KR20140026993A - System and method for packetizing data stream in streaming service based on peer to peer - Google Patents

Abstract

Disclosed are a system and a method for packetizing data stream in a streaming service based on P2P. The packetizing system indexing the data stream, received from the encoder, to a piece, includes a piece generator for generating the piece of the received data stream every preset time interval. [Reference numerals] (1010) Piece generator; (1020) Encoder; (1030) Delivery server

Description

[0001] SYSTEM AND METHOD FOR PACKETIZING DATA STREAM IN STREAMING SERVICE BASED ON PEER TO PEER [0002]

Embodiments of the present invention relate to a system and method for packet-timing a data stream in a P2P-based streaming service.

Peer to Peer Service (P2P) is a service that receives information directly from all personal computers connected to the Internet, unlike the conventional method of searching for information on the Internet through a search engine.

In general, "streaming" refers to a technique of transmitting and receiving a file in real time via a network, and the most commonly used technique is progressive streaming.

When the streaming technology is used for P2P technology, the load on the server and the service side can be reduced, thereby achieving a cost reduction effect. To implement this, a client using a P2P-based streaming service communicates with a P2P-based streaming server by including a function for using a P2P-based streaming service in a dedicated player for playing content such as a multimedia file, Thereby realizing a streaming service. More specifically, the server provides the user with a list of contents that can be provided through the web page, and the client clicks and selects the contents to be served from the list. Thereafter, when content to be serviced is selected, a dedicated player installed in the client is executed to reproduce the selected content. At this time, the dedicated player uses the P2P technology to transmit the selected content to another peer and reproduce the selected content.

When a user client who wishes to use a P2P-based streaming service accesses a server and receives content from a user client, he or she searches for a peer having the selected content and connects the client to provide the content to the user client. That is, the dedicated player of the user client receives and reproduces pieces of the selected content from the server or other peers holding the content.

At this time, in the P2P-based streaming service according to the related art, the data stream is generated as a piece of a certain size. However, since the data stream transmitted from the encoder is not constant, the generation time of each piece is not constant, and thus the synchronization with the peer is not performed and the sharing rate is lowered. In addition, since the bit rate of the encoder needs to be set in advance in the system, the system must be restarted in case of encoder failure or replacement, and the VBR (Variable Bit Rate) can not be supported.

In this specification, a system and method are provided for efficiently packetizing a data stream in a P2P-based streaming service.

A packetizing system and method for generating a piece for a data stream received at regular intervals are provided.

It is possible to cope with inconsistent encoder traffic by generating pieces for a data stream received at a predetermined time instead of generating a piece at a predetermined size and to operate the system irrespective of the bit rate, Variable Bit Rate) can be supported.

There is provided a packetizing system and method capable of generating a free piece to maintain synchronization with a peer and preventing a decrease in the sharing rate when the received data stream does not exist in the current time interval to generate the piece.

A packet-timing system for indexing a data stream received from an encoder as a piece, the packet-timing system comprising a piece generating unit for generating a piece for a received data stream at predetermined time intervals.

According to one aspect, the piece generation unit may generate an empty-piece for a time interval during which a data stream is not received.

According to another aspect, the generated piece can be transmitted to the client via the delivery server at the request of the client.

According to another aspect, the server usage can be adjusted according to the ratio of the number of the current concurrent users to the constant determined using the concurrent visitor inflow rate.

According to another aspect, the constant is determined to be a relatively large value as the inflow rate is high, and may be determined to be a relatively small value as the inflow rate is low.

A packet timing method for indexing a data stream received from an encoder to a piece, the method comprising: generating a piece for a received data stream at predetermined time intervals.

It is possible to cope with inconsistent encoder traffic and to operate the system irrespective of the bit rate by generating pieces for a data stream received at a predetermined time instead of generating a piece in a predetermined size, .

If there is no received data stream in the current time interval to generate the piece, the empty piece can be generated to maintain synchronization with the peer, and the decrease in the sharing rate can be prevented.

1 is a diagram illustrating a logical configuration of a system for providing a P2P-based streaming service in an embodiment of the present invention.
2 is a diagram illustrating a physical configuration of a system for providing a P2P-based streaming service in an embodiment of the present invention.
FIG. 3 is a view for explaining redundancy of a packetizing server in an embodiment of the present invention. FIG.
4 is a diagram for explaining redundancy of a delivery server in an embodiment of the present invention.
5 is a view for explaining redundancy of an index server according to an embodiment of the present invention.
6 is a diagram illustrating a peer management system in an embodiment of the present invention.
7 is a diagram for explaining a buffer available in a peer according to an embodiment of the present invention.
FIG. 8 shows a graph when a piece is generated with a predetermined size and a graph when a piece is generated at a predetermined time interval.
9 shows a graph when a data stream does not exist for a specific time interval and a graph when a bin piece is generated.
10 is a block diagram for explaining a packet-taining system in an embodiment of the present invention.
11 is a flowchart for explaining a packet-taining method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram illustrating a logical configuration of a system for providing a P2P-based streaming service in an embodiment of the present invention. 1 shows a media source 110, a packetizing server 120, a delivery server group 130, a client 140, and a plurality of peers 150.

Here, the media source 110 may refer to an encoder that provides a stream. Packet-tinging server 120 may receive and index the data stream from such a media source 110 as a piece. As a method of indexing a piece to provide a stream through a P2P service, one of various known methods can be used. One packetizing server 120 may exist for each media source 110. For example, if four games are played during a professional baseball relay, and four media sources 110 exist for four matches, four packetizing servers 120 may also be used. The packetizing server 120 will be described later in more detail with reference to FIG.

The delivery server group 130 may include at least one delivery server 131. At this time, the number of delivery servers 131 operating in the delivery server group 130 can be adjusted according to the number of simultaneous users. The delivery server 131 can receive the indexed pieces from the packetizing server 120 and buffer them. At this time, the client 140 can transmit the piece to the client 140 at the request of the client 140.

Here, the client 140 may refer to a user terminal such as a PC, and may include a peer 141 and a player 142 as shown in FIG. The peer 141 can receive the piece from at least one of the delivery server 131 and the plurality of peers 150 and transmit the data stream to the player 142. [ In one example, the peer 141 may be a program installed and operating in the client 140. A plurality of peers 150 may also be installed and operated in a plurality of clients, respectively.

In other words, the data stream converted into pieces by the packet-tensing server 120 may be transmitted to at least a part of all the clients through the delivery server group 130, and may be transmitted to the delivery server 131, And transmits the data stream to the player 142, so that the user is provided with a P2P-based streaming service.

Here, the server usage amount control system according to an embodiment of the present invention may refer to the system itself described with reference to FIG. 1, and may be included in or linked to the system described with reference to FIG. 1 as needed. The server usage control system adjusts the server usage adaptively considering the number of concurrent users, so that low server traffic can be maintained regardless of the number of concurrent users. At this time, the server usage amount can be adjusted based on the ratio of the constant determined by using the concurrent visitor inflow rate to the current number of concurrent users. For example, the server usage can be calculated as shown in Equation 1 below.

Here, 'c' may be a constant determined by using the inflow rate of concurrent users, and 'n' may be a number of concurrent users. In this case, 'c' is a constant that can be determined by the administrator or system through empirical test. Basically, the higher the inflow rate of the concurrent user, the higher the value, and the lower the inflow rate, the smaller the value have. Equation (1) can be used to cover 'n' users with 'c' traffic. For example, if 'c' is 6, it means traffic to provide to 6 people, covering 'n' names (current concurrent users).

That is, even if the number of concurrent users is increased due to the increase of the concurrent access rate, the constant 'c' is fixed, while the server usage is decreased because 'n' increases. Therefore, it is possible to always maintain a low server usage regardless of the number of concurrent users.

To this end, the server usage control system may be configured to determine a current concurrent visitor number based on a ratio of a current concurrent visitor recipient providing a current concurrent visitor number (not shown) and a constant determined by using the concurrent visitor & And a server usage control unit (not shown) for controlling usage.

2 is a diagram illustrating a physical configuration of a system for providing a P2P-based streaming service in an embodiment of the present invention. 2 is a block diagram illustrating an external IDC 210 providing a device for a P2P-based streaming service, an encoder vendor 220 providing a data stream, and an external IDC 210 managing a process of providing stream data to the client 240 Lt; RTI ID = 0.0 > IDC 230 < / RTI >

The external IDC 210 may include a packetizing server 211, a main net checker 213, a subnet checker 214, a delivery server 212 and an index server 215 and a plurality of switches 216 have. Here, each of the packetizing server 211, the delivery server 212, the main net checker 213, the subnet checker 214, and the index server 215 may be composed of a plurality of servers instead of one server Each of the plurality of switches 216 may be used to transmit data to or receive data from a corresponding server among the plurality of servers. As each of the plurality of switches 216, for example, the switch L4 may be used.

The packetizing server 211 receives the data stream from the media encoder 221 of the encoder vendor 220 and processes the received data stream into data for use in the system. That is, the packetizing server 211 can convert the data stream into a plurality of pieces. As described above, one packetizing server 211 can be executed for each media encoder 221. [

The delivery server 212 delivers a piece received from the packet-ting server 211 to the client 240 in response to a request from the client 240. In addition, the index server 215 maintains a list of clients 240 and provides search services. Here, the main net checker 213 and the subnet checker 214 may refer to a relay server that relays peer-to-peer connections.

Table 1 below shows the number of servers required when the number of concurrent users is 150,000, the content bit rate is 500kbps, and the sharing rate is 80%.

The LMS 231 included in the internal IDC 230 may refer to a peer management system. The LMS 231 is a server that manages the packetizing server 211, the delivery server 212 and the index server 215. The LMS 231 includes an upload download state or traffic of the server, the number of queries, a resource (CPU, memory) Rate and so on. In addition, the LMS 231 may generate statistical data on the number of concurrent users, unique visitors, sharing rate, user speed distribution, average viewing time, and viewing channel number, and may store the data in the DB 232. That is, the DB 232 may refer to a server that stores statistical data.

At this time, as described above, the system according to an embodiment of the present invention adjusts the server usage amount adaptively considering the number of concurrent users, so that low server traffic can always be maintained regardless of the number of concurrent users. At this time, the server usage amount can be adjusted based on the ratio of the constant determined by using the concurrent visitor inflow rate to the current number of concurrent users. For example, the server usage can be calculated according to Equation (1) described above, and the server usage can refer to the traffic of the delivery server 212 described with reference to FIG. The server usage control system described with reference to FIG. 1 may refer to the system itself described with reference to FIG. 2, or may be included in the LMS 231 described above.

FIG. 3 is a view for explaining redundancy of a packetizing server in an embodiment of the present invention. FIG. FIG. 3 shows that four packetizing servers can operate in the main packetizing server group 350 when a data stream is provided through four encoders 310 to 340. FIG. That is, 'PS-1', 'PS-2', 'PS-3' and 'PS-4' may refer to the four packetizing servers described above.

At this time, the four packetizing servers can convert the data stream received from the four encoders 310 to 340 into pieces of data that can be used in the P2P, and transmit data converted to all the delivery servers. At this time, the four packetizing servers and all the delivery servers can transmit and receive data using the switch 370.

The packetizing server does not use a large amount of resources such as CPU and traffic. For example, when 20 delivery servers (based on 150,000 concurrent users) are connected, only 20 Mbps of traffic is used for 1 Mbps-sized content.

In addition, since different data streams may be transmitted from the four encoders 310 to 340 for each packet-timing server, the pieces may be divided into a main packet-testing server group 350 and a sub-packet- Server group 360 of the server group. Accordingly, the main packetizing server group 350 and the sub packetizing server group 360 can operate in an active / standby mode. The secondary packetizing server group 360 may be used to respond to a failure of the primary packetizing server group 350.

4 is a diagram for explaining redundancy of a delivery server in an embodiment of the present invention. The pieces of data, which are converted by the packetizing server group 410, are transmitted to the delivery server group 420. In this case, 'DS-n' means 'n' th delivery server, which means that there are 'n' delivery servers in the delivery server group 420.

Each delivery server receives a piece from a packet-tensing server, performs buffering for a predetermined number of packets, and transmits the corresponding piece to the peers according to a request from peers that are clients. At this time, the delivery servers of the delivery server group 420 can also be bound to the switch 430, and the traffic can be adjusted by increasing the delivery server as the number of concurrent users increases.

5 is a view for explaining redundancy of an index server according to an embodiment of the present invention. The index server group 510 may include a plurality of index servers, 'IS-n' denotes an 'n' th index server, and there are 'n' index servers in the index server group 520 . ≪ / RTI >

At this time, each of the plurality of index servers manages peers as clients. More specifically, it manages peers installed in clients and delivers search results according to the request of the peer. In addition, the index server can perform message delivery and maintain a persistent connection with the peer. Each of these index servers can be bound to the switch 520 and can increase the index server according to the number of concurrent users.

6 is a diagram illustrating a peer management system in an embodiment of the present invention. The LMS 610 shown in FIG. 6 may denote a peer management system, the IS 620 an index server, the PS 630 a packetizing server, and the DS 640 a delivery server. At this time, the peer management system can perform management, distribution, update, and monitoring functions for the index server, the packetizing server, and the delivery server, and can perform statistics collection and analysis functions for peers (peers). For example, the peer management system can monitor the status of the above-described servers, such as CPU or memory usage or traffic, and provide event notification functions (SMS, e-mail) for the failure situation.

7 is a diagram for explaining a buffer available in a peer according to an embodiment of the present invention. The peer is software that can be installed and operated on the client as described above to determine which piece to receive from. Table 2 below shows how to select a piece at a peer and the available sources.

In Table 2, 'Rarest First' means a piece selection method in which a rare piece is first received on a network, and 'Progressive' means a piece selection method in which pieces are sequentially received from the beginning. At this time, the piece can be received through the delivery server or received from another peer.

At this time, the peer can use a buffer for storing the received pieces to facilitate playback and increase sharing efficiency. FIG. 7 shows the first buffer 710 to the third buffer 730. FIG.

First, the first buffer 710 may be divided into a first range 711 for storing pieces received through 'progressive' and a second range 712 for storing pieces received through 'Rarest First' have. At the same time, the first buffer 710 can be divided into a third range 713 for storing the pieces received via the delivery server and a fourth range 714 for storing pieces received via other peers.

The second buffer 720 indicates areas of the buffer actually divided. That is, the second buffer 720 includes a first region 721 for storing pieces received in a 'progressive' manner via a delivery server, a second region 721 for storing pieces received in a 'progressive' manner via another peer 722) and a third area 723 for storing pieces received as 'Rarest First' through other peers.

The third buffer 730 indicates that the sizes of the divided areas may be different from each other. That is, the third buffer 730 shows an example in which the area A 731, the area B 732, and the area C 733 are configured at different sizes. For example, the size ratio of the area A 731, the area B 732, and the area C 733 may have a ratio of 1: 4: 16. These ratios can be adjusted as needed. In addition, a certain piece of the piece may be received from the delivery server and stored in the area B 732 and the area C 733 as needed.

The packetizing system according to an embodiment of the present invention to be described later may be a system corresponding to the packetizing server described above or a system included in the packetizing server. As described above, when a piece is generated with a predetermined size, since the data stream transmitted from the encoder is not constant, the generation time of each piece is not constant and synchronization with the peer is not performed, have.

FIG. 8 shows a graph when a piece is generated with a predetermined size and a graph when a piece is generated at a predetermined time interval. 8, the x-axis of the first graph 810 and the second graph 820 represent the flow of the packetizing time, and the y-axis represents the size of the generated piece, respectively.

That is, the first graph 810 indicates that the pieces of the same size are generated over time in the prior art, but the generation time of the pieces is not constant.

On the other hand, the second graph 820 shows that the generation time of the pieces is constant by generating the pieces at a predetermined time interval in the packet-tensing system or the packet-ting method according to the embodiments of the present invention. In this case, the size of the pieces may differ from piece to piece.

When a piece is generated every predetermined time, it is possible to cope with traffic of an encoder that is not constant, and it is possible to operate a system that provides a streaming service irrespective of a bit rate. In addition, VBR (Variable Bit Rate) can be supported.

However, when packet timing is performed at such a constant time interval, a time interval during which no data stream is received from the encoder may occur. In this case, since the piece can not be generated, synchronization with the peer can not be achieved and the sharing rate may be lowered.

Therefore, in the packetizing system or the packet-tacking method according to the embodiments of the present invention, in a time interval when no data stream is received, an empty-piece is generated to maintain synchronization with the peer, .

9 shows a graph when a data stream does not exist for a specific time interval and a graph when a bin piece is generated. First, in the first graph 910 and the second graph 920, the x-axis represents the flow of packet timing time, and the y-axis represents the size of the generated pieces.

Here, the first graph 910 indicates that no data stream has been received from the encoder during a particular time interval and that no pieces have been generated during that time interval. At this time, since the pieces can not be generated at the corresponding time intervals, synchronization with the peer is difficult to be performed, and thus the sharing rate may decrease.

On the other hand, the second graph 920 represents a case of generating a bin piece during the corresponding time interval. Thus, by using the empty piece, the synchronization with the peer can be maintained and the reduction in the sharing rate can be prevented.

10 is a block diagram for explaining a packet-taining system in an embodiment of the present invention. The packetizing system 1000 according to the present embodiment includes a piece generation unit 1010, as shown in FIG.

The packetizing system 1000 indexes the data stream received from the encoder 1020 as a piece and for this, the piece generation unit 1010 generates a piece for the received data stream every predetermined time interval do. As described above, synchronization with a peer (client) can be achieved by generating a piece for a data stream every predetermined time interval instead of generating a certain size.

In addition, the piece generation unit 1010 generates an empty-piece for a time interval during which a data stream is not received. That is, if a piece can not be generated by generating a free piece for a time interval during which the data stream is not received by the piece generation unit 1010, the problem that the synchronization with the peer is not performed and the sharing rate is lowered can be solved , It is possible to cope with the traffic of an encoder which is not constant, and it can support VBR.

At this time, the generated pieces can be transmitted to the client through the delivery server 1030 according to the request of the client. At this time, the server usage can be adjusted according to the ratio of a constant determined by using the inflow rate of the concurrent visitor to the current number of concurrent users. In this case, the constant is determined to be a relatively large value as the inflow rate is high, The lower the value, the smaller the value can be determined. Since the control of the server usage has been described in detail, repetitive description will be omitted.

The packetizing system according to the embodiments of the present invention may correspond to the above-described packetizing server or may be included in the packetizing server. The packetizing system 1000 described in FIG. 10 is included in the packetizing server and operates. When the packetizing system 1000 is a system corresponding to the packetizing server, (Not shown) for receiving the data stream from the encoder 1020 and a transmission unit (not shown) for transmitting the generated pieces to the delivery server 1030. [

11 is a flowchart for explaining a packet-taining method according to an embodiment of the present invention. The packetizing method according to the present embodiment can be performed by the packetizing system 1000 described with reference to FIG. 11, a packet timing method performed by the packetizing system 1000 will be described.

In step 1110, the packet-tensing system 1000 receives the data stream from the encoder. As already explained, the data stream received from the encoder is not constant over the same time interval. Therefore, when such a data stream is generated with a piece of a certain size, synchronization with the peer may not be performed, and the sharing rate may be lowered.

In order to solve such a problem, in the packetizing method according to the present embodiment, it is possible to generate pieces at predetermined time intervals instead of a predetermined size as in the below step 1120.

That is, in step 1120, the packet-tensing system 1000 generates a piece for the received data stream every predetermined time interval. As described above, synchronization with a peer (client) can be performed by generating pieces for a data stream at predetermined time intervals instead of generating a certain size, and it is possible to cope with traffic of an encoder that is not constant, and VBR .

In addition, the packet-tensing system 1000 generates a bin-piece for a time interval during which a data stream is not received. That is, if the packet generation system 1000 does not generate a piece by generating a free piece for a time interval during which a data stream is not received, it can solve the problem that synchronization with the peer is not performed and the sharing rate is lowered have.

In step 1130, the packet-tensing system 1000 transmits the generated piece to the client via the delivery server at the request of the client. At this time, the generated pieces can be transmitted to the client through the delivery server according to the request of the client. At this time, the server usage can be adjusted according to the ratio of a constant determined by using the inflow rate of the concurrent visitor to the current number of concurrent users. In this case, the constant is determined to be a relatively large value as the inflow rate is high, The lower the value, the smaller the value can be determined. Since the control of the server usage has been described in detail, repetitive description will be omitted.

11, the packetizing method performed by the packetizing system 1000 corresponding to the packetizing server 1000 has been described. When the packetizing system 1000 is a system included in the packetizing server, The packet timing method performed by the packetizing system 1000 may include only the step 1120. [

As described above, by using the packetizing system and method according to the embodiments of the present invention, it is possible to generate pieces for a data stream received every predetermined time instead of generating a piece in a predetermined size, Can operate the system regardless of the bit rate, and can support Variable Bit Rate (VBR). In addition, if there is no received data stream in the current time interval to generate the piece, it is possible to generate a bin piece to maintain synchronization with the peer, and to prevent a decrease in the sharing rate.

Methods according to an embodiment of the present invention may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, 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. Furthermore, the above-described file system can be recorded on a computer-readable recording medium.

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.

1000: Packetizing system
1010:
1020: encoder
1030: Delivery server

Claims (1)

System for packetizing data stream in P2P based streaming service.

Images