{\σ}}_{i-1},{\mathrm{\σ}}_i)=P\{{\mathrm{\σ}}_i,{\hat{x}}_i/{\mathrm{\σ}}_{i-1}\},G_i$ For σ _i-1→ σ _ibranch transition probability; for backward recursion.For branch transition probability G _i(σ _i-1, σ _i), owing to separating in code tree from state σ _i-1be updated to state σ _ibranch just in time correspond to code word bits x _i, thus have

$G_i({\mathrm{\σ}}_{i-1},{\mathrm{\σ}}_i)=P\{{\mathrm{\σ}}_i,{\hat{x}}_i/{\mathrm{\σ}}_{i-1}\}=P\{{\hat{x}}_i,x_i\}=P\{{\hat{x}}_i/x_i\}P\left(x_i\right)---\left(3\right)$

Suppose that arithmetic code used is desirable, the code word bits so obtained should be the general distribution such as independent, namely can think P (x _i) ≈ 0.5.In this case, G _i(σ _i-1, σ _i) value just only and prior probability relevant, this prior probability is also the input of soft-output coding associating arithmetic code decoder.

And forward direction A _i-1with backward B _iderivation can be continued obtain:

$A_i=\frac{{\mathrm{\Σ}}_{{\mathrm{\σ}}_{i-1}}{A}_{i-1}{G}_i/P\left({\hat{x}}_1^{k-1}\right)}{{\mathrm{\Σ}}_{{\mathrm{\σ}}_i}{\mathrm{\Σ}}_{{\mathrm{\σ}}_{i-1}}{A}_{i-1}{G}_i/P\left({\hat{x}}_1^{k-1}\right)}=\frac{{\mathrm{\Σ}}_{{\mathrm{\σ}}_{i-1}}{A}_{i-1}{G}_i}{{\mathrm{\Σ}}_{{\mathrm{\σ}}_i}{\mathrm{\Σ}}_{{\mathrm{\σ}}_{i-1}}{A}_{i-1}{G}_i}---\left(4\right)$

$B_{i-1}=\frac{B_{i-1}}{P({\hat{x}}_k^M/{\hat{x}}_1^{k-1})}=\frac{{\mathrm{\Σ}}_{{\mathrm{\σ}}_i}{B}_i{G}_i}{P({\hat{x}}_{k+1}^M/{\hat{x}}_1^k)P\left({\hat{x}}_1^k\right)/P\left({\hat{x}}_1^{k-1}\right)}=\frac{{\mathrm{\Σ}}_{{\mathrm{\σ}}_i}{B}_i{G}_i}{{\mathrm{\Σ}}_{{\mathrm{\σ}}_i}{\mathrm{\Σ}}_{{\mathrm{\σ}}_{i-1}}{A}_{i-1}{G}_i}---\left(5\right)$

In addition represent the prior probability of information source binary sequence, the decoding paths of its correspondence terminates in state σ _m.For forward recursion A _i-1with backward recursion B _ican realize respectively by backward recursion before as follows.

$\left\{\begin{array}{c}{A}_i=A_{i-1}{G}_i\\ A_0\left(0\right)=1\end{array}\right.i=\mathrm{1,2},...M---\left(6\right)$

$\left\{\begin{array}{c}{B}_{i-1}=\underset{{\mathrm{\σ}}_i}{\mathrm{\Σ}}{G}_i{B}_i\\ B_S={\mathrm{\Π}}_{i=0}^{M-1}P\left(x_i\right)\end{array}\right.i=\mathrm{1,2},...M---\left(7\right)$

In order to simplify calculating, above-mentioned formula being converted into logarithmic form, obtaining maximum a posteriori probability by log-likelihood ratio and exporting:

$\mathrm{LLR}\left({\hat{x}}_i\right)=\ln \left[\frac{{\mathrm{\λ}}_i(m=1)}{{\mathrm{\λ}}_i(m=0)}\right]=\ln \left[\frac{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=1}{A}_{i-1}{G}_i{B}_i}{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=0}{A}_{i-1}{G}_i{B}_i}\right]---\left(8\right)$

Simplify this formula, molecule denominator with divided by and make α _i=ln A _i, β _i=ln B _i, above formula becomes

$\mathrm{LLR}\left(x_i\right)=\ln \left[\frac{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=1}\exp ({\mathrm{\α}}_{i-1}+{\mathrm{\λ}}_i+{\mathrm{\β}}_i)}{{\mathrm{\Σ}}_{\left({\mathrm{\σ}}_{i-1,{\mathrm{\σ}}_i}\right);x_i=0}\exp ({\mathrm{\α}}_{i-1}+{\mathrm{\λ}}_i+{\mathrm{\β}}_i)}\right]---\left(9\right)$

α _i-1and β _icalculating respectively as follows:

$\left\{\begin{array}{c}{\mathrm{\α}}_i={\mathrm{\α}}_{i-1}+{\mathrm{\γ}}_i\\ {\mathrm{\α}}_0\left(0\right)=0\end{array}\right.i=\mathrm{1,2},...M---\left(10\right)$

$\left\{\begin{array}{c}{\mathrm{\β}}_{i-1}=\ln \underset{{\mathrm{\σ}}_i}{\mathrm{\Σ}}\exp ({\mathrm{\γ}}_i+{\mathrm{\β}}_i),\\ B_S={\mathrm{\Π}}_{i=0}^{M-1}\ln P\left(x_i\right)\end{array}\right.i=\mathrm{1,2},...M---\left(11\right)$

Obtain the prior information γ of channel _iwith the prior probability P (x of binary sequence information source _i) after, according to above formula (10), (11) carry out forward and backward recursion, just can obtain the result LLR (x of soft output arithmetic decoding _i).And x _iwith b _ithe relation mapped one by one, can according to LLR (x _i), obtain

If according to derivation above, directly in the words of separating the enterprising row operation of code tree, need to use the codeword sequence that all length is M, and the source sequence length of the practical application of its correspondence usually is all greater than 1000, due to path number along with search depth exponentially increases, the computing realizing high like this complexity is a challenge greatly to system resource, therefore, is necessary that emphasis is considered and designs the strategy deleting Invalid path.

2, M stack search algorithm

The code that counts is two-value tree-shaped code, therefore, considers to adopt the mode of sequence search to obtain optimal estimation sequence.But it is unpractical for directly searching for optimal path according to path metric formula (10) in all possible path, because along with the increase of decoded bits length, the path of separating in code tree also can increase with index rank fast.For this reason, the sequence search mode finding suboptimum is needed.Adopt the M stack algorithm (Multiple stack algorithm) of breadth First herein, at each search depth i, N number of state node with maximum metric is only retained for subsequent expansion according to formula (10), and the expansion of each state node has two paths, one is 0, and another is 1.In addition, deletion path policy be in the past all mainly adopt in arithmetic code, add availability constraint, then in decoding end once detect that availability constraint just deletes this path.If this method of adding error correction information when arithmetic code is encoded is applied to CABAC, the change of the probabilistic model of current CABAC will certainly be caused, affect its code efficiency.Therefore, do not consider to add availability constraint at coding side, but in path search process, consider that correct decoding paths should be to meet the code word that follow-up VLC (variable length code) decodes, thus, whether the output sequence detected when extensions path on current decoding paths is legal vlc code word, if not, then delete current path.When this search depth reaches the length of binary sequence, search terminates, and the maximum backward progressively recursion in path of selectance value just can obtain optimal estimation sequence.

The realization of combined signal source channel decoding length changeable code device (JVLD):

Also need to carry out variable length code combined decoding from combining yard bit information for decoder output that counts.Variable length code combined signal source channel decoding adopts two kinds to search for trellis structure usually.The trellis structure that first is-symbol is limited, proposed by Sayood, (Khalid Sayood and HasanH, " Joint source/cahnnel coding for variablelength codes; " ieee communication transactions, 2000,48th volume the 5th phase, pp.787-794.) each paths all comprises identical symbolic number.Another is then the limited trellis structure of bit, by Park, (M.Park andD.J.Miller is proposed, " Joint source-channel decoding for variable length encoded data by exactand approximate MAP sequence estimation ", ieee communication transactions, 2000,48th volume pp.1-6.), each paths all comprises identical bit number.The performance of performance lower than the limited combined decoding of bit of the limited combined decoding of symbol is found in l-G simulation test.This is because the decoding accuracy that decoding accuracy that bit is limited is limited compared with symbol is high.Also find simultaneously, in the decode procedure that bit is limited, in each bit decoding moment, to all retain K bar optimal path under each decoded state, in fact increasing the Path complexity that will search for, also limit the selection of optimal path simultaneously.Based on above consideration, propose the bit limited variable length combined decoding algorithm of improvement, as shown in Figure 3.In each bit decoding moment, some paths of its best might not intersperse among various possible decoded state, and be likely only present in one and several decoded state, therefore, be different from the delet method of Park and Miller, retain some optimal paths in each decoded state, the deletion mode of employing is in each bit decoding moment, all active paths are sorted, only retains optimum K paths.In each decoding moment in schematic diagram Fig. 3, only retain two optimal paths in various possible path.Clearly, when the number of coded identification collection increases, the method can reduce computation complexity greatly.

As shown in Figure 3, in horizontal elements represent bit-time, be a group code of n at each status representative bit length of time n.When depth of decode reaches S bit, more current all paths retained, have maximum path metric value and that paths that its decoding symbols number exported meets coding symbol number just can think optimal path.

The derivation of path metric formula is as follows:

Suppose one group of information source a=(a ₀, a ₁... a _n-1) include N number of source symbol a _i, this information source becomes b=(b through two-value arithmetic code encoder encodes ₀, b ₁... b _n-1), wherein each a _iall be encoded into the bit vectors of different length b _urepresent b _ibit length.These bit vectors are connected in series to and just can form the binary sequence that length is S together.Then this sequence a is by becoming after noisy communication channel so just can obtain the posterior probability of this sequence:

$P\left(B\right|\tilde{B})=\underset{i=0}{\overset{S-1}{\mathrm{\Π}}}\frac{P\left(\widetilde{b_i}\right|b_i)P\left(b_i\right|b_{i-1})}{2^{-V\left(\widetilde{b_i}\right)}}=\underset{i=0}{\overset{S-1}{\mathrm{\Π}}}\frac{\left(\underset{j=1}{\overset{i_U}{\mathrm{\Π}}}P\left(\widetilde{b_i^j}\right|b_i^j)(P\left(b_i\right|b_{i-1}\right)}{2^{-V\left(\widetilde{b_i}\right)}}---\left(12\right)$

Here, channel transition probability, P (b _i| b _i-1) then depend on source statistics, representative coding required bit number. the Soft Inform ation that can be exported by previous stage combined signal source channel arithmetic code decoder obtain after conversion.It is defined as in order to use it for JVLD, doing as down conversion to it, obtaining

$P(\widetilde{b_k}=l)=\left\{\begin{array}{ccc}\frac{\exp \left(\mathrm{LLR}\left(\widetilde{b_k}\right)\right)}{1+\exp \left(L\left(\widetilde{b_k}\right)\right)}& \mathrm{if}& l=0\\ \frac{1}{1+\exp \left(\mathrm{LLR}\left(\widetilde{b_k}\right)\right)}& \mathrm{if}& l=1\end{array}\right.---\left(13\right)$

Thus, the logarithmic form of formula (12) is adopted to represent path metric criteria:

$\log P\left(B\right|\tilde{B})=\underset{i=0}{\overset{S-1}{\mathrm{\Σ}}}[\left(\underset{j=1}{\overset{i_U}{\mathrm{\Σ}}}\log P\left(\widetilde{b_i^j}\right|b_i^j)\right)+\log P\left(b_i\right|b_{i-1})+V\left(\widetilde{b_i}\right)\log 2]---\left(14\right)$

Such integrated source and channel variable-length decoder JVLD just contains the process of the expansion cumulative path of a forward direction and a backward rollback path search process finds the best path with maximum a posteriori probability.The tolerance of each paths is represented by formula (14).Numerical value is larger, and representing it may be that the probability of optimal path is larger.The length that whole decode procedure is performed until decoding paths equals to receive the length of bit sequence.Finally, sweep backward finds optimal path sequence, first in all path lists, finds out the correct decoding paths of decoding symbols number, more therefrom finds out the output sequence of the maximum path of cumulative metric value as final decoding.At the end of decoding, having correct symbolic number and bit number, and the maximum path of metric is as the Output rusults of decoder.If do not have a paths to have correct symbolic number, the path that decoder selects cumulative metric value maximum is as the Output rusults of decoding.

The above is only the preferred embodiment of the present invention; it should be pointed out that for those skilled in the art, without departing from the inventive concept of the premise; can also make some improvements and modifications, these improvements and modifications also should be considered as in scope.

Claims (2)

1., based on the combined signal source channel decoding method of variable length code and arithmetic code, comprising length is the source symbol sequence a={a of N ₁, a ₂... a _i... a _nbinaryzation coding is done through the binaryzation encoder of CABAC, export the binary sequence b={b that length is S ₁, b ₂... b _s; Through the two-value arithmetic code encoder of CABAC, two-value arithmetic code coding is carried out to described binary sequence, exports the binary sequence x={x that length is M ₁, x ₂... x _m, become sequence y={y through binary phase shift keying BPSK modulation and chnnel coding ₁, y ₂... y _nsend into noisy communication channel afterwards; Sequence is received at receiving terminal after obtain output information through channel-decoding

Described sequence for sequences y={ y ₁, y ₂... y _nthe sequence that obtains through noisy communication channel; Described for the sequence that x sequence obtains through noisy communication channel;

It is characterized in that, comprise the steps:

11) the likelihood ratio LLR (x of output sequence is calculated at combined signal source channel arithmetic code decoder by formula (a), formula (b) and formula (c) _i), and then according to x _iwith b _ibe the relation mapped one by one, obtain

Described α _i=ln A _i, β _i=ln B _i, described σ _irepresent the decoded state of each state node in binary tree; Described a _i-1for forward recursion; g _ifor σ _i-1→ σ _ibranch transition probability; b _ifor backward recursion, in for sequence a middle kth symbol represents;

12) decoding of combined signal source channel arithmetic code adopts the M stack algorithm of breadth First to search for optimal path, at each search depth i, N number of state node with maximum metric is only retained for subsequent expansion according to formula (a), and the expansion of each state node has two paths, one is 0, and another is 1; The variable length code code word whether output sequence on current decoding paths is legal follow-up decoding length changeable code device is detected when extensions path, if not, then delete current path, when search depth reaches the length of binary sequence, search terminates, and the maximum backward progressively recursion in path of selectance value just can obtain optimal estimation sequence;

13) will through step 11) and step 12) Output rusults be input to combined signal source channel decoding length changeable code device, in described combined signal source channel decoding length changeable code device, carry out path metrics value by formula (d) and formula (e);

14) in combined signal source channel decoding length changeable code device, adopt trellis structure way of search to calculate the optimal path of decoding, in each bit decoding moment in trellis structure, all active paths are sorted by measure formulas (e) value, only retain the maximum K paths of metric, when depth of decode equals the length of the bit sequence received, more current all paths retained, there is maximum path metric value and that paths that its decoding symbols number exported meets coding symbol number is exactly optimal path, this path can obtain output decoded symbol sequence by backward recursion.

2. the combined signal source channel decoding method based on variable length code and arithmetic code according to claim 1, it is characterized in that, in combined signal source channel decoding length changeable code device part, if do not have a paths to have correct symbolic number in the backward derivation of trellis structure search, the combined signal source channel decoding length changeable code device path of selecting cumulative metric value maximum is as the Output rusults of combined signal source channel decoding length changeable code.