CN103532674A - File transmission encoding method based on complex network - Google Patents
File transmission encoding method based on complex network Download PDFInfo
- Publication number
- CN103532674A CN103532674A CN201310507723.6A CN201310507723A CN103532674A CN 103532674 A CN103532674 A CN 103532674A CN 201310507723 A CN201310507723 A CN 201310507723A CN 103532674 A CN103532674 A CN 103532674A
- Authority
- CN
- China
- Prior art keywords
- grouping
- coded identification
- original
- limit
- another
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Landscapes
- Mobile Radio Communication Systems (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a file transmission encoding method based on a complex network, and relates to the field of reliable transmission of wireless network data. The method comprises the processes as follows: the sizes of to-be-transmitted files are counted, and the to-be-transmitted files are grouped; a degree distribution is selected; and a new group, namely, a code symbol, is generated on the basis of multiple groups of to-be-output original files, so that the original file is transmitted in the network in a code symbol sequence form. According to the method, a mapping relation between middle and short ring structures of LT (Luby transform) codes and a corresponding complex network structure is established, detection is performed according to the complex network structure, short ring structures in the encoding process of the LT codes are avoided to the largest extent to improve the success probability of coding, and an efficient and reliable file transmission encoding scheme based on the complex network is constructed. By means of the method, transmission of data with a time-variant characteristic in a wireless channel is particularly facilitated, and a receiving end can receive less data and can restore original data, so that the bandwidth utilization ratio of the channel can be effectively increased, and the transmission delay is reduced.
Description
Technical field
The present invention relates to the reliability transmission field of wireless network data, particularly the becate of a kind of employing based on complex network detects and the LT code scheme of evading algorithm carries out the method for transfer files.
Background technology
Luby Transform code (being called for short LT code) is that the class being proposed in 2002 by Luby has the digital fountain code without code check characteristic.Its coding and decoding performance directly depends on that the degree of its coded identification distributes.Desirable orphan is distributed in and can realizes in theory optimum LT code coding/decoding performance, and the coding symbol number that recovering K incoming symbol needs is K.Yet because the LT code distributing based on desirable orphan can not guarantee that the element number of preliminary treatment collection in whole decode procedure is always 1, the fluctuation that preliminary treatment collection is small causes whole decode procedure failure.In order to address the above problem, Luby has further proposed the LT code distributing based on sane orphan, and it makes in whole decode procedure, and in preliminary treatment, element number is stabilized in one and is greater than in 1 scope.The performance of the LT code distributing based on sane orphan is better than the performance of the LT code based on desirable orphan's distribution, becomes a standard design method of current LT code.Existing LT code scheme is mostly that to take the LT code distributing based on sane orphan be the improvement that basis is done, as the LT code distributing based on inferior goodness, multiple-objection optimization LT code, adopt the LT code of chaology design etc.2012, Zhao etc. utilized the short average path length characteristic of scale-free networks network, have proposed a class and have had the LT code without characteristics of scale, make the coding and decoding efficiency of LT code and successfully decoded probability have larger improvement.And based on this, associated ideal orphan degree of being distributed in discharges the advantage of probability aspect, has proposed a class and has had the sane LT code without characteristics of scale, has further improved the coding and decoding performance of LT code.
The research of recent years finds that the appearance of becate has a certain impact to the volume of LT code (translating) code performance tool.For cataloged procedure, coded identification expectation obtains the information that more incoming symbols are transmitted, yet the existence of becate makes the information of incoming symbol repeat to pass to coded identification, causes the waste of coded identification.Zhang etc. proposed to adopt extraction related column method to evade the becate structure in LT code in " Little Cycle elimination of Fountain codes using Distilling Correlative Columns " literary composition in 2008, at " A cycle elimination algorithm for construction of LT codes " literary composition, by PEG algorithm, the becate for LT code detects and evades algorithm Zhou etc., makes data in ideal communication channel, obtain better successfully decoded probability.
Above-mentioned two kinds of methods all only propose for the Fourth Ring structure occurring in Tanner figure, the impact of the becate structure of not considering to be more than or equal to six rings on LT code.In addition, because the coded identification of LT code generates as required, when channel condition is poor, the random generation of degree will inevitably produce Fourth Ring and six ring structures, and now, the degree that this two schemes has changed original coding symbol distributes.Cause the incoming symbol of decoding equal number need to transmit more coded identification, take more channel width.Therefore, the maximization becate problem of LT code remains a problem demanding prompt solution.
Summary of the invention
The deficiency existing for prior art, the object of this invention is to provide a kind of file transfer coding method based on complex network, evades Fourth Ring and six ring structures of LT code, to overcome the problem of the maximization becate of LT code in testing process.
Technical scheme of the present invention is achieved in that a kind of file transfer coding method based on complex network, comprises the following steps:
Step 1: add up file size to be transmitted, file waiting for transmission is divided into groups;
Described grouping process is: file to be transmitted is divided into k group by bit-order, and wherein, k is positive integer, and k>10
3if the data volume in last grouping is less than average amount, by the remaining bits position zero padding in this grouping;
Step 2: select a degree according to k value and distribute;
Step 3: generate new grouping in a plurality of packet by packet basis of original to be exported, coded identification, makes original transmit in network with the form of coded identification sequence;
Step 3.1: generate first coded identification;
Step 3.1.1: the degree of being selected by step 2 distributes and generates at random a value d, d different grouping of random selection in former output file, wherein, d is natural number;
Step 3.1.2: XOR is carried out in this d different grouping, forms a new grouping, this grouping is first coded identification for generating just;
Step 3.1.3: the coded identification that the step 3.1.2 of take is generated is start node, generates a network configuration being comprised of a node;
Step 3.1.4: this coded identification is placed in channel and is transmitted;
Step 3.2: generate next coded identification;
Step 3.2.1: distribute according to the degree of step 2, by this degree, distribute and generate at random another positive integer d ', from k grouping of original, select a grouping at random, the grouping that this is grouped into first original selected that generates next coded identification selection is also first the optional grouping that generates next coded identification.In network configuration, add a new node.If a grouping of the original of selecting be generate other certain or some there is a grouping of coded identification, so, one of having between corresponding node in network configuration that the weights that connect ，Qie limit, limit are the original selected of this new node and above-mentioned certain or some already present coded identification is grouped in the sequence number in the grouping of all k original; If d '=1, performs step 3.2.2; Otherwise execution step 3.2.3;
Grouping complete zero bit sequence isometric with it of the original that step 3.2.2: step 3.2.1 selects carries out XOR, form a new grouping, the next coded identification that this is grouped into generation, is placed on this coded identification in channel and transmits, and to step 3.3, continues to carry out;
Step 3.2.3: except the grouping of having selected for the next coded identification of generation, select another grouping of original in remaining grouping in all groupings of original, this is grouped into another grouping selected that generates next coded identification;
Step 3.2.4: find out that all another groupings in step 3.2.3 participate in to generate there is coded identification and in network configuration corresponding node, check that whether corresponding with these node of new node in step 3.2.1 exists and connect limit.If the company of existence limit, does not show that another grouping of selecting can not cause the Fourth Ring structure of LT code, continues to carry out from step 3.2.5; If the company of existence limit, shows that this another grouping will cause the Fourth Ring structure of LT code, this another grouping is not the optional grouping that generates next coded identification.Now, if still exist for, generate the not grouping of the original of selected mistake of next coded identification, get back to step 3.2.3 and continue to carry out, otherwise continue to carry out to step 3.2.9;
Step 3.2.5: find out that all another groupings in step 3.2.3 participate in to generate there is coded identification and in network configuration corresponding node, check in network configuration with new node and exist the node that connects limit whether to have even limit with the coded identification corresponding node in network configuration that exists that another grouping participates in generating.If the weights on the company of existence limit or the company of existence limit and limit do not equal the sequence number in another all groupings that are grouped in original, show that so another grouping can not cause six ring structures of LT code, this another be grouped into the optional grouping that generates next coded identification, to step 3.2.6, continue to carry out; If the company of existence limit, and the weights that connect limit are not equal to the sequence number in another grouping that is grouped in original, show to exist six ring structures, i.e. this another grouping will cause six ring structures of LT code, and this another grouping is not the optional grouping that generates next coded identification.Now, if still there is the not grouping of the original of selected mistake, get back to step 3.2.3 and continue to carry out; Otherwise, from step 3.2.9, continue to carry out;
Step 3.2.6; There is coded identification and corresponding node in network configuration in that finds out that all another groupings in step 3.2.3 participate in to generate, in network configuration, add the company limit of these corresponding nodes and new node, the weights on limit are grouped in the sequence number in all groupings of original for another;
Step 3.2.7: if equaling d ', optional grouping number continues to carry out from step 3.2.8, individual if optional grouping number is less than d ', from step 3.2.9, continue to carry out;
Step 3.2.8: the individual grouping of this d ' is carried out XOR and obtained a new grouping, this is new is grouped into next coded identification.This coded identification is placed in channel and is transmitted; Execution step 3.3;
Step 3.2.9: if still there is the not grouping of the original of selected mistake, repeating step 3.2.3; If there is no the not grouping of selected mistake, beyond optional grouping, select arbitrarily the grouping of original so, until selecteed packet count is d '.The individual grouping of this d ' is carried out XOR and is generated a new grouping, i.e. next coded identification, and this coded identification is placed in channel and is transmitted;
Step 3.3: the process of repeated execution of steps 3.2, generates remaining coded identification, until transmitting terminal receives the feedback information that file that receiving terminal returns to has correctly recovered.
Beneficial effect of the present invention: the present invention has set up the mapping relations between the complex network structures that LT code short-and-medium ring structure is corresponding with it, according to complex network structures, detect and evade to greatest extent the becate structure in LT code cataloged procedure, improve its successfully decoded probability, constructed a kind of efficient and reliable file transfer encoding scheme based on complex network.Especially for the transfer of data having in the wireless channel of time-varying characteristics, this invention can effectively improve the bandwidth availability ratio of channel so that receiving terminal receives less data just can be recovered initial data, reduces propagation delay time.
Accompanying drawing explanation
Fig. 1 is that the file that embodiment of the present invention is N=kL by bit number is on average divided into k grouping schematic diagram;
Fig. 2 is that the Divide File that embodiment of the present invention is N=kL-δ (δ < L) by bit number becomes k grouping schematic diagram;
Fig. 3 is the embodiment of the present invention schematic diagram of degree of generation value d at random;
Fig. 4 is that embodiment of the present invention is selected d=3 different grouping schematic diagram in original;
Fig. 5 is the network G schematic diagram that embodiment of the present invention only has a node;
Fig. 6 is that embodiment of the present invention generates the schematic network structure after the 2nd coded identification;
If being embodiment of the present invention, Fig. 7 select one of the 3rd coded identification of generation to be grouped into S
2, cause the schematic network structure of Fourth Ring structure;
Fig. 8 is that embodiment of the present invention selects one of the 4th coded identification of generation to be grouped into S
4, cause the schematic network structure of six ring structures;
Fig. 9 is that embodiment of the present invention generates the schematic network structure forming after the 5th coded identification;
Figure 10 is the file transfer coding method flow chart of embodiment of the present invention based on complex network;
Figure 11 is first coding generative process flow chart of embodiment of the present invention;
Figure 12 is LT code and the comparison curves of traditional LT code aspect successfully decoded probability that embodiment of the present invention forms, and wherein packet count is 10
4.
Embodiment
Below in conjunction with accompanying drawing, embodiment of the present invention is described in further detail.
The file transfer coding method based on complex network adopting in present embodiment, as shown in figure 10.Concrete steps are as follows:
Step 1: add up file size to be transmitted, file waiting for transmission is divided into groups.
If data waiting for transmission are file, first want statistics file size, adopt in the present embodiment the fopen () function of C language to open file, adopt fseek () function pointer to be navigated to the position of file, with ftell () function, by the position of the last byte of identification, obtain the size of file.If data waiting for transmission are bit stream, directly treat transmission bit stream and divide into groups.Described grouping process is: file to be transmitted is divided into K group by bit-order, and wherein, K is positive integer, and K>10
3.Suppose that file to be transmitted can be converted into the bit stream of 2KB, can adopt said method that this bit stream is divided into K=2000 group, wherein each grouping comprises L=8 bit.As shown in Figure 1, wherein
First grouping: total 8bits, content is 01011000;
Second grouping: total 8bits, content is 10110011;
……
K=2000 grouping: total 8bits, content is 11100101.
In the present embodiment, the file of 2KB (16000bit) just in time can be divided into the grouping of 2000 8bit.Yet for some file, it just in time can not be divided into a plurality of groupings, that is, the data volume of last grouping after division is less than packeting average data volume.Now, need the end zero padding of an in the end grouping, the bit number that last grouping is comprised equals the average number of bits of each grouping.For example: suppose the file of a 15994bit to divide into groups, still bit stream is divided into 2000 groups, every group comprises L=8 Bit data, last grouping is by 8 of less thaies so, and we need to mend 0 thereafter.As shown in Figure 2, wherein
First grouping: total 8bits, content is 01011000;
Second grouping: total 8bits, content is 10110011;
……
K=2000 grouping: total 8bits, content is 11000000.Wherein last 6bit is filler.
Through said process, we can be 2000 groups by the Divide File of a 2KB.
Step 2: select a degree according to the value of packet count K and distribute;
Described degree distributes and can adopt sane orphan to distribute, also can adopt the degree of the coded identification of scale-free degree distribution and sane existing common LT codes such as scale-free degree distribution to distribute, and what user can be according to oneself need to select distribution function voluntarily.It is example that the sane orphan of take in present embodiment divides μ (d) cloth, and the deterministic process that degree of explanation distributes is as follows:
μ (d) is defined as follows,
Wherein
And
In formula, the coding symbol number that in each step that S is decode procedure, the remaining degree of expectation is 1:
In formula, β ∈ (0,1) is a free variable, when δ represents that receiving terminal is received K coded identification, and tolerable decoding error probability.
For all K values, sane orphan distributes and need to meet following formula:
Parameter beta=0.1 distributing as sane orphan is found in research, δ=1.0 o'clock, and the LT code distributing based on sane orphan has good performance.By β=0.1, δ=1.0, K=10000 substitution formula (2), can obtain S=92.10.And then can obtain K=10000 time, the degree of coded identification distributes.
In like manner, can adopt similar method, try to achieve the parameter in scale-free degree distribution and sane scale-free degree distribution, determine that the degree to the LT code of defining K value distributes.
Be grouped into example with K=2000 that in present embodiment, 2KB file is divided into, by β=0.1, δ=1.0, the value that K=2000 substitution formula (2) obtains S is 33.99.The value substitution formula (1) of S, the δ number that just can obtain dividing into groups, be that 2000 the degree based on sane orphan's calibration distributes.
Step 3: treating to generate new grouping in 2000 packet by packet basis of output size for the original of 2KB, coded identification, makes original transmit in network with the form of coded identification sequence; 2000 groupings of original are called as incoming symbol.This implementation process is simple in order to describe, and only having selected 5 incoming symbols in 2000 incoming symbols is the generative process that example is introduced its part coded identification.These 5 groups of incoming symbols are designated as respectively S
1, S
2, S
3, S
4, S
5, S wherein
1=01011000, S
2=10110011, S
3=11100101, S
4=01011000, S
5=10110010.
Step 3.1: generate first coded identification, as shown in figure 11.
Step 3.1.1: in the present embodiment, the degree of selecting according to step 2 distributes and generates at random a value d
1, the degree that work done in the manner of a certain author is first coded identification.From S
1, S
2, S
3, S
4, S
5in these 5 groups of incoming symbols, select d
1group is used for generating first coded identification.Value d
1selection course as follows:
The degree of selecting according to step 2 distributes, and the variable d in formula (1) gets respectively 1 ..., it is 1 that K different value can be spent ..., the probability that K coded identification is shared.Because carry out example with the cataloged procedure of 5 groups in 2000 groups in the present embodiment, therefore, we suppose K=5 herein, and limit d and get the integer between [1,5].First, 0-1.0 scope is divided into 5 sections according to this probability; Then use the rand () function in C language to produce a random number, judge this random number is in which segmentation of these 5 segmentations of 0-1.0, in which segmentation, the degree value of selected first coded identification of step 3.1.1 is just several so.As shown in Figure 3, the probability that degree of hypothesis is 1 is 0.1, the probability 0.4 that degree is 2, and the probability that degree is 3 is 0.3, and the probability that degree is 4 is 0.15, and the probability that degree is 5 is 0.05.Application cumulative distribution, if the size of random number between 0～0.1, the degree value of selecting is 1; If the size of random number is between 2～0.5, the degree value of selecting is 2; By that analogy.
Now, suppose to produce a random number that value is 0.51, as shown in Figure 3, known 0.51 is greater than 0.5, is less than 0.8, drops in the 3rd segmentation, therefore, plans the degree d of first coded identification
1assignment is 2.
3 different groupings of random selection in 5 groupings of original, the probability of arbitrary grouping of selection original is identical, and in the present embodiment, this probability is 1/5.
In present embodiment, adopt the rand () function in C language to produce 2 random numbers, be respectively 0.29,0.17.They drop in the 2nd segmentation and the 1st segmentation as shown in Figure 4, and therefore, select by this step 2 groupings that are used for generating the original of first coded identification are respectively the 1st grouping and the 2nd grouping.
Step 3.1.2: first grouping in present embodiment is designated as C
1, C
1value equal the XORs of 2 different original groupings that step 3.1.1 selects.That is, C
1=S
1⊕ S
2=11101011;
Step 3.1.3: the coded identification that the step 3.1.2 of take is generated is start node, generates a network configuration G who is comprised of a node, as shown in Figure 5.In the present embodiment, refer to coded identification C
1corresponding node V
0generating network structure G, this step is mainly for G is carried out to initialization, the content relating to mainly comprises node and the limit that adopts malloc () the function initialization G in C language.
Step 3.1.4: coded identification C
1be placed in channel and transmit;
Step 3.2: in the present embodiment, generate the 2nd coded identification;
Step 3.2.1: distribute according to the degree of step 2, by the degree d of random second coded identification of this degree distribution
2=1, at random from S
1, S
2, S
3, S
4, S
5in these 5 groupings, select 1 grouping, suppose this selected S of being grouped into
1.In network configuration G, add a new node V
1, because S
1to generate C
1a grouping, so, C
2and C
1the node V of correspondence on G
1, V
0between exist to connect limit, and the weight on limit is S
1, as shown in Figure 6.
Because the degree value of the 2nd coded identification is 1, and selected unique S that is grouped into
1, therefore, by step 3.2, generate second coded identification C
2=S
1=01011000.And by C
2be placed in channel and transmit.
Step 3.3: in the present embodiment, generate the 3rd coded identification;
Step 3.3.1: distribute and select at random the degree of the 3rd coded identification according to the degree of step 2, suppose that its value is for d
3=3, now we need to be at S
1, S
2, S
3, S
4, S
5in these 5 groupings, select 3 groupings that satisfy condition.Its process is as follows:
At random from S
1, S
2, S
3, S
4, S
5in these 5 groupings, select to generate the 1st grouping of the 3rd coded identification, suppose this selected S of being grouped into
1.In network configuration G, add a new node V
2, due to S
1to generate C
1and C
2the grouping of an original, in G, increase so a connected node V
2with node V
0weights be S
1company limit, an and connected node V
2with node V
1weights be S
1company limit.
At random from S
2, S
3, S
4, S
5in the grouping of these 4 originals, select a grouping to select grouping as the 2nd that generates the 3rd coded identification, suppose that it is S
2.S
2participation has generated coded identification C
1, while new node V
2with C
1corresponding node V
0having weights is S
1company limit, as shown in Figure 7.Yin Yin, selects S herein
2will cause Fourth Ring structure, it is not an optional grouping.
Random from remaining grouping S
3, S
4, S
5in these 3 groupings, select a grouping S
3, because it does not form Fourth Ring and six ring structures, so S
3can be used as an optional grouping that generates the 3rd coded identification.Again from remaining grouping S
4, S
5in random select a grouping, be assumed to be S
4if find S simultaneously
4be placed on and in complex network structures, also do not form Fourth Ring or six ring structures so S
4also can be used as an optional grouping that generates the 3rd coded identification.
Therefore, the 3rd coded identification C
3can there is S
1, S
3, S
4this three coded identification XORs and obtaining, i.e. C
3=S
1⊕ S
3⊕ S
4=01100101, by coded identification C
3be placed in channel and transmit.
Step 3.4: in the present embodiment, generate the 4th coded identification C
4;
Step 3.4.1: distribute and select at random the degree of the 4th coded identification according to the degree of step 2, suppose its value d
4=2, now we need to be at S
1, S
2, S
3, S
4, S
5in these 5 groupings, select 2 groupings that satisfy condition.Process is as follows:
At random from S
1, S
2, S
3, S
4, S
5in these 5 groupings, select to generate the 1st grouping of the 4th coded identification, suppose this selected S of being grouped into
2.In network configuration G, add a new node V
3, due to S
2to generate coded identification C
1the grouping of an original, therefore, in G, add a connected node V
3with node V
0weights be S
2company limit.
Random 4 grouping S from remainder
1, S
3, S
4, S
5grouping S of middle selection
4as the 2nd that generates the 4th coded identification, select grouping.Due in network configuration G, new node V
3connected node V
0, with grouping S
4there is coded identification C in what participate in to generate
3corresponding node V
2between exist to connect ，Qie Lian limit, limit weights be S
1≠ S
4, as shown in Figure 8.So, select S herein
4can cause six ring structures, that is, and S
4not to generate C
4an optional grouping.
Random 3 grouping S from remainder
1, S
3, S
5grouping of middle selection, supposes the selected S of being grouped into
5, that is, and S
5as the next one that generates the 4th coded identification, selected grouping.Due to S
5can not form Fourth Ring structure and six ring structures, therefore, S
5to generate coded identification C
4another optional grouping.
By step 3.4, generate in the present embodiment the 4th coded identification C
4=S
2⊕ S
5=00000001.Coded identification C
4putting into channel transmits.
Step 3.5: in the present embodiment, generate the 5th coded identification C
5;
According to the degree of step 2, distribute and select at random the degree of the 5th coded identification, suppose its value d
5=2, now need to be at S
1, S
2, S
3, S
4, S
5in these 5 groupings, select 2 groupings that satisfy condition.Process is as follows:
At random from S
1, S
2, S
3, S
4, S
5in these 5 groupings, select a grouping as an optional grouping that generates the 5th coded identification, suppose that this is grouped into S
3.In G, add a new node V
4, due to S
3participation has generated coded identification C
3, therefore, in G, add a connected node V
2and V
4weights be S
3limit.
Random 4 grouping S from remainder
1, S
2, S
4, S
5in these 4 groupings, select 1 grouping, be assumed to be S
2.Due in G with new node V
4the node V on the company of existence limit
2with S
2participate in the coded identification C generating
1the node V of correspondence in G
0the weights on the company of existence ，Qie limit, limit are S
1≠ S
2, therefore, if herein by S
2as optional grouping, will cause six ring structures.
Random 3 grouping S from remainder
1, S
4, S
5in these 3 groupings, select 1 grouping, be assumed to be S
1, as selecting grouping.Due to S
1participate in the coded identification C generating
3the node V of correspondence in G
2with new node V
4having weights is S
3company limit, therefore, S
1as optional grouping, will cause Fourth Ring structure.
Random 2 grouping S from remainder
4, S
51 grouping of middle selection, as generating coded identification C
5the next one selected grouping, be assumed to be S
4.Due to S
4participate in the coded identification C generating
3the node V of correspondence in G
2with new node V
4having weights is S
3company limit, therefore, if S
4as optional grouping, will cause Fourth Ring structure.
Random 1 grouping S from remainder
51 grouping of middle selection, this is grouped into S
5if select S herein
5can not cause Fourth Ring structure can not cause six ring structures yet, so, S
5to generate coded identification C
5an optional grouping.Due to S
5participation has generated coded identification C
4so,, in G, add a connected node V
3and V
4weights be S
5company limit.As shown in Figure 9.
By step 3.5, generate in the present embodiment the 5th coded identification C
5=S
3⊕ S
5=01010111, C
5be placed in channel and transmit.
Step 3.6: in the present embodiment, generate the 6th coded identification C
6;
According to the degree of step 2, distribute and select at random the degree of the 6th coded identification, suppose its value d
6=3.Now need to be at S
1, S
2, S
3, S
4, S
5in these 5 groupings, select 3 groupings that satisfy condition.Process is as follows:
At random from S
1, S
2, S
3, S
4, S
5in these 5 groupings, select a grouping as 1 the optional grouping that generates the 6th coded identification, suppose that this is grouped into S
2.In G, add a new node V
5, due to S
2participation has generated coded identification C
1and C
4, therefore, in G, add and connect V
5, V
0weights be S
2company limit be connected V
5, V
3weights be S
2company limit.
At random from S
1, S
3, S
4, S
5in these 4 groupings, select 1 grouping to select grouping as the 2nd that generates the 6th coded identification, suppose that this is grouped into S
1.Due in network configuration G, S
1participate in the coded identification C generating
1the node V of correspondence in G
0with new node V
5having weights is S
2company limit, therefore, if select S herein
1will cause Fourth Ring structure.
At random from 3 remaining grouping S
3, S
4, S
51 grouping of middle selection is as generating C
6the 3rd select grouping, suppose that this is grouped into S
3.Due in G with new node V
5the node V on the company of existence limit
0with S
3participate in the coded identification C generating
3the node V of correspondence in G
2between exist to connect Qie Lian limit, limit weights be S
1≠ S
3, therefore, S herein
3it not an optional grouping.
At random from 2 remaining grouping S
4, S
5grouping of middle selection is as generating C
6the 4th select grouping, be assumed to be S
4.Due in G with new node V
5the node V on the company of existence limit
0with S
4participate in the coded identification C generating
3the node V of correspondence in G
2between exist to connect Qie Lian limit, limit weights be S
1≠ S
4, therefore, if S herein
4as optional grouping, will form six ring structures.
Analyze last grouping S
5, due to S
5participate in the coded identification C generating
4the node V of correspondence in G
3with new node V
5having weights is S
2company limit, therefore, S
5as optional grouping, will cause Fourth Ring structure.
In summary, if generate the 6th selected first of coded identification, be grouped into S
2, so no matter to select S
1, S
3, S
4or S
5becate structure is caused in capital.Now, from removing S
2outer S
1, S
3, S
4, S
52 groupings of these 4 the random selections of grouping, are assumed to be S
1, S
4, as generating coded identification C
6two other optional grouping.
By step 3.6, obtain the 6th coded identification C
6=S
2⊕ S
1⊕ S
4=10110011, C6 is placed in channel and is transmitted.
By said process, generated 6 coded identifications, they are all 5 generations of dividing into groups in original.In the present embodiment, for the original that comprises 2000 groupings, also can adopt identical method, generate coded identification sequence.At receiving terminal, according to the relation of a plurality of groupings of the value of coded identification and coded identification and original, original is recovered.
The present invention can directly apply to the transmittings of Volume data in wireless channel such as video, image, for the ease of verifying feasibility and the advantage of described method, only from the encoding scheme of successfully decoded probability angle and traditional LT code, be analyzed herein.
It is 10 that 1000 groups of incoming symbol numbers are take in the present invention
4, block length is that the sane LT code without characteristics of scale of having of 1 bit and the LT code based on sane orphan's distribution are that example is verified validity of the present invention, the LT code that wherein participates in contrast comprises;
(1) RD-SF (cycle-6priority): parameter is selected P
1the sane LT code without characteristics of scale of having of Fourth Ring and six ring structures is reduced in=0.1, λ=1.9 as far as possible;
(2) RD-SF (cycle-4priority): parameter is selected P
1fourth Ring is reduced in=0.1, λ=1.9 as far as possible, does not consider the sane LT code without characteristics of scale of having of six rings;
(3) RD-SF (random): parameter is selected P
1=0.1, λ=1.9, that does not consider Fourth Ring and six rings has a sane LT code without characteristics of scale;
(4) RSD (cycle-6priority): parameter is selected c=0.1, the LT code distributing based on sane orphan of Fourth Ring and six ring structures is reduced in δ=1 as far as possible;
(5) RSD (cycle-4priority): parameter selection c=0.1, Fourth Ring is reduced in δ=1 as far as possible, does not consider the LT code distributing based on sane orphan of six rings;
(6) RSD (random): parameter is selected c=0.1, δ=1, does not consider the LT code distributing based on sane orphan that Fourth Ring and six is encircled.
The degree that this invention does not change coded identification distributes, therefore, average XOR number of times and original coding scheme that generating coded identification needs are more or less the same, but due to Fourth Ring and six ring probabilities of occurrence reductions, the coding symbol number that recovering identical incoming symbol needs will decrease, and the average XOR number of times of this scheme decode procedure is reduced.As shown in table 1 is the relativity of redundancy factor in different encoding schemes, and redundancy factor is lower, and the coding symbol number that shows to recover original incoming symbol needs is fewer, and required decoding delay is lower:
The LT code of table 1 the present embodiment structure and the Performance Ratio of traditional LT code are
Analytical table 1 is known, the average redundancy factor of the LT code of application six ring priority algorithms is less than the average redundancy factor of the LT of application Fourth Ring priority algorithm, they are all less than the average redundancy factor of the LT code that does not carry out becate optimization, and the coding symbol number that the LT code that therefore adopts becate to evade algorithm recovers the incoming symbol needs of similar number is less than the LT code that does not carry out becate optimization.
Because the degree of each coded identification is independent generation, as long as and there is the sane LT code without characteristics of scale and the LT code that distributes based on desirable orphan receives enough coded identifications and just can guarantee that recipient can Complete Decoding, irrelevant with the channel condition of transmission data, therefore, only consider LT code that the becate of application based on Complex Networks Theory the evade algorithm successfully decoded probability under desirable channel condition herein.
Figure 12 is that incoming symbol number is 10
4, construct respectively 1000 groups of above-mentioned six kinds of LT codes, obtain their successfully decoded probability and the relation between redundancy factor.
Analysis Figure 12 is known, under the identical prerequisite of redundancy factor, application becate is evaded the successfully decoded probability with the sane LT code without characteristics of scale that algorithm generates higher than not considering the sane LT code without characteristics of scale of having of becate structure and other three kinds of LT codes that distribute based on sane orphan.And application becate is evaded the LT code distributing based on sane orphan that algorithm generates, receiving terminal receives the successfully decoded probability of coded identification N=(1+ α) K of similar number higher than the LT code distributing based on sane orphan of not considering becate structure.
Above-mentioned research shows, application becate is evaded the LT code of algorithm design and can be compiled under the prerequisite of (translating) code efficiency guaranteeing, successfully decoded probability is greatly improved.
Although more than described the specific embodiment of the present invention, the those skilled in the art in this area should be appreciated that these only illustrate, and can make various changes or modifications to these execution modes, and not deviate from principle of the present invention and essence.Scope of the present invention is only limited by appended claims.
Claims (2)
1. the file transfer coding method based on complex network, is characterized in that: comprise the following steps:
Step 1: add up file size to be transmitted, file waiting for transmission is divided into groups;
Step 2: select a degree according to k value and distribute;
Step 3: generate new grouping in a plurality of packet by packet basis of original to be exported, coded identification, makes original transmit in network with the form of coded identification sequence;
Step 3.1: generate first coded identification;
Step 3.1.1: the degree of being selected by step 2 distributes and generates at random a value d, d different grouping of random selection in former output file, wherein, d is natural number;
Step 3.1.2: XOR is carried out in this d different grouping, forms a new grouping, this grouping is first coded identification for generating just;
Step 3.1.3: the coded identification that the step 3.1.2 of take is generated is start node, generates a network configuration being comprised of a node;
Step 3.1.4: this coded identification is placed in channel and is transmitted;
Step 3.2: generate next coded identification;
Step 3.2.1: distribute according to the degree of step 2, by this degree, distribute and generate at random another positive integer d ', from k grouping of original, select a grouping at random, the grouping that this is grouped into first original selected that generates next coded identification selection is also first the optional grouping that generates next coded identification.In network configuration, add a new node.If a grouping of the original of selecting be generate other certain or some there is a grouping of coded identification, so, one of having between corresponding node in network configuration that the weights that connect ，Qie limit, limit are the original selected of this new node and above-mentioned certain or some already present coded identification is grouped in the sequence number in the grouping of all k original; If d '=1, performs step 3.2.2; Otherwise execution step 3.2.3;
Grouping complete zero bit sequence isometric with it of the original that step 3.2.2: step 3.2.1 selects carries out XOR, form a new grouping, the next coded identification that this is grouped into generation, is placed on this coded identification in channel and transmits, and to step 3.3, continues to carry out;
Step 3.2.3: except the grouping of having selected for the next coded identification of generation, select another grouping of original in remaining grouping in all groupings of original, this is grouped into another grouping selected that generates next coded identification;
Step 3.2.4: find out that all another groupings in step 3.2.3 participate in to generate there is coded identification and in network configuration corresponding node, check that whether corresponding with these node of new node in step 3.2.1 exists and connect limit.If the company of existence limit, does not show that another grouping of selecting can not cause the Fourth Ring structure of LT code, continues to carry out from step 3.2.5; If the company of existence limit, shows that this another grouping will cause the Fourth Ring structure of LT code, this another grouping is not the optional grouping that generates next coded identification.Now, if still exist for, generate the not grouping of the original of selected mistake of next coded identification, get back to step 3.2.3 and continue to carry out, otherwise continue to carry out to step 3.2.9;
Step 3.2.5: find out that all another groupings in step 3.2.3 participate in to generate there is coded identification and in network configuration corresponding node, check in network configuration with new node and exist the node that connects limit whether to have even limit with the coded identification corresponding node in network configuration that exists that another grouping participates in generating.If the weights on the company of existence limit or the company of existence limit and limit do not equal the sequence number in another all groupings that are grouped in original, show that so another grouping can not cause six ring structures of LT code, this another be grouped into the optional grouping that generates next coded identification, to step 3.2.6, continue to carry out; If the company of existence limit, and the weights that connect limit are not equal to the sequence number in another grouping that is grouped in original, show to exist six ring structures, i.e. this another grouping will cause six ring structures of LT code, under this another grouping, is the optional grouping that generates next coded identification.Now, if still there is the not grouping of the original of selected mistake, get back to step 3.2.3 and continue to carry out; Otherwise, from step 3.2.9, continue to carry out;
Step 3.2.6: that finds out that all another groupings in step 3.2.3 participate in to generate exists coded identification and corresponding node in network configuration, in network configuration, add the company limit of these corresponding nodes and new node, the weights on limit are grouped in the sequence number in all groupings of original for another;
Step 3.2.7: if equaling d ', optional grouping number continues to carry out from step 3.2.8, individual if optional grouping number is less than d ', from step 3.2.9, continue to carry out;
Step 3.2.8: the individual grouping of this d ' is carried out XOR and obtained a new grouping, this is new is grouped into next coded identification.This coded identification is placed in channel and is transmitted; Execution step 3.3;
Step 3.2.9: if still there is the not grouping of the original of selected mistake, repeating step 3.2.3; If there is no the not grouping of selected mistake, beyond optional grouping, select arbitrarily the grouping of original so, until selecteed packet count is d '.The individual grouping of this d ' is carried out XOR and is generated a new grouping, i.e. next coded identification, and this coded identification is placed in channel and is transmitted;
Step 3.3: the process of repeated execution of steps 3.2, generates remaining coded identification, until transmitting terminal receives the feedback information that file that receiving terminal returns to has correctly recovered.
2. the file transfer coding method based on complex network according to claim 1, is characterized in that: described grouping process is: file to be transmitted is divided into k group by bit-order, and wherein, k is positive integer, and k>10
3if the data volume in last grouping is less than average amount, by the remaining bits position zero padding in this grouping.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310507723.6A CN103532674B (en) | 2013-10-23 | 2013-10-23 | The transmission coded method of a kind of file based on complex network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310507723.6A CN103532674B (en) | 2013-10-23 | 2013-10-23 | The transmission coded method of a kind of file based on complex network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103532674A true CN103532674A (en) | 2014-01-22 |
| CN103532674B CN103532674B (en) | 2016-08-17 |
Family
ID=49934391
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310507723.6A Active CN103532674B (en) | 2013-10-23 | 2013-10-23 | The transmission coded method of a kind of file based on complex network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103532674B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103929275A (en) * | 2014-04-30 | 2014-07-16 | 电子科技大学 | Distributed SF-LT code coding method with scale-free characteristic |
| CN107612897A (en) * | 2017-09-07 | 2018-01-19 | 唐冬香 | A kind of data transmission method |
| CN108259930A (en) * | 2018-01-16 | 2018-07-06 | 深圳市力沃信息科技有限公司 | The transfer control method and system of a kind of electronics class board |
| CN108267950A (en) * | 2018-01-19 | 2018-07-10 | 东北大学 | Simple fractional order complex network external mix synchronous method |
| CN115333673A (en) * | 2022-07-29 | 2022-11-11 | 南京信息工程大学 | Block coding transmission method based on ILT in block chain network |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011139122A2 (en) * | 2010-05-07 | 2011-11-10 | Samsung Electronics Co., Ltd. | Apparatus and method for channel coding in a communication system |
| CN102324998A (en) * | 2011-05-11 | 2012-01-18 | 浙江大学 | Raptor Codes encoding/decoding method suitable for medium/short code lengths of additive white Gaussian noise channel |
| CN103023937A (en) * | 2011-09-26 | 2013-04-03 | 北大方正集团有限公司 | Method and system for distributing network files |
| CN103078707A (en) * | 2013-01-03 | 2013-05-01 | 北京理工大学 | File transmission method in deep space communication |
-
2013
- 2013-10-23 CN CN201310507723.6A patent/CN103532674B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2011139122A2 (en) * | 2010-05-07 | 2011-11-10 | Samsung Electronics Co., Ltd. | Apparatus and method for channel coding in a communication system |
| CN102324998A (en) * | 2011-05-11 | 2012-01-18 | 浙江大学 | Raptor Codes encoding/decoding method suitable for medium/short code lengths of additive white Gaussian noise channel |
| CN103023937A (en) * | 2011-09-26 | 2013-04-03 | 北大方正集团有限公司 | Method and system for distributing network files |
| CN103078707A (en) * | 2013-01-03 | 2013-05-01 | 北京理工大学 | File transmission method in deep space communication |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103929275A (en) * | 2014-04-30 | 2014-07-16 | 电子科技大学 | Distributed SF-LT code coding method with scale-free characteristic |
| CN103929275B (en) * | 2014-04-30 | 2017-10-03 | 电子科技大学 | Distributed SF-LT code coding method with scale-free characteristic |
| CN107612897A (en) * | 2017-09-07 | 2018-01-19 | 唐冬香 | A kind of data transmission method |
| CN108259930A (en) * | 2018-01-16 | 2018-07-06 | 深圳市力沃信息科技有限公司 | The transfer control method and system of a kind of electronics class board |
| CN108259930B (en) * | 2018-01-16 | 2020-08-14 | 深圳市力沃信息科技有限公司 | Transmission control method and system for electronic class board |
| CN108267950A (en) * | 2018-01-19 | 2018-07-10 | 东北大学 | Simple fractional order complex network external mix synchronous method |
| CN115333673A (en) * | 2022-07-29 | 2022-11-11 | 南京信息工程大学 | Block coding transmission method based on ILT in block chain network |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103532674B (en) | 2016-08-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1620760B (en) | Multi-stage code generator and decoder for communication systems | |
| JP5485302B2 (en) | File download and streaming system | |
| CN1836394B (en) | Apparatus and method for coding/decoding block ldpc codes in a mobile communication system | |
| CN102164026B (en) | Fountain code compiling method based on deep space communication environment | |
| CN101425871B (en) | Multi-element error correcting code transmitting and receiving apparatus, data communication system and related method | |
| CN103532674A (en) | File transmission encoding method based on complex network | |
| CN102770911A (en) | Method for encoding information object and encoder using the same | |
| CN106209302B (en) | Data transmission processing method and device | |
| CN108282265A (en) | Error correction/encoding method, device, equipment and computer readable storage medium | |
| JP2015520990A (en) | Packet transmitting / receiving apparatus and method in broadcasting and communication system | |
| CN105991227A (en) | Data encoding method and apparatus | |
| CN106059712A (en) | High bit error arbitrary bit rate convolutional code coding parameter blind identification method | |
| CN111447160A (en) | Signal detection method of large-scale MIMO passive random access based on neural network | |
| CN102938682A (en) | Methods and apparatus for providing linear erasure codes | |
| CN106877980A (en) | Mixing Sparse Code multiple access method | |
| CN104378122B (en) | A kind of Compilation Method of variable-length Turbo code | |
| CN103152652A (en) | Video frame data network transmission method based on Cauchy reed-solomon (RS) code | |
| CN107437976B (en) | Data processing method and equipment | |
| Larsson | Multicast multiuser ARQ | |
| CN102195743B (en) | Coding scheme of dynamic real-time fountain code | |
| CN101645753B (en) | Coding method without rate codes | |
| CN109150397B (en) | A kind of method and device constructing polarization code sequence | |
| CN104219032A (en) | Minimum packet loss retransmission method in multicast scene in wireless local area network | |
| CN102820952A (en) | Method for optimizing throughput of wireless network system based on no-rate encoding | |
| CN101127581B (en) | A mapping and reverse mapping method and device for LDPC coding modulation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |