WO2009043253A1

WO2009043253A1 - Peer-to-peer media streaming determining method and device

Info

Publication number: WO2009043253A1
Application number: PCT/CN2008/072194
Authority: WO
Inventors: Jing Liu; Tingyong Liu; Zhixin Sun; Jing Gong
Original assignee: Huawei Technologies Co., Ltd.; BU, Kai
Priority date: 2007-09-14
Filing date: 2008-08-29
Publication date: 2009-04-09
Also published as: CN101170496A; CN101170496B

Abstract

A peer-to-peer media streaming determining method and device involving communication technique has solved the problem that existing technique cannot effectively identify PMS so as to cause the running instability of the network system. The method includes: gathering media streaming, the media streaming comprising data packets; determining the media streaming according to the protocol adopted by the data packets and the characteristic of the data packets in the media streaming. The technical solution can be applied to the network using P2P technique.

Description

Method and apparatus for determining peer-to-peer media flow The present application claims priority to Chinese Patent Application entitled "A Method and Apparatus for Determining a Point-to-Point Media Stream" submitted to the Chinese Patent Office on September 14, 2007, application number 200710145490. The entire contents of which are incorporated herein by reference.

Technical field

The present invention relates to a communication technology, and in particular, to a method and apparatus for determining a P2P Media Streaming (PMS).

Background technique

In the network model using P2P (Peer-to-Peer, P2P) technology, each node is peer-to-peer, has the same responsibilities and capabilities, and can easily share various information resources. Due to the importance and extensiveness of information resource sharing, P2P technology with information resource sharing has been rapidly popularized and developed, attracting more and more people to share various information resources using P2P technology. Thus, using P2P technology The traffic of various media streams (P2P Media Streaming, PMS, also known as P2P media streams) is also increasing in the proportion of the entire network traffic. According to statistics, in just a few years, the traffic of P2P media streams has occupied more than 60% of the traffic in the fixed network. Correspondingly, the communication of P2P media streams occupies a large bandwidth, for WEB, Email, etc. Network services pose a serious threat. Therefore, how to identify the P2P media stream to control it has gradually become a concern of people.

In the prior art, there are two methods for identifying a PMS: a port-based identification method and a keyword-based recognition method. The two methods are described separately below.

1, based on port identification method

The method determines if the media stream is a PMS by determining a port used by the media stream. When the port used by the media stream is the same as the predetermined port, the media stream is considered to be a PMS; otherwise, the media stream is considered not to be a PMS. For the port-based PMS identification method, since some PMSs no longer use fixed ports (ie, use dynamic ports), such PMSs cannot be identified through the ports. 2, based on keyword recognition method

The method is to determine the PMS by studying the keywords that the PMS has. The packet is deeply detected. When a specific location in a packet of the media stream has a specific keyword, it can be determined that the media stream belongs to the PMS. For the keyword-based PMS identification method, some or all of the bytes of the keyword may change depending on the operating environment, version, etc. of the software. In order to identify the PMS, it is necessary to track the latest keyword information of various PMSs to ensure that the keywords that are summarized are PMS keywords. Therefore, this method is time consuming and laborious.

In summary, there is currently no effective PMS identification method, which can not effectively control the PMS, so that the current network system is unstable under the influence of PMS, and sometimes even paralyzed. Summary of the invention

The embodiment of the invention provides a method and a device for determining a point-to-point media stream, which can effectively identify a PMS, so as to perform flow control on a host that performs PMS communication, so that the network system runs stably.

An embodiment of the present invention provides a method for determining a peer-to-peer media stream, including:

Collecting a media stream, where the media stream includes a data packet;

A point-to-point media stream is determined based on protocol and packet characteristics employed by the data packet in the media stream.

Another embodiment of the present invention further provides a point-to-point media stream determining apparatus, including: a collecting unit, configured to collect a media stream, where the media stream includes a data packet;

And a determining unit, configured to determine a point-to-point media stream according to a protocol and a packet feature used by the data packet in the media stream collected by the collection unit.

The embodiment of the present invention identifies the PMS by using the packet characteristics. The method is simple, and can meet different requirements according to needs, such as performing flow control on the host performing PMS communication, so as to improve the operating efficiency of the network system. Or implement a business guarantee for users who sign up for the PMS service, and implement a "best effort" strategy for users who are not contracted to the PMS service.

DRAWINGS

FIG. 1 shows a process of determining a PMS according to an embodiment of the present invention; 2 shows a packet length characteristic of a TCP-PMS according to an embodiment of the present invention;

FIG. 3 is a diagram showing a packet length matching template of a PMS according to an embodiment of the present invention; FIG.

FIG. 4 is a diagram showing packet length characteristics of a UDP-PMS according to an embodiment of the present invention; FIG.

FIG. 5 is a diagram showing a determining apparatus of a PMS according to Embodiment 3 of the present invention;

FIG. 6 shows an identification sample of a PMS media stream according to an embodiment of the present invention; FIG.

Fig. 7 shows the recognition result of the PMS media stream of the embodiment of the present invention.

detailed description

Embodiments of the present invention are now described in conjunction with the accompanying drawings in which FIG.

Referring to FIG. 1, a PMS identification method of the present invention is described, including:

Step 11. Monitor the traffic change of the host in the network.

Step 12: Determine whether the traffic change amount of the host is greater than the traffic change threshold LR, and if yes, perform step 13; otherwise, return to step 11.

The flow change threshold LR can be set to 100 times.

You can also monitor the traffic change of the host in the network and directly perform step 13 to collect the media stream passing through the host.

Step 13: Perform sampling on the media stream passing through the main body, where the media stream includes a data packet. For example: 500 data packets can be collected as a sample to identify the PMS.

Step 14: Determine, according to the protocol and the data packet feature used by the data packet in the media stream, whether the media stream obtained by the sampling set is a PMS (for the determination method, refer to Embodiment 1 and Embodiment 2), where the protocol is a transmission control protocol. Or a user datagram protocol, the data packet feature including a packet length feature or a packet address feature.

Step 15: Record the IP address corresponding to the identified PMS, so as to provide the IP address information to the traffic management device, so that the traffic management device manages and controls the traffic of the host corresponding to the IP address (such as speed limit and access prohibition). Internet, etc.).

Since the PMS can pass the Transmission Control Protocol (TCP) or the user The data is transmitted by the User Datagram Protocol (UDP). The method for identifying the PMS using TCP and UDP is described by Embodiment 1 and Embodiment 2, respectively.

Embodiment 1

For the PMS using TCP as the transmission protocol, after analyzing a large number of PMSs, the feature of the packet length in the PMS has a relatively fixed feature, so that the packet length feature can be considered to identify the PMS.

As shown in Figure 2, for a PMS that uses TCP as the transport protocol, the two hosts interact very frequently, usually only transmitting a small number (or even one, two, generally less than 4) of data packets containing program data ( After the data packet is also referred to as a long data packet, the data packet including the control information is again exchanged (hereinafter, such a data packet is also referred to as a short data packet). The long data packet is above 1400bytes; the short packet length is generally below lOOOObytes. Therefore, the PMS that uses TCP as the transport protocol is a long-packet, short-packet-phase media stream.

In order to determine whether a certain data is a long data packet or a short data packet, a length threshold may be set to compare the packet length with the length threshold. If the length is greater than the length threshold, the packet is considered to be a long packet. Otherwise, , think that the packet is a short packet. In this embodiment, the length threshold can be set to 1000 bytes.

In order to improve the accuracy of the identification, the matching template (sPS1, sLS, sPS2) is used as the matching template of the PMS stream, where sPS1 is the number of the first standard continuous long data packet, and sLS is the standard continuous short data packet. Length, sPS2 is the number of second standard continuous long packets. Figure 3 shows the matching template (sPSl, sLS, sPS2). Where sPS1 shows the number of first standard continuous-length data packets, sLS shows the total number of bytes of the standard short data packet, and sPS2 shows the number of second standard continuous-length data packets. For example: In the matching template (sPS1, sLS, sPS2) of this embodiment, sPS1=3, sLS=1000 bytes, and sPS2=3.

Thus, the number of consecutive long packets in the media stream, the total number of bytes LS of consecutive short packets, and the number of consecutive long packets PS2 are sequentially determined. The method of determination is as follows:

First, a data packet is taken from the collected media stream to determine whether the data packet is long data. Packet, if not, take a packet and repeat this step until the packet is a long packet or arrives at the last packet.

Next, determine the number of long data packets, as follows: Set an initial value for PS1 or PS2, for example: set to 1 in this embodiment; determine the next data packet of the current data packet in the media stream as the current data packet, Whether the current data packet is a long data packet, and if so, adding a preset value increment to PS1 or PS2, which is set to 1 in this embodiment, repeating this step until the current data packet is a short data packet or reaches the last data packet. , returns the value of PS1 or PS2.

Finally, determine the total number of bytes of the short packet as follows: Set an initial value for the LS, for example, set the initial value of the LS to 0; and use the first short packet in the collected media stream as the current data packet. And adding the length of the first short data packet to the initial value of the LS; then, using the next data packet of the current data packet in the media stream as the current data packet, determining whether the current data packet is a short data packet, and if so, LS plus the current packet length, repeat this step until the current packet is a long packet or arrives at the last packet, returning the value of LS.

Repeat the above steps of determining the number of long data packets (PS1, PS2), determining the total number of bytes (LS) of the short data, and obtaining the packet characteristics of a group of media streams (PS, LS, PS, ..., ). Matching the matching template with the packet characteristics of the media stream determines whether the media stream is a PMS. The method of judgment is as follows:

Μ = (sPS - Ρ ) & 8L SLS - LS) &&(sPS - PS ₂ ) =

Where M is the identification factor. When M=l, it indicates that the media stream is a TCP-PMS stream. When M=0, it indicates that the media stream is not a TCP-PMS stream; "&&" represents a logical AND operation. The calculation rule is that the result is only 1 when all the values participating in the operation are non-negative (preventing the case of PSl=sPS or PS2=sPS from being misidentified as non-PMS traffic), otherwise 0.

Embodiment 2

For UDP-PMS, the data length of the UDP-PMS transmission program data is also large, and a small number of data packets including control information are interspersed with each other, in order to determine whether a certain data is For a long data packet or a short data packet, a length threshold may be set, and the data packet length is compared with the length threshold. If the length is greater than the length threshold, the data packet is considered to be a long data packet. Otherwise, the data packet is considered to be a short data packet. . In this embodiment, the length threshold can be set to 900 bytes. FIG. 4 shows the packet length characteristics of the UDP-PMS of the embodiment of the present invention.

Starting from the detected long data packet, until the end of the short data packet is detected, the short data packet is set as the current data packet, if the current data packet source address is equal to the previous data packet source address, and the current data packet destination address is If the destination address of the previous packet is equal, the media stream is determined to be a point-to-point media stream. The judgment method can be expressed by the following formula:

Μ

Wherein, M' is a recognition factor. When M' =1, it indicates that the media stream is a UDP-PMS stream. When M' =0, it indicates that the media stream is not a UDP-PMS stream; the operation rule of " == " is : If the values on both sides of the symbol are equal, the operation result is 1; otherwise, it is 0; Src(ei') indicates the source address of the current data packet; Src(ei, -1) is the source address of the previous data packet of the current data packet. Dst(ei, ) indicates the destination address of the current packet; Dst(ei, -1) is the destination address of the previous packet of the current packet.

Embodiment 3

As shown in FIG. 5, this embodiment describes a peer-to-peer media stream determining apparatus, including: a collecting unit 51, configured to collect a media stream, where the media stream includes a data packet, and a determining unit 52, configured to The protocol and packet characteristics used by the data packet in the media stream collected by the set unit 51 determine a point-to-point media stream, and the protocol is a transmission control protocol or a user datagram protocol, and the data packet feature includes a packet length feature or data. Package address characteristics.

The determining unit 52 may include: a determining unit 521, configured to determine a protocol used by the data packet in the media stream, and when the determining result is a transmission control protocol, start the first determining unit 522; when the determining result is the user data^ protocol The first determining unit 522 is configured to determine, in sequence, the number of consecutive long packets in the media stream, the total number of bytes LS of consecutive short packets, and the number of consecutive long packets PS2; sPS-PSl ) && ( sLS-LS ) && ( sPS-PS2 ) is true, The media stream is a point-to-point media stream, where (sPS-PSl), (sLS-LS), and (SPS-PS2) are non-negative, the value is true, otherwise the value is ^ && is logical AND; sPS is The first standard consecutive number of data packets, sPS2 is the number of second standard continuous long data packets, sLS is the total number of bytes of the standard continuous short data packet; and the second determining unit 523 is configured to detect the long data packet Start, until the end of the short data packet is detected, the short data packet is set as the current data packet; if the current data packet source address is equal to the previous data packet source address, and the current data packet destination address is equal to the previous data packet destination address , the media stream is determined to be a peer-to-peer media stream.

In order to explain the correctness of the method for identifying PMS in the embodiment of the present invention, an experimental verification is performed. As shown in Figure 6, multiple types of traffic sample data are now available, including five sets of PMS media streams: PPLive, MySee, TVKoo, UUSee, PPStream, and many other groups. Type of traffic: BT (a large file sharing software), KuGoo (cool dog). Plot based on the length of the packet in each media stream. Among them, PPLive (TCP-PMS1), UUSee (TCP-PMS2), PPStream (TCP-PMS3) are TCP-PMS, My See (UDP-PMS1), and TVKoo (UDP-PMS2) are UDP-PMS. The TCP-PMS and UDP-PMS media streams are identified by the method of the first embodiment and the second embodiment, and the recognition result is shown in FIG. 7. All PMS media streams are identified in Figure 7, so the recognition success rate is 100%; on the other hand, since there is no non-PMS media stream in Figure 7, no non-PMS media streams are misclassified as PMS media. Flow, so the false positive rate is zero.

In summary, the embodiment of the present invention identifies the PMS by the length feature of the data packet, and the method is simple, and the experiment proves that the accuracy is high and the PMS can be effectively identified. Therefore, the PMS can be effectively controlled according to requirements, such as performing flow control on the host performing PMS communication to improve the operating efficiency of the network system; or implementing service guarantee for users who subscribe to the PMS service, and implementing users of the non-signed PMS service. Do your best to "strategy."

A person skilled in the art can understand that all or part of the steps of implementing the above embodiments can be completed by a program to instruct related hardware, and the program can be stored in a computer readable storage medium, such as ROM/RAM, magnetic. Disc or CD. The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any change or replacement that can be easily conceived by those skilled in the art within the technical scope of the present invention is All should be covered by the scope of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

Claim

A method for determining a peer-to-peer media stream, comprising:

Collecting a media stream, where the media stream includes a data packet;

A peer-to-peer media stream is determined based on protocol and packet characteristics employed by the data packet in the media stream.

2. The method according to claim 1, wherein the protocol is a transmission control protocol or a user datagram protocol, and the data packet feature comprises a data packet length feature or a data packet address feature.

The method according to claim 2, wherein the protocol is a transmission control protocol, the data packet feature is a data packet length feature, and the data packet length feature includes a first continuous length data packet, The total number of bytes of consecutive short packets and the number of second consecutive long packets;

Determining the point-to-point media stream according to the protocol and the packet characteristics used by the data packet in the media stream specifically includes:

Determining, in the media stream, the number of the first consecutive long data packets, the total number of consecutive bytes of the consecutive short data packets, and the number of the second consecutive long data packets;

When the number of the first standard continuous-length data packets minus the number of the first consecutive long data packets is non-negative, the total number of standard consecutive short data packets minus the total number of bytes of the consecutive short data packets is When the number of non-negative numbers and the number of second standard consecutive long packets minus the number of the second consecutive long data packets is non-negative, it is determined that the media stream containing the data packet is a point-to-point media stream.

The method according to claim 3, wherein the step of determining the number of the first consecutive long data packets or determining the number of the second consecutive long data packets comprises:

Collecting a first long data packet in the media stream, and setting an initial value for the first consecutive long data packet number or the second continuous long data packet number;

Determining whether the current data packet is a long data packet by using the next data packet of the first long data as the current data packet, and if yes, the first continuous long data packet number or the second continuous long data packet The initial value of the number of packets plus a preset value added, repeating this step until the current data packet is a short data packet or reaches the last data packet, and returns the number of the first continuous long data packet or the second Even The number of continuation packets.

The method according to claim 4, wherein the step of collecting the first long data packet in the media stream specifically includes:

Taking a data packet from the collected media stream, determining whether the data packet is a long data packet, and if not, taking a data packet, repeating this step until the data packet is a long data packet or reaches the last data packet. .

The method according to claim 3, wherein the step of determining the total number of bytes of the consecutive short data packets specifically includes:

Setting an initial value for the total number of bytes of the consecutive short packets;

The next data packet of the current data packet in the media stream is used as the current data packet, and it is determined whether the current data packet is a short data packet. If yes, the total number of bytes of the consecutive short data packet is added to the current data packet length, and the current packet is repeated. Step, until the current data packet is a long data packet or reaches the last data packet, the total number of bytes of the consecutive short data packet is returned.

The method according to claim 1, wherein the protocol is a user datagram protocol, the data packet feature is a data packet address feature, and the data packet is used according to the data packet. The protocol and packet characteristics determine the point-to-point media flow steps specifically include:

Starting from the detected long data packet until the short data packet is detected, the short data packet is set as the current data packet;

If the current data packet source address is the same as the previous data packet source address, and the current data packet destination address is the same as the previous data packet destination address, the media stream is determined to be a point-to-point media stream.

The method according to any one of claims 1 to 7, wherein the method further comprises: before collecting the media stream, the method further comprising:

Monitor the amount of traffic change on the host;

When the traffic variation of the host is greater than the traffic change threshold, the media stream is collected.

The method according to any one of claims 1 to 7, wherein after the step of determining a point-to-point media stream, the method further comprises: recording a source address and a destination address of the media stream.

10. A point-to-point media stream determining apparatus, comprising:

a collection unit, configured to collect a media stream, where the media stream includes a data packet;

The device according to claim 10, wherein the determining unit specifically includes: a determining unit, configured to determine a protocol used by the data packet in the media stream, and when the determining result is a transmission control protocol, start a first determining unit; when the determination result is a user datagram protocol, starting the second determining unit;

The first determining unit is configured to sequentially determine the number of the first continuous long data packet in the media stream, the total number of consecutive short data packets, and the second consecutive long data packet; when the first standard continuous long data packet The number of the first consecutive long data packets is reduced to be non-negative, the total number of bytes of the standard continuous short data packet minus the total number of bytes of the consecutive short data packets is non-negative, and the second standard is consecutive When the number of long data packets minus the number of the second consecutive long data packets is non-negative, determining that the media stream containing the data packet is a point-to-point media stream;

The second determining unit is configured to start from the detected long data packet until the end of the short data packet is detected, and set the short data packet as the current data packet; if the current data packet source address and the previous data packet source address Equal, and the current packet destination address is equal to the previous packet destination address, then the media stream is determined to be a point-to-point media stream.