KR101204543B1 - Method and apparatus for chunk selection in p2p streaming service - Google Patents

Abstract

A method and apparatus for selecting a chunk in a P2P streaming service is provided. One or more chunks that make up the file to be played by streaming are downloaded. For each of the one or more chunks, the priority of the chunk is calculated, and the chunk to request transmission is selected based on the calculated priority. The priority is calculated based on the scarcity of the chunk and the location of the chunk within one or more chunks.

Description

METHOD AND APPARATUS FOR CHUNK SELECTION IN P2P STREAMING SERVICE}

The following embodiments are directed to a method and apparatus for selecting a chunk in a P2P streaming service.

Recently, as a system for playing a video through the Internet has received attention, various methods for improving the performance of such a system have been developed.

Video files generally have a large size.

In the case of tree-based systems, a large burden is placed on the server by a video file having a large size. Therefore, such a system has a problem that it is difficult to use by many users.

For this reason, the focus is on mesh-based systems, which are distributed systems that put less strain on the server.

In addition, mesh-based streaming services show simpler and more robust performance in environments where users frequently enter and leave.

In mesh-based systems, swarming protocols such as BitTorrent are mostly used. In this swarm protocol, users share files with each other by exchanging pieces of files with other users.

Chunk scheduling refers to a policy of which file fragments (ie blocks or chunks) should be exchanged first (first) during file sharing.

Various policies on chunking scheduling have been developed, and such chunking scheduling provides streaming services using peer-to-peer (P2P) such as video-on-demand (VOD) and live streaming. Can be used in any system (eg, Internet Protocol TeleVision (IP-TV)).

An embodiment of the present invention may provide a chunk selection apparatus and method for providing a P2P video streaming service.

One embodiment of the present invention can provide a chunk selection apparatus and method that takes into account the sparsity of the block and the playback deadline together.

According to one aspect of the present invention, a method for downloading one or more chunks included in a file in a P2P network connected to one or more neighboring nodes, the method comprising: calculating a priority of the chunk for each of the one or more chunks, the priority Selecting a chunk to request transmission based on a rank, requesting transmission of the selected chunk to one or more nodes of the neighboring nodes, and one or more nodes of one or more nodes requested to transmit the selected chunk. Downloading the chunk that requested the transmission from the second value, wherein the priority is based on a first value calculated based on the scarcity of the chunk and a second value calculated based on a position in the one or more chunks of the chunk. A file download method is provided, which is calculated based on this.

The file may be a file being played or may be a file being transmitted through streaming for the playback.

The file downloading method may further include selecting the one or more chunks of the chunks constituting the file as an interval for downloading.

The one or more chunks may be chunks after the location of the playback in the file, may be chunks that have not been downloaded, and may be chunks within a specific range from the location of the playback.

The priority may be a weighted sum of the first value and the second value.

The weight may be calculated based on the ratio of chunks that may be provided by nodes below a certain number of the one or more chunks.

According to another aspect of the present invention, in a P2P file sharing unit for downloading one or more chunks included in a file from a P2P network connected with one or more neighbor nodes, the priority of calculating the priority of the chunks for each of the one or more chunks is high. A transfer request chunk selecting unit for selecting a chunk to request transmission based on the priority, wherein the priority calculating unit calculates a first value based on the scarcity of the chunk, and the at least one of the chunks. A P2P file sharing unit is provided that calculates a second value based on a location in the chunks and calculates the priority based on the first value and the second value.

The P2P file sharing unit may further include a window selector for selecting the one or more chunks as the section for the download among the chunks constituting the file.

The scarcity of the chunk may represent the ratio of the chunk to all the chunks that are spread to the neighbor nodes.

A relatively preceding chunk of the one or more chunks may have the second value that is higher than a relatively subsequent chunk.

The priority may be a weighted sum of the first value and the second value.

The weight may be calculated based on the number of chunks of the one or more chunks spread below a predefined value in the network.

The priority calculator may increase the weight for the first value as the number of chunks spread below a predefined value in the network among the one or more chunks is smaller.

The priority calculator may increase the weight for the second value as the number of chunks spread below a predefined value in the network among the one or more chunks increases.

According to another aspect of the present invention, in a node for downloading one or more chunks included in a file through a P2P network connected to one or more neighboring nodes, calculating the priority of the chunk for each of the one or more chunks, A control unit for selecting a chunk to request transmission based on the priority, one or more nodes of one or more nodes requesting transmission of the selected chunk to one or more nodes of the neighboring nodes, and requesting transmission of the selected chunk; And a storage unit for storing the downloaded chunk as part of the file, and a storage unit for downloading the chunk requesting the transmission from the memory, wherein the priority is the first value calculated based on the scarcity of the chunk and the chunk of the chunk. Based on a second value calculated based on a location within one or more chunks On top of that, a P2P network node is provided.

Provided are a chunk selection apparatus and method for providing a P2P video streaming service.

A chunk selection apparatus and method are provided that take into account the sparsity of the block and the playback deadline together.

1 illustrates a BiToS chunk selection method according to an embodiment of the present invention.
2 compares the chunk selection method according to an embodiment of the present invention.
3 illustrates an ENR chunk selection method according to an embodiment of the present invention.
4 is a flowchart of a file download method using ENR technology according to an embodiment of the present invention.
5 is a structural diagram of a node of a P2P network according to an embodiment of the present invention.
6 compares an effective download ratio (EDR) between chunk selection methods according to an embodiment of the present invention.
7 compares the miss ratio (MR) between the chunk selection methods according to an embodiment of the present invention.
8 compares a startup delay (SD) between chunk selection methods according to an embodiment of the present invention.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited to or limited by the embodiments. Like reference symbols in the drawings denote like elements.

1 illustrates a BiToS chunk selection method according to an embodiment of the present invention.

Among the file sharing (sharing) technology using P2P is BitTorrent technology. BitTorrent Streaming (BiToS) is one of the techniques of applying BitTorrent technology to VOD streaming.

BiToS uses "Rarest First" of existing chunk selection techniques used in BitTorrent.

The sparsity first technique is to preferentially receive (or obtain) the least spread chunks on a network overlay. Sparsity-first approach allows files to be shared more efficiently across the entire network.

The sparse priority of BitTorrent simply schedules chunks (ie, the transmission of the chunks) based solely on the scarcity. Therefore, BitTorrent's sparsity priority cannot be used properly in streaming services where the head of a video file must first be received.

The video file that is the subject of transmission is separated into chunks. Each chunk has a sequence of numbers in order within the video file. Chunks are largely classified into four types. Each of the four species is as follows.

1) downloaded piece 112: The chunk has already been downloaded by the client. The downloaded portion is moved to the player buffer and used by the player to output the video.

2) currently downloading piece 114: A chunk being sent from the server to the client.

3) not requested or not downloaded piece 116: A chunk that is the subject of a transfer but has not yet been sent.

4) Missed player deadline piece 118: A missing player deadline piece 118 is a part that is not sent because the deadline requested by the player has already passed. For example, when the player is reproducing a portion corresponding to the fourth chunk 194 or the fifth chunk 195, the third chunk 193 does not need to be transmitted because it is already excluded from the object of reproduction.

The portion of the video file in the server other than the fragment 112 that has already been downloaded or the fragment 118 that misses the deadline of the player is the portion 120 to be downloaded. That is, the portion 120 to be downloaded includes the fragment 114 currently being downloaded and the fragment 116 not requested or not downloaded.

Received pieces 150 in the client include downloaded pieces 112, unsolicited or undownloaded pieces 116, and pieces that missed the deadline of the player, and the video file played. As it moves to player buffer 160.

BiToS divides the portion 120, which is the object of download of the video file 110 provided by VoD, into two sections 130 and 140 so that BitTorrent is used more efficiently in streaming.

BiToS enables a streaming service to be provided by increasing the weight of the leading portion 140 of the divided two sections 130 and 140.

BiToS bundles some chunks into a high priority set 140 in order close to the point at which they are played by the player, and the remaining chunks into a low priority set 130. Tie.

BiToS applies different probabilities to each of the two sets 130 and 140 in selecting the chunks. For example, the probability that the higher priority set 140 will be selected (ie, the probability that one of the chunks in the higher priority set will be selected) is 80%, and the probability that the remaining set 130 will be selected is 20%. %to be.

In BiToS, if one set 130 or 140 of two sets 130 and 140 is selected, a chunk is selected according to the sparsity priority within the selected set.

Thus, BiToS is designed to be suitable for streaming services that preferentially require the leading part of a video file.

2 compares the chunk selection method according to an embodiment of the present invention.

In the video file 110 that is the object of transmission, the part that is not downloaded among the parts after the playback position (that is, the chunk) is the object of the transmission.

As described above, BiToS selects the high priority portion 140 and the low priority portion 130 of the portion 120 to be downloaded. Within each of the portions 130 and 140, sparsity priority is applied.

G2G and beta selection are ways to improve the performance of VoD streaming over BiToS (i.e., to increase the probability that a particular chunk will be received before it is played).

The G2G method is similar to BiToS as a whole. However, the G2G method selects the high priority portion 210, the middle priority portion 220, and the low priority portion 230 from the portion 120 to be downloaded in the order close to the playback position.

Between the chunks in the high priority portion 210 an earliest first (a method of selecting from the preceding chunks in order) is applied. In other words, the preceding chunk is preferentially selected.

The sparsity priority is applied between the chunks in middle priority portion 220.

The Beta Selection method is similar to BiToS as a whole.

Leading priority is applied between the chunks in high priority portion 240.

Beta random is applied between the chunks in the low priority portion 250.

Beta random is a method that uses a beta distribution in which the probability that chunks are selected decreases from 1 to 0 constantly. That is, the probability that the chunk located relatively ahead of the chunks is selected is higher than the probability that the chunk located relatively later is selected.

BiToS separates sets 130 and 140. However, within each of the separated sets 130 and 140, a sparsity first technique is used. Therefore, chunk misses (if they are not received before the chunks are played) frequently occur during VoD streaming.

In G2G and beta selection, precedence techniques are used within a high priority set. The precedence first method selects the chunks in order. Therefore, flexible chunk selection cannot be performed by the prioritized first technique. In addition, in both methods, since two or more sets are selected in the portion 120 to be downloaded, bandwidth cannot be efficiently used.

3 illustrates an ENR chunk selection method according to an embodiment of the present invention.

This embodiment enables more flexible chunk selection by complementing the problems of the techniques mentioned with reference to FIG. In this embodiment, by flexibly selecting the chunk, the performance of the VoD streaming service can be improved, and the performance of the entire network can be improved.

In the ENR chunk selection method, a window 310 for downloading among files is selected.

As shown, the section 310 for download includes only chunks after the location 320 being played, excluding the chunks 330 already downloaded. In addition, the section 310 for downloading may exclude the chunks 340 that are out of a predetermined distance from the position 320 being played.

That is, in the other methods described above, the ENR chunk selection method selects a single set as the interval 310 for download, compared to when two or more sets 130 and 140 are selected.

One or more chunks in the download section 310 are subject to transmission, and priority is calculated by a biased random method and a beta selection method for each of the one or more chunks.

The biased random scheme means a method of determining the probability that the chunk is selected based on the scarcity of the chunk in the network.

The beta selection method refers to a method of determining a probability of selecting the chunk based on the relative position of the chunk in the interval for downloading.

That is, the priority of the chunks may be calculated based on the scarcity in the chunk's network and the (relative) location within the interval for downloading the chunks.

In the following, the method of calculating the priority of the chunks according to the ENR chunk selection method is described in detail.

4 is a flowchart of a file download method using ENR technology according to an embodiment of the present invention.

In this embodiment, the node downloading the file is connected with one or more neighboring nodes via a network. The network may be a P2P network.

A file consists of one or more chunks. Neighbor nodes may provide all or part of the chunks.

A node receives a file by downloading one or more chunks contained in the file from one or more neighboring nodes.

The file may be a video file or a file transmitted through streaming. That is, the node may play a file being transmitted while receiving a file from one or more neighboring nodes through streaming.

In step S410, one or more chunks of the chunks constituting the file are selected as the section 310 for downloading.

The one or more chunks selected are the chunks after the playback position in the file being transmitted. Also, the selected one or more chunks are chunks that have not yet been downloaded and are chunks that are within a predetermined range from the playback position.

In step S420, the priority of the chunk is calculated for each of the selected one or more chunks.

The priority may be calculated according to the ENR chunk selection method described above.

The priority is based on a first value calculated based on the scarcity of the chunk and a second value calculated based on the location of the chunk within one or more chunks (ie, the chunks in the interval 310 for download). Can be calculated. That is, the first value is a value calculated according to the biased random scheme, and the second value is a value calculated by the beta selection scheme.

The first value may be calculated based on Equation 1 below.

Where b _i represents the i th chunk, r (b _i ) is a sparsity function representing the number of neighbor nodes including chunk b _i of the neighbor nodes, and k is one or more chunks (ie, for download) Number of chunks in section 310 or the total number of chunks constituting the file.

That is, the sparsity of a particular chunk may indicate the ratio of that particular chunk to all chunks that are spread to neighboring nodes.

The second value may be calculated based on Equation 2 below.

Here, d represents a beta distribution function, and d (b _i ) is a value of beta distribution function of b _i .

The beta distribution function gives the probability that the chunk is selected according to the location of the chunk in the section 310 for download. By the beta distribution function, the probability of selecting a particular chunk is constantly reduced from 1 to 0.

That is, according to the beta distribution, among the chunks in the section 310 for download, the chunk located relatively ahead (that is, the chunk corresponding to the front part of the file) has a greater probability of being selected than the chunk located relatively later. (Ie, higher second value).

The priority may be a weighted sum of the first value and the second value.

That is, the priority may be calculated based on Equation 3 below.

Here, p is a weight given to the first value and (1-p) is a weight given to the second value.

The weight may be calculated based on the sparsity of each of the one or more chunks.

The scarcity of a particular chunk may be the number of neighboring nodes that can provide the particular chunk among neighboring nodes serving files in the network.

That is, the weight may be calculated based on the ratio of chunks that may be provided by up to a predefined number of nodes of one or more chunks (ie, the chunks in interval 310 for download). .

For example, there are five chunks (first to fifth chunks) in the interval 310 for download. The predefined number for calculating the weight is three. One of the five chunks may be provided by one neighbor node. Two of the five chunks may be provided by three neighboring nodes, and the remaining two of the five chunks may be provided by five neighboring nodes. Thus, there are three chunks that can be provided by nodes of three or less (predefined number) of the five chunks. Therefore, the weight p is 3/5.

That is, the weights assigned to each of the first value and the second value may be calculated based on Equation 4 below.

Here, t is a predefined value, r (b _i ) is the number of nodes that can provide b _i of the neighbor nodes, W is the number of the one or more chunks (that is, the interval 310 for downloading) Size).

That is, ρ is a value representing the number of chunks spread in t or less in the network among the entire windows (that is, the section 310 for downloading) based on the predefined value t.

Therefore, the more chunks spread below t, the larger the value of p. As the value of ρ increases, the proportion of scarcity increases in the priority according to equation (3). Thus, rarer chunks spread better within the network.

The fewer chunks spread below t, the smaller the value of p. As the value of p decreases, the weight of the beta distribution increases in the priority according to equation (3). Therefore, the preceding chunk is preferentially selected in the window, and the above selection improves the performance of the VoD streaming service. That is, the probability that the chunk to be reproduced is missed due to the above selection is reduced.

In step S430, a chunk to request transmission is selected based on the calculated priority.

In descending order of priority, one or more predefined numbers of chunks may be selected as chunks to request transmission.

In addition, in descending order of priority, the chunks to request transmission may be sequentially selected.

In step S440, transmission of the selected chunk is requested to neighboring nodes. For example, a message indicating to send the selected chunk is sent to neighboring nodes.

Some of the neighbor nodes that can provide the chunk to be selected (eg, those identified as capable of providing the selected chunk) may be requested to send the chunk to be selected.

In step S450, the chunk requested to be downloaded is downloaded from one or more nodes of the one or more nodes that are requested to send the chunk.

Technical contents according to an embodiment of the present invention described above with reference to FIGS. 1 to 3 may be applied to the present embodiment as it is. Therefore, more detailed description will be omitted below.

5 is a structural diagram of a node of a P2P network according to an embodiment of the present invention.

The node 500 may include a transceiver 510, a controller 520, and a storage 530.

The storage unit 520 may include a P2P file sharing unit 540.

The P2P file sharing unit 540 may include a window selector 550, a priority calculator 560, and a transfer request chunk selector 570.

The node 500 may download one or more chunks included in the file via a P2P network connected with one or more neighboring nodes. The download may be performed by the P2P file sharing unit 540.

The file containing one or more chunks to be downloaded may be a file being played by node 500 and may be a file being transmitted via streaming for playback.

The window selector 550 may select one or more chunks as a section 310 for downloading among the chunks constituting the file.

The one or more chunks selected are the chunks after the location being played in the file and are not yet downloaded. Also, the selected one or more chunks may be chunks within a predefined range from the location being played.

The priority calculator 560 calculates the priority of the chunks for each of the selected one or more chunks.

The priority calculator may calculate a first value based on the sparseness of the chunk, calculate a second value based on a location within the selected one or more chunks of the chunk, and determine the chunk based on the first value and the second value. The priority can be calculated.

The priority calculator may calculate the priority of the first value, the second value, and the chunk according to Equations 1 to 5 described above.

The scarcity of a chunk may indicate the ratio of the chunk to all chunks that are spread to the neighbor nodes.

The relatively preceding chunk of the selected one or more chunks may have a second value that is higher than the later located chunk.

The priority may be a weighted sum of the first value and the second value.

The weight may be calculated based on the number of chunks spread out below a predefined value in the network of the selected one or more chunks.

The priority calculator 560 may increase the weight for the first value as the number of chunks spread below a predefined value in the network among the selected one or more chunks is smaller.

The priority calculator 560 may increase the weight for the second value as the number of chunks spread below a predefined value in the network among one or more selected chunks increases.

The transmission request chunk selector 570 selects a chunk to request transmission from among one or more chunks based on the calculated priority.

The transceiver 510 requests transmission of the selected chunk to one or more nodes of the neighboring nodes, and downloads the chunk requested to be transmitted from one or more nodes among the one or more nodes for which the transmission of the selected chunk is requested.

The storage unit 530 stores the downloaded chunk as part of the file.

Technical contents according to an embodiment of the present invention described above with reference to FIGS. 1 to 4 may be applied to the present embodiment as it is. Therefore, more detailed description will be omitted below.

6 compares an effective download ratio (EDR) between chunk selection methods according to an embodiment of the present invention.

The efficient download ratio is the most important of the performances of VoD streaming and represents the ratio between the number of chunks played per second and the number of chunks downloaded per second (does not include the number of chunks to be duplicated). That is, the EDR may be (the number of chunks downloaded per second) / (the number of chunks reproduced per second).

In general, the more chunks to download than the chunks to play back, the better the performance of the VoD streaming.

Graph 610 is a graph comparing the EDR value of the ENR with the EDR value of other technologies (BiToS, G2G and Beta Selection). An EDR value of 1.0 indicates that the number of chunks to play and the number of downloaded chunks are the same.

According to graph 610, it can be seen that ENR downloads a larger amount of chunks compared to other techniques. This indicates that more chunks can be received by solving the problem that the ENR wastes bandwidth due to dividing the set previously described.

As in the conventional method, when the interval is divided into sets and different bandwidths are used according to the interval, 1) a problem arises by dividing the bandwidth, and 2) a high priority is given to the high priority set despite the small size. The allocation of specific bandwidth causes a problem that bandwidth is not used efficiently.

Since ENR does not divide sets, bandwidth can be used more efficiently. Thus, the ENR improves the performance of services for individual users, while at the same time actively spreading chunks on the P2P network.

7 compares the miss ratio (MR) between the chunk selection methods according to an embodiment of the present invention.

The miss ratio refers to the percentage of chunks that are not down before the playback point has passed while streaming VoD. That is, the miss ratio refers to the number of chunks not received compared to the number of chunks played per second while the node receives the Vod streaming service.

For example, when the value of MR is 0.1, while 10 chunks are to be reproduced, it indicates that the reproduction time of the chunk has elapsed without one chunk being received.

Graph 710, along with the other techniques described above, shows how much chunk miss occurred during VoD streaming when the ENR technique was used.

Looking at the results of graph 710, the ENR technique has a lower MR value than other techniques. That is, the ENR technology may receive a chunk to be reproduced in advance when compared to other technologies when playing VoD streaming. As such, ENR technology can improve the performance of VoD streaming services.

When the MR value becomes 0 at the start of VoD reproduction, a phenomenon occurs due to a startup delay, which will be described later. A significant increase in MR value after the initial startup delay indicates that chunk misses have occurred for the trailing portion that was not downloaded at the initial startup delay as playback begins.

The ENR also has a relatively lower MR value in this early part compared to other techniques. This indicates that the amount of chunks down by the ENR technique is more than the amount of chunks down by the other techniques because the ENR technique uses bandwidth efficiently as described above.

8 compares a startup delay (SD) between chunk selection methods according to an embodiment of the present invention.

When starting the VoD streaming service, if the received video file is played after the leading portion of the video file is received for a predetermined amount or more, the performance of streaming can be further improved. Therefore, waiting time is required at the beginning of file playback. This required waiting time is called startup delay.

Depending on the startup delay, some of the performance of the VoD streaming may be determined.

Graph 810 shows the ENR and SD of each of the other techniques described above.

According to graph 810, the G2G and beta selection techniques that receive the higher priority heads by dividing the set have a relatively lower SD value than the ENR technique.

However, there is no significant difference between the SD values of the two technologies (G2G technology and the beta selection technology) and the SD values of the ENR technology, and the ENR technology uses bandwidth compared to the two technologies (G2G technology and beta selection technology). And better performance in terms of chunk miss.

Thus, in overall terms, ENR technology has better performance than other technologies.

ENR technology can allow each user to receive a larger amount of chunks by minimizing waste of bandwidth within the user's device (ie, node) and network.

Thus, ENR technology can help with VoD streaming, thereby allowing more chunks to spread on the P2P network. In addition, ENR technology can provide a better performance VoD streaming service to the user by scheduling the downloaded chunk appropriately for Vod streaming.

Method according to an embodiment of the present invention is implemented in the form of program instructions that can be executed by various computer means may be recorded on a computer readable medium. The computer readable medium may include program instructions, data files, data structures, etc. alone or in combination. The program instructions recorded on the medium may be those specially designed and constructed for the present invention or may be available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks, such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

As described above, the present invention has been described by way of limited embodiments and drawings, but the present invention is not limited to the above embodiments, and those skilled in the art to which the present invention pertains various modifications and variations from such descriptions. This is possible.

Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined by the equivalents of the claims, as well as the claims.

500: P2P network node
510: transceiver
520: control unit
530: storage unit
540: P2P file sharing unit

Claims (18)

A method of downloading one or more chunks included in a file in a P2P network connected to one or more neighboring nodes, the method comprising:
Calculating a priority of the transmission of the chunk for each of the one or more chunks;
Selecting a chunk to request transmission from among the one or more chunks based on the priority;
Requesting transmission of the selected chunk to one or more of the neighboring nodes; And
Downloading the chunk that requested the transmission from one or more of the one or more nodes for which the transmission of the selected chunk was requested.
Including,
Wherein the priority is calculated based on a first value calculated based on the sparsity of each of the one or more chunks and a second value calculated based on a position in the one or more chunks of each of the one or more chunks. How to download. The method of claim 1,
And the file is a file being played and is a file being transmitted through streaming for the playback. The method of claim 2,
Selecting the one or more chunks of the chunks constituting the file as an interval for download;
Further comprising, the file download method. The method of claim 2,
Wherein the one or more chunks are chunks after the location of the playback in the file, are chunks that have not been downloaded, and are chunks within a specific range from the location of the playback. The method of claim 1,
The first value is calculated based on Equation 1 below.
[Equation 1]

Where b _i represents the i th chunk, r (b _i ) is a sparsity function indicating the number of neighbor nodes including the chunk b _i among the neighbor nodes, and k is the total number of chunks constituting the file. . The method of claim 1,
The second value is calculated based on Equation 2 below.
&Quot; (2) "

Where b _i represents the i th chunk, d represents the beta distribution function, and d (b _i ) is the value of beta distribution function of b _i The method of claim 1,
Wherein the priority is a weighted sum of the first value and the second value. The method of claim 7, wherein
And the weight is calculated based on the proportion of chunks that can be provided by a node below a certain number of the one or more chunks. The method of claim 1,
The priority is calculated based on Equation 3 below.
&Quot; (3) "

Where b _i denotes the i th chunk, r (b _i ) is a sparse function indicating the number of neighbor nodes including chunk b _i among the neighbor nodes, k is the number of the one or more chunks, and d is A beta distribution function, d (b _i ) represents a beta distribution function value of b _i , ρ is a weight assigned to the first value, and (1-ρ) is a weight given to the second value. 10. The method of claim 9,
The p is calculated based on Equation 4 below.
&Quot; (4) "

Where t is a predefined value, r (b _i ) is the number of nodes that can provide b _i of the neighboring nodes, and W is the number of the one or more chunks. A P2P file sharing unit for downloading one or more chunks included in a file from a P2P network connected with one or more neighboring nodes,
A priority calculator for calculating a priority of transmission of the chunk for each of the one or more chunks; And
A transmission request chunk selecting unit that selects a chunk to request transmission from among the one or more chunks based on the priority;
Wherein the priority calculator calculates a first value based on the sparsity of each of the one or more chunks, calculates a second value based on a position in the one or more chunks of each of the one or more chunks, and And calculating the priority based on the first value and the second value. The method of claim 11,
And a window selector for selecting the one or more chunks as the section for downloading among the chunks constituting the file. The method of claim 11,
Wherein the sparsity of each of the one or more chunks represents a ratio of each of the one or more chunks to all the chunks that are spread to the neighboring nodes. The method of claim 11,
And wherein the relatively preceding chunk of the one or more chunks has the second value higher than the later located chunk. The method of claim 11,
And said priority is a weighted sum of said first value and said second value. 16. The method of claim 15,
And wherein the weight is calculated based on the number of chunks of the one or more chunks that have spread below a predefined value in the network. 16. The method of claim 15,
And the priority calculator increases the weight for the first value as the number of chunks spread below the predefined value in the network among the one or more chunks increases. 16. The method of claim 15,
And the priority calculator increases the weight for the second value as the number of chunks spread out below a predefined value in the network among the one or more chunks increases.

Images