WO2013017077A1

WO2013017077A1 - Method for coding thereof related device and communication system

Info

Publication number: WO2013017077A1
Application number: PCT/CN2012/079469
Authority: WO
Inventors: 李挥; 李江; 张明龙; 周艳; 赵书江; 魏凯
Original assignee: 华为技术有限公司
Priority date: 2011-08-01
Filing date: 2012-08-01
Publication date: 2013-02-07
Also published as: CN102916761B; CN102916761A

Abstract

Disclosed in the embodiment of the present invention are a method for coding thereof related device and communication system. In the embodiment of the present invention, DNCP(DNCP: Doublesource Network Coding Protocol) is provided for supporting coding and decoding between two signal sources, wherein, the coding device performs random linear coding to two IP packages whose arrival interval are smaller than the setting duration and the destination addresses are same, through this ,it is useful for partly eliminating the influence of mismatch of transmitting rate and transfer delay in the actual network; at the same time, the header of DNCP package carries two source IDs, two IP package IDs and coding coefficient vector information which is related to the coding coefficient vector, which lays the foundation for the coding device continuously to code multi times or for the decoding device to decode rapidly, and benefits improving the generality of network coding in the actual network; and as the present invention is carried out based on network coding technology, therefore it can make full use of network resource and increase the network throughout rate obviously and achieve higher transmission efficiency.

Description

Codec method and related equipment and communication system The application is filed on August 1, 2011, the Chinese Patent Office, the application number is 201110217978.X, and the invention is entitled "Codec and Decoding Method and Related Equipment and Communication System" Priority is hereby incorporated by reference in its entirety. TECHNICAL FIELD The present invention relates to the field of communications technologies, and in particular, to a codec method and related devices and communication systems. Background technique

The network coding (NC, Network Coding) born in 2000 proposed how to achieve the network communication capacity determined by the "maximum flow minimum cut theorem" when a single source or multiple sources in a communication network multicast or broadcast to multiple receiving points. An idea of the limit.

The routing switch on the traditional communication network node only performs the "store-and-forward" function, and regards the information as "goods", so that the information is considered to be non-superimposable. This cognitive limitation makes it difficult for the network to achieve maximum streaming. The NC indicates that if the routing switch is allowed to encode multiple incoming streams and then send it, the network communication can achieve maximum streaming. At present, the industry proposes a coding scheme for a single source to multiple receiving point scenes, which is based on the idea of dividing a data packet, and transmitting the divided data packet after network coding.

Research and practice have found that the prior art is mainly directed to a scenario where a single source is connected to multiple receiving points, and in the actual network, there are more scenarios in which multiple sources are connected to multiple receiving points, which makes the prior art unable to satisfy the actual network. The application requires, and the prior art uses a partition coding mechanism, which greatly increases the data processing load of the network node.

SUMMARY OF THE INVENTION Embodiments of the present invention provide a codec method, a related device, and a communication system, with a view to improving The flexibility of network codec reduces the complexity of network node codec.

An encoding method, including:

Receiving a first internet protocol IP data packet from the first source;

Receiving, by the set time duration after receiving the first IP data packet, a second IP data packet from the second source, and the destination address of the first IP data packet and the second IP data packet Similarly, the coding coefficient vector is randomly selected for the first IP data packet and the second IP data packet in the finite field;

And randomly encoding the first IP data packet and the second IP data packet by using the coding coefficient vector to obtain a first linearly encoded data packet;

The first linearly encoded data packet is encapsulated by a dual source network coding protocol (DNCP) to obtain a first DNCP data packet, where the DNCP header of the first DNCP data packet includes an identifier of the first source, An identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and coding coefficient vector information related to the coding coefficient vector; performing IP encapsulation on the first DNCP data packet The third IP packet is obtained and sent. An encoding method, including:

Receiving a fifth internet protocol IP data packet;

Receiving a sixth IP data packet within a set duration after receiving the fifth IP data packet, and the sixth IP data packet and the fifth IP data packet have the same destination address, and the The IP layer payloads of the six IP data packets and the fifth IP data packet both contain a dual source network coding protocol DNCP data packet, and the DNCP header of the DNCP data packet in the IP layer payload of the sixth IP data packet includes two The source identifier and the two IP packet identifiers are the same as the two source identifiers and the two IP packet identifiers included in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet, and are limited. Randomly selecting a coding coefficient vector for the load of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload in the domain;

Performing random linear coding on the payload of the DNCP packet in the IP layer payload of the fifth IP data packet and the sixth IP data packet by using the coding coefficient vector to obtain second linear encoded data Package

DNCP encapsulating the second linearly encoded data packet to obtain a third DNCP data packet, where the DNCP header of the third DNCP data packet includes the two source identifiers, the two IP packet identifiers, and Encoding coefficient vector information related to the coding coefficient vector; IP encapsulating the third DNCP data packet to obtain a seventh IP data packet and transmitting. A decoding method, including:

Receiving Internet Protocol IP data packets;

If the IP layer payload of the received IP data packet includes a dual source network coding protocol DNCP data packet, the DNCP data packet is stored in the buffer area;

When the DNCP data packet stored in the buffer area reaches a preset number, the DNCP data packet is read from the buffer area;

If the DNCP header of the read DNCP data packet includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, whether the DNCP header is present in the buffer area Another DNCP packet containing the two source identifiers and two IP packet identifiers, if any, reads the other DNCP packet from the buffer area, wherein the DNCP header of the other DNCP packet Also including another set of coding coefficient vector information related to the coding coefficient vector;

The load of the read DNCP packet is randomly linearly decoded using the two sets of coding coefficient vector information.

An encoding device comprising:

a first receiving module, configured to receive a first Internet Protocol IP data packet from the first source, and a first encoding coefficient selection module, configured to: after the first receiving module receives the first IP data packet, Receiving a second IP data packet from the second source within a set duration, and the first IP data packet and the second IP data packet have the same destination address, respectively, the first in the finite field The IP data packet and the second IP data packet randomly select a coding coefficient vector;

An encoding module, configured to use the first coding coefficient selection module to randomly select a coding coefficient Performing random linear coding on the first IP data packet and the second IP data packet to obtain a first linear encoded data packet;

a DNCP encapsulation module, configured to perform a DNCP encapsulation of the first linearly encoded data packet by using a dual source network coding protocol (DNCP), where a DNCP header of the first DNCP data packet includes an identifier of the first source And an identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and coding coefficient vector information related to the coding coefficient vector;

The IP encapsulation module is configured to perform IP encapsulation on the first DNCP data packet obtained by the DNCP encapsulation module to obtain a third IP data packet and send the same.

An encoding device comprising:

a second receiving module, configured to receive a fifth internet protocol IP data packet;

a second coding coefficient selection module, configured to: if the second receiving module receives the sixth IP data packet within a set duration after receiving the fifth IP data packet, and the sixth IP data packet and the first The destination addresses of the five IP data packets are the same, and the IP layer payloads of the sixth IP data packet and the fifth IP data packet both include a dual source network coding protocol DNCP data packet, and the sixth IP data packet IP address Two source identifiers and two IP packet identifiers included in the DNCP header of the DNCP packet in the layer payload, and two sources included in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet The identifier and the two IP data packet identifiers are the same, and the coding coefficient vector is randomly selected in the finite field for the load of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload respectively;

An encoding module, configured to perform random linear coding on a load of the DNCP data packet in the fifth IP data packet and the IP layer payload of the sixth IP data packet by using a coding coefficient vector randomly selected by the coding coefficient selection module a second linearly encoded data packet;

a DNCP encapsulating module, configured to perform DNCP encapsulation on a second linearly encoded data packet obtained by the encoding module to obtain a third DNCP data packet, where a DNCP packet header of the third DNCP data packet includes the two source identifiers, The two IP data packet identifiers and the Coding coefficient vector correlation coding coefficient vector information;

The IP encapsulation module is configured to perform IP encapsulation on the third DNCP data packet obtained by the DNCP encapsulation module to obtain a seventh IP data packet and send the same.

An encoding device comprising:

An IP layer module, configured to receive a first Internet Protocol IP data packet from the first source;

a DNCP layer module, configured to receive, by the IP layer module, a second IP data packet from a second source within a set duration after receiving the first IP data packet, and the first IP data packet And the same as the destination address of the second IP data packet, the coding coefficient vector is randomly selected for the first IP data packet and the second IP data packet in the finite field; An IP data packet and the second IP data packet are randomly linearly encoded to obtain a first linearly encoded data packet; and the first linearly encoded data packet is encapsulated by a dual source network coding protocol (DNCP) to obtain a first DNCP data packet, where The DNCP header of the first DNCP data packet includes an identifier of the first source, an identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and Coding coefficient vector information related to the coding coefficient vector; performing IP encapsulation on the first DNCP data packet to obtain a third IP data packet;

The IP layer module is further configured to send the third IP data packet obtained by the DNCP layer module. A decoding device, comprising:

a third receiving module, configured to receive an internet protocol IP data packet;

a first storage module, configured to: when the IP layer payload of the IP data packet received by the third receiving module includes a DNCP data packet, store the DNCP data packet into a buffer area;

a first reading module, configured to read a DNCP data packet from the buffer area when a DNCP data packet stored in the buffer area reaches a preset number;

a second reading module, configured to: if the DNCP packet of the DNCP packet read by the first reading module includes two source identifiers, two IP packet identifiers, and a set of coding coefficients related to the coding coefficient vector Vector information, then it is found in the buffer area that another DNCP packet whose DNCP header contains the two source identifiers and two IP packet identifiers, if any, And reading the another DNCP data packet from the buffer area, where the DNCP packet header of the another DNCP data packet further includes another set of coding coefficient vector information related to the coding coefficient vector; and a decoding module, configured to utilize two sets of coding The coefficient vector information performs a random linear decoding on the load of the read DNCP packet.

A decoding device, comprising:

An IP layer module, configured to receive an Internet Protocol IP data packet;

a DNCP layer module, configured to: if the IP layer payload of the IP data packet received by the IP layer module includes a dual source network coding protocol DNCP data packet, store the DNCP data packet into a buffer area; When the DNCP data packet of the buffer area reaches a preset number, the DNCP data packet is read out from the buffer area; if the DNCP header of the read DNCP data packet includes two source identifiers, two IP addresses The data packet identifier and a set of coding coefficient vector information related to the coding coefficient vector, and whether there is another DNCP data packet whose DNCP header contains the two source identifiers and two IP packet identifiers in the buffer area, if If yes, the another DNCP data packet is read from the buffer area, where the DNCP header of the another DNCP data packet further includes another set of coding coefficient vector information related to the coding coefficient vector; The vector information performs random linear decoding on the load of the read DNCP data packet;

The IP layer module is further configured to send a seventh IP data packet that is randomly and linearly decoded by the DNCP layer module.

A communication system comprising:

An encoding device and/or a decoding device as described in the above embodiments.

As can be seen from the above, the embodiment of the present invention proposes DNCP to support codec between two sources. Since the encoding device performs random linear coding on two IP data packets whose arrival interval is less than the set duration and the destination address is the same, this is advantageous. To some extent, the effect of mismatching problems such as transmission delay and transmission rate in the actual network is eliminated. Meanwhile, the DNCP header of the DNCP packet carries two source identifiers, two IP packet identifiers, and an encoding associated with the coding coefficient vector. Coefficient vector information, which lays the foundation for the encoding device to continue encoding multiple times or for decoding the decoding device quickly. It is beneficial to improve the versatility of network coding in actual networks; and because it is implemented based on network coding technology, it can fully utilize network resources, significantly improve network throughput and achieve higher transmission efficiency. BRIEF DESCRIPTION OF THE DRAWINGS In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings to be used in the embodiments or the description of the prior art will be briefly described below, and obviously, in the following description The drawings are only some of the embodiments of the present invention, and other drawings may be obtained from those skilled in the art without departing from the drawings.

FIG. 1 is a schematic diagram of a data packet structure after extending a DNCP layer according to an embodiment of the present disclosure;

FIG. 1 is a schematic diagram of a packet header structure of a DNCP data packet according to an embodiment of the present invention; FIG. 1 is a schematic diagram showing a comparison between a value and a meaning of a flag field in a DNCP packet header according to an embodiment of the present invention;

2 is a schematic diagram of codec decoding of a butterfly network according to an embodiment of the present invention;

3 is a schematic flowchart of an encoding method according to an embodiment of the present invention;

4 is a schematic flowchart of another coding method according to an embodiment of the present invention;

FIG. 5 is a schematic flowchart of another coding method according to an embodiment of the present invention;

6 is a schematic flowchart of another coding method according to an embodiment of the present invention;

FIG. 7 is a schematic flowchart diagram of a decoding method according to an embodiment of the present invention;

FIG. 8 is a schematic flowchart of another coding method according to an embodiment of the present disclosure;

FIG. 8 is a schematic diagram of an IP packet coding structure according to an embodiment of the present invention;

FIG. 9 is a schematic flowchart diagram of another decoding method according to an embodiment of the present invention;

10 is a schematic diagram of a hardware hierarchy of a codec device according to an embodiment of the present invention; FIG. 11 is a schematic structural diagram of a hardware module of an encoding device according to an embodiment of the present invention; FIG. 12 is a schematic structural diagram of a hardware module of a decoding apparatus according to an embodiment of the present invention; FIG. 13 is a schematic diagram of a topology structure of a cell network according to an embodiment of the present invention;

FIG. 14 is a schematic diagram of an encoding device according to an embodiment of the present invention;

FIG. 15 is a schematic diagram of an encoding device according to an embodiment of the present invention;

FIG. 16 is a schematic diagram of a decoding device according to an embodiment of the present invention;

FIG. 17 is a schematic diagram of an encoding device according to an embodiment of the present invention;

FIG. 18 is a schematic diagram of a decoding device according to an embodiment of the present invention;

DETAILED DESCRIPTION OF THE EMBODIMENTS The embodiments of the present invention provide a codec method, a related device, and a communication system, so as to improve the flexibility of network codec and reduce the complexity of coding and decoding of a network node.

In order to make the object, the features and the advantages of the present invention more obvious and easy to understand, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. The described embodiments are only a part of the embodiments of the invention, and not all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

Wherein the terms "first", "second", "third", "fourth", etc. (if present) in the specification and claims of the present invention are used to distinguish similar objects, and are not necessarily used for Describe a specific order or order.

In order to facilitate the specification and implementation of the network codec, the embodiment of the present invention proposes a Double Source Network Coding Protocol (DNCP).

The following describes the relevant content of the DNCP protocol.

The IP headers of Internet Protocol (IP) packets from different sources have different information. When IP packets are transmitted in the network, the network intermediate nodes can encode or decode them. For better specification and implementation of network codec, the embodiment of the present invention proposes For DNCP that supports codec operations between dual sources, DNCP can be added, for example, between the IP layer and the transport layer, ie the DNCP header is inserted between the IP header and the transport layer header (see Figure 1-a).

A data format of a DNCP packet header provided by an embodiment of the present invention is shown in FIG. 1-b. The meaning of each field of the DNCP header shown in Figure 1-b can be defined as follows:

Version field: For example, 4 bits (or other length), can be used to record the evolved version of the DNCP packet format, in order to distinguish between the subsequent development research version and the previous version.

The length field of the header: for example, 4 bits (or other length), which can be used to record the first degree;

The header length refers to the part of the DNCP packet except the valid data payload. The unit is 4 bytes and the minimum value is 3. For example, when the header length is 3, it may mean that the payload of the DNCP packet is not encoded, but the DNCP header is added.

The total length field: for example, 16 bits (or other length), can be used to record the total length; where the total length is the length of the sum of the header length and the payload, in bytes. The flag field: for example, 2 bits (or other lengths), can be used to identify the encoding of the DNCP packet payload. The meaning of the different values of the flag field can be as shown in Figure 1-c.

The number of encoding fields: for example, 2 bits (or other lengths), can be used to record the number of times the packet is encoded, that is, the number of times the packet is encoded from the original packet.

Among them, in an actual network, the encoding of the data packet may be recursive, that is, it may be encoded multiple times, and sometimes the DNCP header may be directly added to a certain data packet without encoding. The number of times of encoding is recorded to facilitate decoding after multiple encodings. Wherein, if the payload of the DNCP packet is the original IP packet, the number of encodings is 0, and in other cases, the number of encodings is >1. For example, when an IP packet and an encoded packet are encoded, referring to the number of encodings, the decoding router can prevent the decoding router from mistaking the DNCP packet as an IP packet and hand it to the destination node.

Original packet length field: For example, 12 bits (or other length) can be used to record the length of the unencoded original data packet. The original packet length field is used to facilitate accurate recovery of the original data packet at the time of decoding. Because the length of the data packets of different data sources may be different, in order to facilitate the coding calculation, it may be complemented by 0 to make the length consistent. Since a typical Ethernet packet is between 500 and 1500 bytes in length, the padding length is typically no more than 1000 bytes. In addition, consider the limitation of the maximum transmission unit (MTU) of the MAC layer, that is, the length of the encoded MAC frame cannot exceed 1518 bytes.

Coding coefficient field: for example, 8 bits (or 8 other lengths), can be used to record the coding coefficient vector, for example, in a Galois field GF2 ⁿ (the value range of n is, for example, an integer of 4-16). For example, GF256) randomly selects the coding coefficient vector in the finite field. In the case of multiple encoding, the new encoding coefficient obtained by performing the specific calculation of the current encoding coefficient vector and the originally recorded encoding coefficient vector may be filled in the field; if the DNCP header is directly added in front of a certain data packet, To encode it, the field of the DNCP header can be used as a reserved field, for example, the field can be filled in with FF or other values.

The original packet identification field: for example, occupying 10 bits (or other lengths), can be used to record the identifier of the encoded original data packet, and its function is to facilitate decoding, and the identifier can be, for example, an order number generated in order or from the original IP data. The identifier of the IP header of the packet is 10 bits or less, and may of course be an identifier generated by other means for identifying the original packet.

Source identification field: For example, 4 bits (or other length), which can be used to record the identifier of the source corresponding to the original data packet, and its function is to facilitate decoding.

Reserved fields: For example, 30 bits (or other length) can be used for simultaneous encoding of multiple sources. Those skilled in the art can understand that some of the fields in the DNCP header of the data format shown in FIG. 1-b are not necessary, and other arrangements may also be used between the fields, that is, the data format of the DNCP header is not limited to the above. For example.

The basic principle of NC is briefly introduced based on the butterfly network shown in Figure 2. In the acyclic, one-way butterfly network shown in Figure 1, the solid line table with arrows Show unit channel. It can be seen from the "maximum flow minimum cut theorem" that the maximum flow of the network is 2, that is, if the sinks R1, R2 simultaneously request 2 bits of information from the sources S1, S2, b, where a comes from SI and b comes from S2, then In the case of maximum streaming, communication takes place in one unit of time. However, if the traditional "store-and-forward" mode is used, congestion will occur at node M, and it takes at least two unit time to complete the above communication target. If NC is used, the M node performs encoding operation (for example, XOR operation) on the two-bit information a, b input at the same time, and encodes it into 1-bit a @ b transmission, so that R1 receives two in one unit time. The bit information a and a @ b, after decoding, can restore the original information a, b. R2 is the same. It can be seen that the introduction of NC can make full use of network resources and make network communication more efficient (the network throughput can be improved without increasing link resources).

The details will be described below by way of specific examples.

One embodiment of the encoding method of the present invention is mainly directed to a scenario in which two IP data packets are encoded. An encoding method may include: the encoding device receives a first internet protocol IP data packet from a first source; and receives a second IP from a second source within a set time period after receiving the first IP data packet a data packet, and the destination addresses of the first IP data packet and the second IP data packet are the same, and the coding coefficient vector is randomly selected for the first IP data packet and the second IP data packet in the finite field respectively; using the coding coefficient vector pair The first IP data packet and the second IP data packet are randomly linearly encoded to obtain a first linearly encoded data packet; the first linearly encoded data packet is encapsulated by a dual source network coding protocol DNCP to obtain a first DNCP data packet, and the first DNCP data packet The DNCP header of the packet includes an identifier of the first source, an identifier of the second source, an identifier of the first IP packet, an identifier of the second IP packet, and coding coefficient vector information related to the coding coefficient vector; The DNCP packet is IP encapsulated to obtain a third IP packet and sent.

Referring to Figure 3, specific steps may include:

301: The encoding device receives the first internet protocol IP data packet from the first source;

302. If the set time duration after receiving the first IP data packet (the set duration may be set according to an actual network, for example, may be set to 1 millisecond, 5 milliseconds, 10 milliseconds, 50 milliseconds, or other values) To a second IP packet from the second source, and the first IP packet and the second IP data If the destination addresses of the packets are the same, the encoding device randomly selects the coding coefficient vector for the first IP data packet and the second IP data packet in the finite field;

In an actual network, the encoding device usually receives IP data packets from the first source and the second source from different links, and of course, it may also receive the first source and the second letter from the same link. Source IP packet.

In an actual application, the encoding device may randomly select a coding coefficient vector for the first IP data packet and the second IP data packet, respectively, in a Galois Field (GF, Galois Field) 2 ⁿ , where the value range of the n For example, an integer between 4 and 16 (for example, an integer between 8 and 16), or n may be an integer greater than 16, and the larger the n, the greater the decoding probability.

303. The encoding device performs random linear coding on the first IP data packet and the second IP data packet by using the selected coding coefficient vector to obtain a first linearly encoded data packet.

304. The encoding device encapsulates the first linear encoded data packet into a dual source network coding protocol DNCP to obtain a first DNCP data packet.

The DNCP header of the first DNCP data packet includes an identifier of the first source, an identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and a coding coefficient related to the coding coefficient vector. Vector information (wherein the coding coefficient vector information related to the above-mentioned coding coefficient vector may be the coding coefficient vector itself, or may be the coding coefficient vector information obtained by passing the above-mentioned coding coefficient vector through a specific calculation, and the decoding device may pass, for example, The inverse of the specific calculation corresponds to the coding coefficient vector.

305. The encoding device performs IP encapsulation on the first DNCP data packet to obtain a third IP data packet and sends the data packet.

In an actual application, the source address recorded in the IP header of the third IP packet is, for example, the same as the source address recorded in the IP header of the first IP packet or the second IP packet, and the IP header of the third IP packet is in the IP header. The recorded destination address may be, for example, the same as the destination address recorded in the IP header of the first IP data packet or the second IP data packet (where the first IP data packet and the IP header of the second IP data packet are recorded in the same IP header. Destination address). In addition, the third IP packet can also be included in the IP header. Contains the DNCP packet identifier (the DNCP packet identifier, for example, a protocol field that can be carried in the IP header of the third IP packet, for example, its protocol field can be padded with FE or other specific value to indicate that its IP layer payload contains DNCP packets) In this way, the other network node can directly know that the IP layer payload of the third IP data packet includes the DNCP data packet according to the DNCP data packet identifier included in the IP packet header of the third IP data packet.

In addition, if the encoding device does not receive the same IP data packet from the second source as the destination address of the first IP data packet within the set duration after receiving the first IP data packet, the encoding device may send the first An IP data packet, or the first IP data packet is DNCP-encapsulated to obtain a second DNCP data packet, and the second DNCP data packet is IP-encapsulated to obtain a fourth IP data packet, and the DNCP packet header of the second DNCP data packet is sent. An identifier of the first source and an identification of the first IP packet may be included. At this time, the source address recorded in the IP header of the fourth IP packet is the same as the source address recorded in the IP header of the first IP packet, and the destination address and the first IP data recorded in the IP header of the fourth IP packet. The destination address recorded in the IP header of the packet is the same. In addition, the IP header of the fourth IP data packet may further include a DNCP data packet identifier (the DNCP data packet identifier may be, for example, a protocol field that may be carried in the IP header of the fourth IP data packet, for example, the protocol field may be filled with FE or Other specific values to indicate that their IP layer payload contains DNCP data packets), so that other network nodes can directly know the IP layer payload of the fourth IP data packet according to the DNCP data packet identifier included in the IP header of the fourth IP data packet. DNCP packet.

As can be seen from the above, this embodiment proposes DNCP to support codec between two sources. Since the encoding device performs random linear coding on two IP packets whose arrival interval is less than the set duration and the destination address is the same, this is advantageous. To some extent, the effect of mismatching problems such as transmission delay and transmission rate in the actual network is eliminated. Meanwhile, the DNCP header of the DNCP packet carries two source identifiers, two IP packet identifiers, and coding coefficients related to the coding coefficient vector. Vector information, which lays the foundation for the encoding device to continue encoding multiple times or for decoding the decoding device quickly, which is beneficial to improve the versatility of the network coding in the actual network; and because it is implemented based on the network coding technology, it can fully utilize the network Resources, significantly improve the throughput of the network, to achieve higher transmission Transmission efficiency.

Another embodiment of the encoding method of the present invention is directed to a scenario in which two encoded data packets are encoded. An encoding method may include: the encoding device receives the fifth internet protocol IP data packet; if the sixth IP data packet is received within a set time period after receiving the fifth IP data packet, and the sixth IP data packet and the fifth The destination address of the IP data packet is the same, and the IP layer payload of the sixth IP data packet and the fifth IP data packet both contain the DNCP data packet, and the DNCP packet header of the DNCP data packet in the IP layer payload of the sixth IP data packet is included. Two source identifiers and two IP packet identifiers, which are the same as the two source identifiers and the two IP packet identifiers contained in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet, are limited Randomly selecting a coding coefficient vector for the payload of the DNCP packet in the fifth IP data packet and the sixth IP data packet IP layer payload in the domain; using the coding coefficient vector, for the fifth IP data packet and the sixth IP data packet The payload of the DNCP packet in the IP layer payload is randomly linearly encoded to obtain a second linearly encoded data packet; the second linearly encoded data packet is DNCP-encapsulated to obtain a third DNCP data packet, and the third DNCP data packet The DNCP header of the packet includes the two source identifiers, the two IP packet identifiers, and the coding coefficient vector information related to the coding coefficient vector. The third DNCP packet is IP-encapsulated to obtain a seventh IP packet and sent.

Referring to Figure 4, specific steps may include:

401. The encoding device receives the fifth internet protocol IP data packet.

402. If the encoding device receives the sixth IP data packet within a set duration after receiving the fifth IP data packet, and the sixth IP data packet and the fifth IP data packet have the same destination address, and the sixth IP data packet And the IP layer payload of the fifth IP packet includes the DNCP packet, and the DNCP header of the DNCP packet in the IP layer payload of the sixth IP packet contains two source identifiers and two IP packet identifiers, and The DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet contains two source identifiers and the same two IP packet identifiers, and the encoding device is the fifth IP packet and the sixth in the finite field respectively. The payload of the DNCP packet in the IP layer payload of the IP packet is randomly selected from the coding coefficient vector; In an actual application, the encoding device can detect, for example, whether the IP header of the fifth IP data packet and the sixth IP data packet includes the DNCP data packet identifier, and if the IP packet header includes the DNCP data packet identifier, the IP layer payload can be obtained according to the identifier. The DNCP data packet is included. Of course, the encoding device can also know whether the IP layer payload includes the DNCP data packet by other possible methods. For example, the IP layer payload of the fifth IP data packet and the sixth IP data packet can be detected to obtain the IP address. Whether the layer payload contains DNCP packets.

The encoding device may randomly select a coding coefficient vector for the load of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload, respectively, in the Galois field GF2 ⁿ , where the value range of the n For example, an integer between 4 and 16 (for example, an integer between 8 and 16), or n may be an integer greater than 16.

403. The encoding device performs random linear coding on the payload of the DNCP data packet in the IP layer payload of the fifth IP data packet and the sixth IP data packet by using the selected coding coefficient vector to obtain a second linear encoded data packet.

404. The encoding device performs DNCP encapsulation on the second linearly encoded data packet to obtain a third DNCP data packet, where the DNCP packet header of the third DNCP data packet includes the foregoing two source identifiers, the foregoing two IP data packet identifiers, and the foregoing coding coefficient vector. Corresponding coding coefficient vector information (wherein the coding coefficient vector information related to the above coding coefficient vector may be the coding coefficient vector itself, or it may be that the coding coefficient vector and the coding coefficient vector information included in the DNCP packet of the DNCP packet are passed Data obtained after a specific calculation).

405. The encoding device performs IP encapsulation on the third DNCP data packet to obtain a seventh IP data packet and sends the data packet.

In practical applications, the source address recorded in the IP header of the seventh IP data packet is, for example, the same as the source address recorded in the IP header of the fifth IP data packet or the sixth IP data packet, and the IP header of the seventh IP data packet is in the IP header. The recorded destination address may be, for example, the same as the destination address recorded in the IP header of the fifth IP packet or the sixth IP packet (where the fifth IP packet and the sixth IP packet are recorded in the same IP header) Destination address); the IP header of the seventh IP packet may also contain DNCP Packet identification (DNCP packet identification, for example, may be carried in the protocol field of the IP header of the seventh IP packet, eg the protocol field may be padded with FE or other specific value to indicate that its IP layer payload contains DNCP packets), thus The other network node can directly know that the IP layer payload of the seventh IP data packet includes the DNCP data packet according to the DNCP data packet identifier included in the IP header of the seventh IP data packet.

In addition, the encoding device does not receive the sixth IP data packet having the same address as the destination address within the set duration after receiving the fifth IP data packet, or the IP layer of the sixth IP data packet received in the set duration The DNCP header of the DNCP packet in the payload contains two source identifiers and two IP packet identifiers, and two source identifiers and two DNCP headers of the DNCP packet in the IP layer payload of the fifth IP packet. If the IP packet identifiers are different, the encoding device can send the fifth IP packet.

As can be seen from the above, this embodiment proposes DNCP to support codec between two sources, because the encoding device is a DNCP packet in the IP layer payload of two IP data packets whose arrival interval is less than the set duration and the destination address is the same. The load is randomly linearly coded, which is beneficial to eliminate the influence of the mismatch problem such as transmission delay and transmission rate in the actual network to some extent; and the DNCP header of the DNCP packet encapsulated by the random linear coding carries two source identifiers. Two IP packet identifiers and coding coefficient vector information related to the randomly selected coding coefficient vector, which provides a basis for the downstream encoding device to continue encoding one or more times or for decoding the decoding device quickly, which is beneficial to improve network coding. The versatility in the actual network; and because it is implemented based on the network coding technology, it can fully utilize the network resources, significantly improve the throughput rate of the network, and achieve higher transmission efficiency.

Another embodiment of the encoding method of the present invention, this embodiment mainly corresponds to a scenario of encoding an unencoded IP data packet and an encoded IP data packet. The encoding method may include: the encoding device receives the fifth internet protocol IP data packet; if the eighth IP data packet is received within the set time period after receiving the fifth IP data packet, and the eighth IP data packet And the destination address of the fifth IP data packet is the same, and the eighth IP data packet and the fifth IP data packet are One IP layer payload contains a DNCP data packet, and the other IP data packet is one of the original IP data packets obtained by random linear coding to obtain the DNCP data packet payload, and the other IP data packet is respectively in the finite field. Randomly selecting a coding coefficient vector from the payload of the DNCP data packet; using the coding coefficient vector to randomly linearly encode the payload of the DNCP data packet and the other IP data packet to obtain a third linearly encoded data packet; The packet is subjected to DNCP encapsulation to obtain a fourth DNCP data packet, where the DNCP header of the fourth DNCP data packet includes two source identifiers, two IP packet identifiers, and coding coefficient vector information related to the selected coding coefficient vector, where The two source identifiers and the two IP packet identifiers are the same as the two source identifiers and the two IP packet identifiers included in the header of the DNCP packet in the IP layer payload; the fourth DNCP packet is IP-enabled. The package gets the eighth IP packet and sends it.

Referring to Figure 5, specific steps may include:

501. The encoding device receives a fifth internet protocol IP data packet.

502. The encoding device receives the eighth IP data packet within a set duration after receiving the fifth IP data packet, and the eighth IP data packet and the fifth IP data packet have the same destination address, and the eighth IP data packet And the IP layer payload of one of the fifth IP data packets includes a DNCP data packet, and the other IP data packet is one of original IP data packets obtained by random linear coding to obtain the DNCP data packet payload, and the encoding device is in a finite field Randomly selecting a coding coefficient vector for the load of the other IP data packet and the DNCP data packet;

In an actual application, the encoding device can detect, for example, whether the IP header of the fifth IP data packet and the eighth IP data packet includes the DNCP data packet identifier, and if the IP packet header includes the DNCP data packet identifier, the IP layer payload can be known according to the data packet header. Include DNCP data packets. Of course, the encoding device can also know whether its IP layer payload contains DNCP data packets by other possible means. For example, the IP layer payload of the fifth IP data packet and the eighth IP data packet can be detected to obtain the IP address. Whether the layer payload contains DNCP packets.

The encoding device may randomly select a coding coefficient vector in the Galois field GF2 ⁿ for the load of the DNCP data packet and the random selection of the coding coefficient vector of the other IP data packet, respectively. The value range of n is, for example, an integer between 4 and 16 (for example, an integer of 8 to 16), or n may be an integer greater than 16.

503. The encoding device randomly encodes the payload of the DNCP data packet and the another IP data packet by using a randomly selected coding coefficient vector to obtain a third linear encoded data packet.

504. The encoding device performs DNCP encapsulation on the third linearly encoded data packet to obtain a fourth DNCP data packet. The DNCP packet header of the fourth DNCP data packet includes two source identifiers, two IP packet identifiers, and selected coding coefficients. Vector related coding coefficient vector information, wherein the two source identifiers and two IP packet identifiers, and the header of the DNCP packet in the IP layer payload include two source identifiers and two IP packet identifiers the same;

505. The encoding device performs IP encapsulation on the fourth DNCP packet to obtain a ninth IP data packet and sends the packet.

In an actual application, the source address recorded in the IP header of the ninth IP packet is, for example, the same as the source address recorded in the IP header of the fifth IP packet or the eighth IP packet, and the IP header of the ninth IP packet is in the IP header. The recorded destination address may be, for example, the same as the destination address recorded in the IP header of the fifth IP data packet or the eighth IP data packet (where the fifth IP data packet and the IP header of the eighth IP data packet are recorded in the same IP header. Destination IP address; the IP header of the ninth IP packet may also include a DNCP packet identifier (the DNCP packet identifier may be, for example, a protocol field that may be carried in the IP header of the ninth IP packet, for example, its protocol field may be filled with FE or Other specific values to indicate that their IP layer payload contains DNCP packets), so that other network nodes can directly know the IP layer payload of the ninth IP packet according to the DNCP packet identifier contained in the IP header of the ninth IP packet. DNCP packet.

As can be seen from the above, this embodiment proposes DNCP to support codec between two sources. If the IP packet arrival interval is less than the set duration and the destination address is the same, and the IP layer payload of one of the two IP packets is included a DNCP packet, and another IP packet is one of original IP packets obtained by random linear coding to obtain the DNCP packet payload, and the encoding device randomly linearizes the load of the other IP packet and the DNCP packet Coding, this is beneficial in To some extent, the effect of mismatching problems such as transmission delay and transmission rate in the actual network is eliminated. Meanwhile, the DNCP header of the DNCP packet encapsulated by the random linear coding carries two source identifiers, two IP packet identifiers, and coding coefficients. Vector information, which provides a basis for the downstream encoding device to continue encoding one or more times or for decoding the decoding device quickly, which is beneficial to improve the versatility of the network coding in the actual network; and since it is implemented based on the network coding technology, It can make full use of network resources, significantly improve network throughput and achieve higher transmission efficiency.

Another embodiment of the encoding method of the present invention, this embodiment primarily corresponds to another scenario of encoding an unencoded IP packet and an encoded IP packet. The encoding method may include: the encoding device receives the fifth internet protocol IP data packet; if the tenth IP data packet is received within the set time period after receiving the fifth IP data packet, and the tenth IP data packet And the destination address of the fifth IP data packet is the same, and the IP layer payloads of the tenth IP data packet and the fifth IP data packet both contain the DNCP data packet, and the IP layer of one of the tenth IP data packet and the fifth IP data packet The DNCP header of the DNCP packet in the payload contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains one of the two source identifiers and One of the two IP packet identifiers randomly selects a coding coefficient vector for the payload of the DNCP packet in the fifth IP packet and the tenth IP packet IP layer payload in the finite field; using the selected encoding a coefficient vector, performing random linear coding on the payload of the DNCP packet in the IP layer payload of the fifth IP packet and the tenth IP packet to obtain a fourth linearly encoded data packet; The encoded data packet is DNCP-encapsulated to obtain a fifth DNCP data packet, wherein the DNCP header of the fifth DNCP data packet includes the above two source identifiers and two IP packet identifiers, and a coding coefficient associated with the randomly selected coding coefficient vector. Vector information; IP encapsulation of the fifth DNCP packet to obtain the eleventh IP packet and sent.

Referring to Figure 6, specific steps may include:

601. The encoding device receives the fifth internet protocol IP data packet.

602. The encoding device receives the tenth in the set duration after receiving the fifth IP data packet. IP data packet, and the destination address of the tenth IP data packet and the fifth IP data packet are the same, and the IP layer payloads of the tenth IP data packet and the fifth IP data packet both contain the DNCP data packet, and the tenth IP data packet and The DNCP header of the DNCP packet in the IP layer payload of one of the fifth IP data packets contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains the One of the two source identifiers and one of the two IP packet identifiers, the encoding device is the DNCP packet in the fifth IP packet and the tenth IP packet IP layer payload in the finite field respectively The load is randomly selected from the coding coefficient vector;

In an actual application, the encoding device can detect, for example, whether the IP header of the fifth IP data packet and the tenth IP data packet includes the DNCP data packet identifier, and if the IP packet header includes the DNCP data packet identifier, the IP layer payload can be known according to the data packet header. Include DNCP data packets. Of course, the encoding device can also know whether its IP layer payload contains DNCP data packets through other possible methods. For example, the IP layer payload of the fifth IP data packet and the tenth IP data packet can be detected to obtain the IP address. Whether the layer payload contains DNCP packets.

For example, the encoding device may randomly select a coding coefficient vector for the load of the DNCP data packet in the fifth IP data packet and the tenth IP data packet IP layer payload in the Galois field GF2 ⁿ , for example, the value range of the n is An integer between 4 and 16, or n may also be an integer greater than 16.

603. The encoding device performs random linear coding on the payload of the DNCP data packet in the IP layer payload of the fifth IP data packet and the tenth IP data packet by using the coding coefficient vector to obtain a fourth linearly encoded data packet.

604. The encoding device performs DNCP encapsulation on the fourth linearly encoded data packet to obtain a fifth DNCP data packet, where the DNCP packet header of the fifth DNCP data packet includes the foregoing two source identifiers, the foregoing two IP data packet identifiers, and the foregoing coding coefficient vector. Correlated coding coefficient vector information;

605. The encoding device performs IP encapsulation on the fifth DNCP packet to obtain the eleventh IP data packet and sends the packet.

In practical applications, the source address recorded in the IP header of the eleventh IP packet is, for example, the same as the source address recorded in the IP header of the fifth IP packet or the tenth IP packet, and the eleventh IP data. The destination address recorded in the IP header of the packet may be, for example, the same as the destination address recorded in the IP header of the fifth IP packet or the tenth IP packet (wherein the IP header of the fifth IP packet and the tenth IP packet) The same destination address is recorded); the IP header of the eleventh IP packet may also contain a DNCP packet identifier (the DNCP packet identifier, for example, a protocol field that can be carried in the IP header of the eleventh IP packet, such as an agreement The field can be filled with FE or other specific values to indicate that its IP layer payload contains DNCP packets. Thus, other network nodes can directly learn the eleventh according to the DNCP packet identifier contained in the IP header of the eleventh IP packet. The IP layer payload of the IP packet contains the DNCP packet.

As can be seen from the above, this embodiment proposes DNCP to support codec between two sources. If two IP packets arrive at an interval less than the set duration and the destination address is the same, and both IP packets contain DNCP packets, and The DNCP header of the DNCP packet in one IP layer payload contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains the two source identifiers. One of the two IP packet identifiers, and the encoding device randomly and linearly encodes the payload of the DNCP packet in the IP layer payload of the two IP packets, which is beneficial to some extent The impact of the mismatch problem such as the transmission delay and the transmission rate in the actual network; meanwhile, the DNCP header of the DNCP packet encapsulated by the random linear coding carries two source identifiers, two IP packet identifiers, and a code coefficient vector. Coding coefficient vector information, which lays the foundation for the downstream encoding device to continue encoding one or more times or for decoding the decoding device quickly. Help to improve the versatility of the network coding in real networks; and because it is implemented based on network coding, which can be full use of network resources, significantly improve network throughput, achieve higher transmission efficiency.

An embodiment of the decoding method of the present invention, a decoding method, may include: the decoding device receives an IP data packet; if the IP layer payload of the received IP data packet includes a DNCP data packet, the decoding device stores the DNCP data packet To the buffer area; when the DNCP data packet stored in the buffer area reaches a preset number, the DNCP data packet is read from the buffer area; if the DNCP packet header of the read DNCP data packet contains two source identifiers, two IP packet identification and a set of encodings Coefficient vector related coding coefficient vector information, then it is found in the buffer area that another DNCP packet whose DNCP header contains the two source identifiers and two IP packet identifiers, if present, is read out from the buffer area The other DNCP packet, wherein the DNCP header of the another DNCP packet further includes another set of coding coefficient vector information related to the coding coefficient vector; and the read DNCP packet is obtained by using the two sets of coding coefficient vector information The load is subjected to random linear decoding.

Referring to Figure 7, specific steps may include:

701. The decoding device receives an IP data packet.

The decoding device may, for example, receive an IP data packet from the first link, or may also receive an IP data packet from the first link and the second link, respectively.

702. If the IP layer payload of the received IP data packet includes a DNCP data packet, the decoding device stores the DNCP data packet into the buffer area.

In an actual application, the encoding device can detect, for example, whether the IP header of the received IP packet includes a DNCP packet identifier, and if the IP header includes the DNCP packet identifier, it can be learned that the IP layer payload includes the DNCP packet. Of course, the encoding device can also know whether the IP layer payload includes the DNCP data packet by other possible means. For example, the IP layer payload of the received IP data packet can be used to know whether the IP layer payload includes the DNCP data packet.

703. When the DNCP data packet stored in the buffer area reaches a preset number, the preset number may be specifically set according to a specific network scenario and a coding scenario, for example, may be set to 32, 64, 128, 256, or other numbers. When the decoding device reads the DNCP data packet from the buffer area;

In an actual application, after the DNCP data packet is stored in the buffer area, the decoding device may further index the DNCP data packet (the index may include, for example, at least one IP data included in the DNCP header of the DNCP data packet. a packet identifier, or at least one IP packet identifier included in the DNCP header of the DNCP packet, and at least one source identifier, or may also include other information that can be indexed to the DNCP packet) and the DNCP packet is The address of the buffer is stored in the Content Addressable Memory (CAM). 704. If the DNCP header of the read DNCP data packet includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, the decoding device searches whether the buffer is stored in the buffer area. The DNCP header contains the two source identifiers and another DNCP packet identified by two IP packets;

Wherein, if the DNCP header of the read DNCP packet includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information, indicating that the payload of the DNCP packet is randomly linearly encoded, of course, The decoding device may also detect whether the payload of the DNCP packet is randomly linearly encoded by detecting a flag field of the DNCP packet header indicating the encoding of the payload of the DNCP packet, and if it is subjected to random linear coding, it needs to be searched. The decoding is performed to other DNCP packets having the same data to be decoded as the DNCP packet (ie, the DNCP header of the DNCP packet containing the two source identifiers and the two DNCP packets identified by the IP packet).

In addition, if the load of the read DNCP data packet is not randomly linearly encoded (ie, the DNCP data packet may be obtained by directly DNCP encapsulating the original IP data packet), then the DNCP header of the DNCP data packet is removed. IP packet (this IP packet is the original IP packet).

705. If yes, the decoding device reads the another DNCP data packet from the buffer area, where the DNCP packet header of the another DNCP data packet may include another set of coding coefficient vector information related to the coding coefficient vector (where The other set of coding coefficient vector information related to the coding coefficient vector may be the same as or different from the coding coefficient vector information related to the coding coefficient vector included in the DNCP packet header of the previously read DNCP data packet) The above two source identifiers and two IP packet identifiers.

In an application scenario, if the decoding device receives the IP data packet from the first link and the second link, respectively, if the IP layer payload of the IP data packet received from the first link includes the DNCP data packet, The decoding device may store the DNCP data packet into the first buffer area; if the IP layer payload of the IP data packet received from the second link includes a DNCP data packet, the decoding device may The data packet is stored in the second buffer area, where the first buffer area and the second buffer area may be located on the same memory chip, or may be located in different memory chips. In this scenario, when the DNCP data packet stored in the first buffer area reaches a preset number and/or the DNCP data packet stored in the second buffer area reaches a preset number, or is stored in the first buffer area. And when the sum of the DNCP data packets of the second buffer area reaches a preset number, according to a predetermined reading order (for example, may be read according to the principle of preemption first, or may be read in order of storage address from low to high) The DNCP packet is read from the first buffer. If the DNCP header of the DNCP packet read from the first buffer includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information corresponding to the two IP packet identifiers, Determining whether there is a DNCP header in the second buffer according to at least one of the two source identifiers and at least one of the two IP packet identifiers, or according to at least one of the two IP packet identifiers Another DNCP packet containing the above two source identifiers and two IP packet identifiers (ie, another DNCP packet having the same data to be decoded), if not present, the decoding device may discard the read DNCP packet; if present, the decoding device reads the other DNCP packet from the buffer.

706. The decoding device performs random linear decoding on the load of the read DNCP data packet by using two sets of coding coefficient vector information.

In practical applications, the decoding device can perform random linear decoding on the load of the read DNCP packet by, for example, matrix multiplication, Gaussian elimination, or Kramer's law. If the random linear decoding succeeds, the decoding device can obtain two original IP data packets from different sources and can send the original IP data packet. If the random linear decoding fails, the decoding device can discard the read out data. The DNCP packet can continue to read other DNCP packets from the buffer for decoding according to the above mechanism.

It can be seen that, in this embodiment, the decoding device receives the IP data packet; if the IP layer payload of the received IP data packet includes the DNCP data packet, the DNCP data packet is stored in the buffer area; when the DNCP is stored in the buffer area When the data packet reaches the preset number, the DNCP data packet is read out from the buffer area; if the DNCP packet header of the read DNCP data packet contains two source identifiers and two IP numbers According to the packet identifier and a set of coding coefficient vector information related to the coding coefficient vector, it is found that another DNCP packet whose DNCP header contains the two source identifiers and two IP packet identifiers in the buffer area, the other The DNCP header of a DNCP packet further includes another set of coding coefficient vector information related to the coding coefficient vector; the decoding device performs random linear decoding on the payload of the read DNCP packet by using the two sets of coding coefficient vector information. Since the DNCP header of the DNCP packet encapsulated by the encoding apparatus after random linear coding carries two source identifiers, two IP packet identifiers, and coding coefficient vector information related to the coding coefficient vector, the decoding device can quickly decode according to the decoding device. The mechanism is beneficial to improve the versatility of the network coding in the actual network; and because it is implemented based on the network coding technology, it can fully utilize the network resources, significantly improve the throughput rate of the network, and achieve higher transmission efficiency.

In order to facilitate a better understanding of the technical solution of the embodiments of the present invention, a codec process in the example scenario is taken as an example for detailed description.

Referring to Figure 8-a, Figure 8-a is a schematic diagram of the coding process. The specific steps may include:

801, the encoding device Eel receives the data packet P _S11 from the source S1;

802, the encoding device Eel determines whether the received data packet is an IP data packet; if not, then step 809;

If yes, go to step 803;

803. The encoding device Eel allocates an identifier to the received data packet P _S11 .

In practical applications, in order to facilitate easy identification of IP packets of the same source during decoding, and to easily recover the original IP packet sequence after decoding, the encoding device Eel can allocate each IP packet received. Identification (such as sequential numbering).

For convenience of description, the identifier assigned by the encoding device Eel to the IP data packet is defined as follows: 殳 In the topology map G=(V, E) of a certain network, for example, there are n source SVs, m sinks Rj ev, Where i=l,2,...! 1 , j=l, 2, — m.

For example, the encoding device Eel can record the first data packet sent by the source S1 as S(1, 0), the second data packet as S(l, l), and so on, and the Xth data of the source Sn. The package is denoted as S(n, x-1). According to According to the actual application requirements, the sequence number of the IP data packet from each source can end, for example, from 0 to 1024 (or other value), and repeats accordingly, that is, the maximum sequence number of the source Sn is S(n, 1024).

804. The encoding device Ec 1 starts a timer.

The purpose of starting the timer is to determine whether the number of packets arriving at the confidence source S2 can be received within the set duration thereafter.

805. If the encoding device Eel receives the data packet of the self-confidence source S2 (for example, represented as the data packet P _S21 ) before the timer expires, and the data packet is an IP data packet, the encoding device Eel determines the IP data packet. Whether the destination address of P _S11 and IP data packet P _S21 are the same;

If they are the same, step 809 is performed;

If it is different, step 806 is performed.

If the packet P _{S21 of the} confidence source S2 is received before the timer expires, the IP packet P _S11 and the IP packet P _{S21 can be} regarded as "simultaneous arrival".

806, the encoding apparatus Eel limited domain are IP packets P _S11 and IP packet P _S21 randomly coded coefficient vectors, and the IP packet P _S11 and IP packet P _S21 linearly encoded with encoding coefficient vector randomly selected A linearly encoded data packet is obtained.

In an actual application, the encoding device Eel can, for example, randomly select a coding coefficient vector for the IP data packet P _S11 and the IP data packet P _S21 from the GF256 (or other GF).

For example, if the coding coefficient vectors selected for IP data packets S(l, x) and IP data packets S(2, y) from different sources are α, β, respectively, the linearly encoded output out e) can be as follows:

= o S(l ,x)+ p S(2,y).

Furthermore, considering the case of multiple encoding and the actual application, the encoded object is, for example, the entire IP packet (IP header + its payload).

807. The encoding device Eel linearly encodes the IP data packet P _S11 and the IP data packet P _S21 to obtain a linear encoded data packet obtained by linearly coding the data packet, and performs DNCP encapsulation to obtain a DNCP data packet. The DNCP header of the giaDNCP data packet may include the identifiers of the source S1 and the source S2, the identifiers of the IP data packet P _S11 and the IP data packet P _S11 , and the randomly selected coding coefficient vector.

808. The encoding device Eel performs IP encapsulation on the obtained DNCP data packet to obtain an IP data packet PDSII;

The format of the IP data packet P _DS11 obtained by the IP encoding of the DNCP data packet by the encoding device Eel can be as shown in FIG. 8-b.

In practical applications, for example, the larger value in the total length field in the IP header of the IP packet P _S11 and the IP packet P _S21 may be filled in the total length field of the IP header of the IP packet P _DS11 , and the IP data may be package P _S11 and P _S21 IP packet IP header TTL field values into larger TTL field PDSIIIP IP packet header, and the source IP address and destination IP address of the IP packet P _DS11, the IP packet The source IP address and destination IP address of the Psii or IP packet P _S2 i are the same.

In addition, the IP packet P _DS11 IP header may also contain a DNCP packet identifier (the DNCP packet identifier may be, for example, a protocol field that may be carried in the IP packet P _DS11 IP header, for example, the protocol field may be filled with FE or other specific value to indicate that the IP layer packet payload contains DNCP), thus, other network node may «DNCP packet identification included in the IP header of the IP packet directly from P _DS11, known IP layer packet payload contains P _DS1 々IP DNCP data pack.

809. Encoding device Eel's IP layer lookup table forwards the data packet.

Referring to FIG. 9, FIG. 9 is a schematic flowchart of a decoding method, and specific steps may include:

901. The decoding device Del (for example, a decoding router) respectively receives IP data packets from two links; 902. The decoding device Del determines whether the received IP data packet IP layer payload includes a DNCP data packet.

If no, go to step 910;

If yes, proceed to step 903;

In an actual application, the decoding device Del can detect, for example, whether the IP header of the received IP packet contains a DNCP packet identifier, and if its IP header contains a DNCP packet identifier, it can be learned that the IP layer payload includes the DNCP packet. , of course, the decoding device Del can also pass its It is possible to know whether its IP layer payload contains DNCP packets. For example, by detecting the IP layer payload of the received IP packet, it is known whether its IP layer payload contains DNCP packets.

903. If the received IP data packet is from the first link, the decoding device Del may store the DNCP data packet in the IP layer payload of the IP data packet into the SRAM cache SR1, if the received IP data packet is from the second. Link, the decoding device Del can store the DNCP data packet in the IP layer payload of the IP data packet into the SRAM buffer SR2;

In an actual application, after the DNCP data packet is stored in the buffer area, the decoding device Del may also index the DNCP data packet (the index may include, for example, at least one IP included in the DNCP packet header of the DNCP data packet. The data packet identifier, or at least one IP data packet identifier included in the DNCP header of the DNCP data packet and at least one source identifier, or may also include other information that can be indexed to the DNCP data packet) and the DNCP data packet The address in the buffer area is stored in the Content Addressable Memory (CAM).

Among them, the reason for using CAM and storing the cache SRAM is that the encoded data packets received by the decoding device may be out of order and not synchronized due to the non-ideal coding and network environment. Therefore, the input DNCP data packet needs to be buffered first in the decoding device. The decoding operation is performed when a sufficient amount of information is collected (for example, the number of buffered DNCP packets reaches a set number).

The order in which DNCP packets are stored in the SRAM can be from the lowest to the highest order of the addresses.

The output of the CAM can be the address corresponding to the DNCP packet in the SRAM.

904. The decoding device D11 reads a DNCP data packet buffered in one of the SRAMs; wherein, for example, when the DNCP data packet stored in one of the SRAMs reaches a preset number and/or the DNCP data packet stored in another SRAM reaches At the preset number, or when the sum of the DNCP packets stored in the two SRAMs reaches a preset number, the decoding device Del can start the decoding operation.

For the decoding device Del, the received DNCP data packet is divided into two cases: a. The payload of the DNCP data packet is not randomly linearly encoded (the decoding device Del can learn DNCP according to the content of the relevant field in the DNCP header of the DNCP data packet, for example. Packet payload Whether it is encoded), but it is packaged in DNCP;

Such a packet can be obtained by simply removing the DNCP header to obtain the original IP packet;

b. The payload of the DNCP packet is randomly linearly encoded, and the payload of the DNCP packet can be expressed as in the form described by equation (1):

aX ( 0, i ) + b Y ( 1 , j ) = K

cX ( 0, i ) + d Y ( 1 , j ) = P ( 1 )

When decoding, for example, a first-in-first-solution strategy can be extracted for the received DNCP data packet, and the IP data packet can be output in the decoding order.

In practical applications, the decoding device Del may, for example, store the DNCP data packets buffered in two SRAMs in turn according to the address size order (for example, the storage may be stored in order of low to high address), and the address may be low according to the address. In the high order, a DNCP packet is read from the buffer SR1 for random linear decoding, then a DNCP packet is read from the buffer SR2 for random linear decoding, and then a DNCP packet is read in the buffer SR1 for randomization. Linear decoding, and so on.

Among them, the purpose of reading two SRAMs in turn is to "deliver as much as possible" the incoming packets.

905. The decoding device Del determines whether the load of the read DNCP data packet is randomly linearly encoded.

If the payload of the read DNCP packet is not encoded, the original DN packet is removed by removing the DNCP header from the DNCP packet, and step 910 is performed;

If the load of the read DNCP data packet is encoded, step 907 is performed;

906. The decoding device Del determines whether the read DNCP data packet has been decoded. After each original IP data packet is decoded, the decoding flag register may be correspondingly marked. Therefore, the decoding device Del may query and decode. A flag register to determine whether the read DNCP packet has been decoded;

Wherein, if the read DNCP data packet has been decoded, the decoding device Del may be paired The flag bit should be reset, and step 904 is executed to continue reading the next DNCP packet for decoding.

If the read DNCP packet is not decoded, execute 907;

Among them, the reason to query whether the data packet is decoded is:

The encoded packet form can always be represented by the equation group (1), that is, each SRAM can store data to be decoded corresponding to the original IP packet X (0, i), Y (1, j). Therefore, if the original IP data packet X (0, i), Y (l, j) has been solved based on a certain SRAM, and the DNCP data packet to be decoded is read when reading the next SRAM next time, It is still possible to read out the DNCP data packet to be decoded corresponding to the original IP data packet X (0, i), Y (1, j), and if it is queried that it has been decoded, it may not need to decode it. It is.

907. The decoding device Del searches, according to the DNCP header information of the read DNCP packet, whether another DNCP data packet having the same information to be decoded exists in another SRAM;

If so, the decoding device Del reads another DNCP packet stored by another SRAM having the same information to be decoded as the read DNCP packet.

The two source identifiers and the two IP packet identifiers included in the DNCP header of the two DNCP packets having the same information to be decoded are the same. For example, the DNCP header of a certain DNCP packet contains the identifiers of the source S1 and the source S2, and the identifier of the packet P _sl oIP packet P _S21 can be considered as having the payload of the two DNCP packets. The same information to be decoded.

The coding information of the two DNCP data packets is included in the DNCP header (see the DNCP header format description for details). For example, the IP packet X (0, 0) is indexed, or the identifier of the IP packet X (0, i) and the identifier of its corresponding source are indexed, and the IP packet X (0, can be detected by the CAM. 0 corresponds to the storage address of the DNCP packet in another SRAM, and the other DNCP packet, cX (0, i) + d Y (1, j), can be read and read according to the storage address.

908. The decoding device Del determines the pair of data to be decoded in the two DNCP data packets that are read out. Whether the vector of the coding coefficient should be linearly related; ― , whether abl is zero.

If the coefficient matrix is zero, the coefficient is linearly correlated. Currently, the data packet cannot be decoded, and the data packet can be discarded. Returning to step 904, the decoding of the next DNCP data packet is continued.

If the coefficient matrix is not zero, it indicates that the linearity is irrelevant and can be decoded. Step 909 is performed. 909. Decoding device Del uses Kramer's law to decode the original IP data packet obtained by reading the data to be decoded in the two DNCP data packets;

There are various decoding methods based on equations, such as matrix multiplication, Gaussian elimination, etc. Considering the implementation of hardware, a decoding algorithm based on Kramer's law is used here. ≠o, then,

p

― .― . a b

After the decoding is completed, the IP packet X ( 0, i ) and the IP packet Y ( l , j ) can be marked as solved.

910, decoding device Del forwards IP packets in the IP layer lookup table.

Returning to step 904, the decoding of the next DNCP packet continues.

In order to facilitate the implementation of the solution of the embodiment of the present invention, the hardware structure of the router DNCP layer with network codec function is also provided below.

To implement the network codec function in the network (and corresponding to the two routers respectively), that is, to change the simple "storage-forward" processing mode of the ordinary router, it is necessary to add a DNCP layer module to the ordinary router.

The DNCP hierarchy is located between the network layer and the transport layer, that is, the block diagram of the changed router architecture is shown in Figure 10. The hardware structure of the entire DNCP layer is shown in Figure 11 and Figure 12. (1) The DNCP layer module that implements the encoding function is described as follows:

Input the Arbitration Module (Input-Arbiter) and set it on the IP layer.

Since the encoding operation is only performed for the IP data packet, the transmitted data packet can be protocol-determined in the Input-Arbiter, and if it is an IP data packet, it is transmitted to the DNCP layer module, and the encoding operation is performed if the condition is satisfied;

If it is a non-IP packet, it is forwarded directly at the IP layer module.

Code Control Module (Control):

a. Extract the IP header information, assign a source identifier to each source, and assign an IP packet identifier to each IP packet (for example, sequence number 0 to 1024).

b. Determine if encoding is required:

1 Determine whether the time interval at which the packet arrives is less than the set duration.

2 If 1 is satisfied, it is judged whether it is a packet addressed to the same destination IP address.

If yes, it is sent to the encoding module (Encoding) for random linear encoding operation, otherwise it is sent directly to the packing module (Packing) for DNCP encapsulation, plus DNCP header.

The encoding module (Encoding) performs the encoding operation.

Encoding module Encoding can include:

The random number generation module, two multipliers, two full adders, and a random number generator (for example, an 8-bit random number). Wherein, when an IP data packet arrives, the random number generator can generate two random numbers, respectively multiply the IP data packets from a certain source, and finally add the obtained results to the packing module (Packing). .

Packing module: Packing the DNCP packet for the linearly encoded data packet obtained by random linear coding to obtain the DNCP data packet, and then encapsulating a new IP packet header for the DNCP data packet to obtain the IP data packet, which is carried by the encapsulated DNCP header and the IP header. For information, see the description of the above embodiment.

The output queue module (Output-Queue) queues the IP packets obtained by the packet module and sends them to the IP layer for table lookup forwarding. (2) The DNCP layer module that implements the decoding function is described as follows:

Input Arbitration Module ( Input_Arbiter ): Located at the IP layer.

Since the decoding operation is only performed for the DNCP data packet, the data packet of the transmitted data packet is judged in the Input-Arbiter to determine whether the payload of the IP data packet received by the IP layer includes the DNCP data packet, for example, when the IP packet IP header protocol field is When OX FE (indicating that the payload of the IP packet includes a DNCP packet), the input arbitration module transmits the DNCP packet in the IP packet payload to the DNCP layer; if the payload of the IP packet does not include the DNCP packet, Look up table forwarding at the IP layer.

SRAM—Control and CAM Control, complete read and write control of SRAM and CAM. Decoding module: Decodes when the amount of information buffered is sufficient (for example, when the DNCP packet buffered in the SRAM reaches a set number).

The decoding flag module (Decoded_flag), for example, the decoding flag position 1 (or 0) corresponding to the decoded original IP data packet, indicates that the original IP data packet has been decoded. When the decoded flag bit corresponding to the decoded packet has been queried, it can be reset to 0 (or 1).

The output queue module (Output-Queue) sorts the original IP data packets obtained by decoding the decoding module and sends them to the IP layer for table lookup forwarding.

It should be noted that the foregoing exemplary hardware structure is only one of many possible codec router hardware structures capable of implementing the codec function of the embodiment of the present invention, and those skilled in the art may, based on the example, codec the hardware structure of the router, Obtaining a variety of other codec router hardware structures, no longer here - repeat.

The embodiment of the present invention also introduces an implementation process in a specific application scenario.

For example, in a cell network, considering the difficulty of the decoding operation, only the codec problem between the two sources is considered first. A typical topology of a cell tree network is shown in Figure 13:

At the same time, sources S1 and S2 want to send IP packets (respectively labeled P1 and P2) to the same broadcast network. In the figure, there are 2 sources (S1, S2) and 10 sink nodes DCi (1 < i < 10). Obviously, if the intermediate node ECj ( lj 5 ) does not use network coding, it cannot The 10 sink nodes can receive the data of S1 and S2 at the same time (only some nodes can). If the network coding is used, the solution node can solve the equations so that each sink node can get the source S1 at the same time. And S2 data.

Further, considering that the cell network structure is relatively simple, in order to improve the probability of decoding, and to eliminate the influence of the rate difference of the source sending data packet, a buffer space is opened on each coding router, and data is obtained from the two sources. The packet is synchronized, which simplifies the coding and increases the probability of decoding. Its synchronization mechanism is for example:

For example, in the DNCP layer of the EC router, two pieces of RAM (such as 72bit 1024) are opened, and IP data packets from two sources are sequentially stored, and when one of the RAMs is full, the IP data in the two RAMs are simultaneously read from the lowest address. The packet is sent to its DNCP layer for encoding operations.

Encoding, S1 and S2, for example, can forward the data packet to ECj (1 j 5 ) through a switch, and judge the destination IP address and output path of the input IP data packet at the ECj (1 j 5 ) node, if it is found to be IP packets destined for the same destination address, and the routing output has a bottleneck (that is, an output channel is competed), and then sent to the DNCP layer for random linear coding. The DNCP layer uses a pseudo-random number generation module to generate a coded coefficient vector in the GF256 domain, and linearly encodes two IP packets that arrive at the same time (i.e., the arrival time interval is less than the set duration).

Then, the DNCP header and the new IP header are generated as required. In the new IP header, the source address is, for example, a local IP address, and the destination IP address is a multicast address, and the protocol field of the IP header is FF.

At the same time, for example, the source S1 can be defined as the source 1 and the source S2 is defined as the source 2, and the corresponding source identifier is filled in the corresponding field of the DNCP header.

It can be foreseen that the results obtained after the ECj (Kj < 5 ) node coding operation is completed are: ajPl+bjP2 ( 1 <j < 5 ) „

The decoding device uses two SRAMs to store the data packets coming from the two input paths, and can use the CAM as an index, when collecting a certain amount of information packets (such as 32, 64, 128 or other numbers) ), you can start the decoding operation. Taking the DC1 node as an example, if it receives two information of alp1+blP2 and a2Pl+b2P2, it can use Kramer which is more conducive to hardware implementation. The law is calculated to recover the information PI and P2. Since the entire IP data packet is encoded during encoding, the decoded IP data packet (original IP data packet) can be directly forwarded.

As can be seen from the above, this embodiment proposes DNCP to solve the codec problem between two sources. Since the encoding device performs random linear coding on two IP data packets whose arrival interval is less than the set duration and the destination address is the same, this is advantageous. To some extent, the effect of mismatching problems such as transmission delay and transmission rate in the actual network is eliminated. Meanwhile, the DNCP header of the DNCP packet carries two source identifiers, two IP packet identifiers, and a randomly selected coding coefficient vector. The decoding device can be quickly decoded according to the same, and the codec mechanism is beneficial to improve the versatility of the network coding in the actual network; and because it is implemented based on the network coding technology, the network resources can be fully utilized, and the throughput of the network is obviously improved. Higher transmission efficiency.

It should be noted that, for the foregoing method embodiments, for the sake of simple description, they are all expressed as a series of action combinations, but those skilled in the art should understand that the present invention is not limited by the described action sequence. Because in accordance with the present invention, certain steps may be performed in other sequences or concurrently. Further, those skilled in the art should also understand that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present invention.

In order to facilitate the better implementation of the above described embodiments of the embodiments of the present invention, related apparatus for implementing the above schemes are also provided below.

Referring to FIG. 14, an encoding device 1400 according to an embodiment of the present invention may include: a first receiving module 1410, configured to receive a first Internet Protocol IP data packet from a first source; a first encoding coefficient selecting module 1420, For receiving the second IP data packet from the second source within the set duration after the first receiving module 1410 receives the first IP data packet, and the purpose of the first IP data packet and the second IP data packet If the addresses are the same, the coding coefficient vector is randomly selected for the first IP data packet and the second IP data packet in the finite field;

In the actual network, the first receiving module 1410 usually receives IP data packets from the first source and the second source respectively from different links, and of course, it may also receive the first source and the same link from the same link. IP packet of the second source. In a practical application, the first coding coefficient selection module 1420 may randomly select a coding coefficient vector for the first IP data packet and the second IP data packet, respectively, in the GF2 ⁿ , where the value range of the n is, for example, 4 to 16. An integer between (for example, an integer between 8 and 16), or n can also be an integer greater than 16, and the larger the n, the greater the decoding probability.

The encoding module 1430 is configured to perform random linear coding on the first IP data packet and the second IP data packet by using a coding coefficient vector randomly selected by the first coding coefficient selection module 1420 to obtain a first linearly encoded data packet;

The DNCP encapsulation module 1440 is configured to perform DNCP encapsulation on the first linearly encoded data packet to obtain a first DNCP data packet, where the DNCP packet header of the first DNCP data packet includes an identifier of the first source, an identifier of the second source, and a first IP address. An identifier of the data packet, an identifier of the second IP data packet, and coding coefficient vector information related to the above-described coding coefficient vector;

The IP encapsulation module 1450 is configured to perform IP encapsulation on the first DNCP data packet obtained by the DNCP encapsulation module 1440 to obtain a third IP data packet and send the same.

The DNCP packet header of the first DNCP data packet may include an identifier of the first source, an identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and the first coding coefficient selection module 1420. Randomly selected coding coefficient vector correlation coding coefficient vector information (where the coding coefficient vector information related to the above coding coefficient vector may be the coding coefficient vector itself, or may be an encoding obtained by passing the above coding coefficient vector through a specific calculation The coefficient vector information, and the decoding device can restore the coding coefficient vector by, for example, the inverse of the specific calculation.

In an actual application, the source address recorded in the IP header of the third IP packet is, for example, the same as the source address recorded in the IP header of the first IP packet or the second IP packet, and the IP header of the third IP packet is in the IP header. The recorded destination address may be, for example, the same as the destination address recorded in the IP header of the first IP data packet or the second IP data packet (where the first IP data packet and the IP header of the second IP data packet are recorded in the same IP header. Destination address). In addition, the IP header of the third IP data packet may further include a DNCP data packet identifier (the DNCP data packet identifier may be carried, for example, in the third IP data packet. The protocol field of the IP header, for example, may fill its protocol field with FE or other specific value to indicate that its IP layer payload contains DNCP packets, so that other network nodes can directly include the IP header according to the third IP packet. The DNCP packet identifies that the IP layer payload of the third IP packet contains the DNCP packet.

In an application scenario, the IP encapsulation module 1450 is further configured to: if the first receiving module 1410 receives the first IP data packet, the first IP data is not received from the second source. The IP packet with the same destination address of the packet transmits the first IP packet.

In another application scenario, the DNCP encapsulation module 1440 is further configured to: if the first receiving module 1410 receives the first IP data packet, does not receive the first IP from the second source within the set duration after receiving the first IP data packet. The IP packet with the same destination address of the data packet is DNCP-encapsulated to obtain the second DNCP data packet;

The IP encapsulation module 1450 is further configured to perform IP encapsulation on the second DNCP data packet to obtain a fourth IP data packet, where the DNCP packet header of the second DNCP data packet includes the identifier of the first source and the first IP data packet. Logo. At this time, the source address recorded in the IP header of the fourth IP packet may be the same as the source address recorded in the IP header of the first IP packet, and the destination address and the first IP recorded in the IP header of the fourth IP packet. The destination address recorded in the IP header of the packet is the same. In addition, the IP header of the fourth IP data packet may further include a DNCP data packet identifier (the DNCP data packet identifier may be, for example, a protocol field that may be carried in the IP header of the fourth IP data packet, for example, the protocol field may be filled with FE or Other specific values to indicate that their IP layer payload contains DNCP data packets), so that other network nodes can directly know the IP layer payload of the fourth IP data packet according to the DNCP data packet identifier included in the IP header of the fourth IP data packet. DNCP packet.

It can be understood that the coding device 1400 in this embodiment may be the coding device in the foregoing method embodiment, and the functions of the respective functional modules may be specifically implemented according to the method in the foregoing method embodiment, and the specific implementation process may refer to the foregoing method. The related description of the embodiment is not described herein again.

Referring to FIG. 15, an encoding device 1500 according to an embodiment of the present invention may include: The second receiving module 1510 is configured to receive the fifth Internet Protocol IP data packet, and the second encoding coefficient selecting module 1520 is configured to receive the first receiving module 1510 after receiving the fifth IP data packet. Six IP data packets, and the destination addresses of the sixth IP data packet and the fifth IP data packet are the same, and the IP layer payloads of the sixth IP data packet and the fifth IP data packet both contain the DNCP data packet, and the sixth IP data packet The DNCP header of the DNCP packet in the IP layer payload contains two source identifiers and two IP packet identifiers, and two letters contained in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet. If the source identifier and the two IP packet identifiers are the same, the coding coefficient vector is randomly selected for the payload of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload in the finite field;

In a practical application, the second coding coefficient selection module 1520 may randomly select a coding coefficient for the load of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload, respectively, in the Galois field GF2 ⁿ . The vector, where n ranges from 4 to 16 (for example, an integer between 8 and 16), or n may be an integer greater than 16.

The coding module 1530 is configured to perform random linear coding on the payload of the DNCP data packet in the IP layer payload of the fifth IP data packet and the sixth IP data packet by using the coding coefficient vector randomly selected by the second coding coefficient selection module 1520. Two linearly encoded data packets;

The DNCP encapsulating module 1540 is configured to perform DNCP encapsulation on the second linearly encoded data packet to obtain a third DNCP data packet, where the DNCP packet header of the third DNCP data packet includes the foregoing two source identifiers, the foregoing two IP packet identifiers, and the encoding. The coding coefficient vector information related to the coding coefficient vector randomly selected by the module 1530;

The IP encapsulation module 1550 is configured to perform IP encapsulation on the third DNCP data packet obtained by the DNCP encapsulation module 1540 to obtain a seventh IP data packet and send the same.

In practical applications, the source address recorded in the IP header of the seventh IP data packet is, for example, the same as the source address recorded in the IP header of the fifth IP data packet or the sixth IP data packet, and the IP header of the seventh IP data packet is in the IP header. The recorded destination address may be, for example, the same as the destination address recorded in the IP header of the fifth IP packet or the sixth IP packet (wherein the IP packet of the fifth IP packet and the sixth IP packet) The IP address header of the seventh IP data packet may also include a DNCP data packet identifier (the DNCP data packet identifier, for example, a protocol field that may be carried in the IP header of the seventh IP data packet, for example, The protocol field can be padded with FE or other specific values to indicate that its IP layer payload contains DNCP packets. Thus, other network nodes can directly learn the seventh IP according to the DNCP packet identifier contained in the IP header of the seventh IP packet. The IP layer payload of the packet contains the DNCP packet.

In addition, if the second receiving module 1510 does not receive the sixth IP data packet with the same destination address within the set duration after receiving the fifth IP data packet, or the sixth IP received in the set duration The DNCP header of the DNCP packet in the IP layer payload of the packet contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet contains two If the source identifier is different from the two IP packet identifiers, the IP encapsulation module 1550 can send the fifth IP data packet.

In an application scenario, the second encoding coefficient selection module 1520 is further configured to: if the second receiving module 1510 receives the eighth IP data packet within the set duration after receiving the fifth IP data packet, and the eighth IP The destination address of the data packet and the fifth IP data packet are the same, and the IP layer payload of one of the eighth IP data packet and the fifth IP data packet includes the DNCP data packet, and the other IP data packet is obtained by random linear coding to obtain the DNCP. One of the original IP data packets of the data packet payload, and the coding coefficient vector is randomly selected for the payload of the DNCP data packet and the other IP data packet in the finite field;

The encoding module 1530 is further configured to use, by using the second encoding coefficient selection module 1520, the payload of the DNCP data packet and the encoding coefficient vector randomly selected by another IP data packet in the finite field, the payload of the DNCP data packet, and the other An IP data packet is randomly linearly encoded to obtain a third linearly encoded data packet;

The DNCP encapsulating module 1540 is further configured to perform DNCP encapsulation on the third linearly encoded data packet obtained by the encoding module 1530 to obtain a fourth DNCP data packet, where the DNCP packet header of the fourth DNCP data packet includes two source identifiers and two IP data packets. Identification and selection with the second coding coefficient Obtaining coding coefficient vector information related to the coding coefficient vector randomly selected by the module 1520, wherein the two source identifiers and the two IP data packet identifiers and the two letters included in the DNCP header of the DNCP data packet in the IP layer payload The source identifier is the same as the two IP packet identifiers;

The IP encapsulation module 1550 is further configured to perform IP encapsulation on the fourth DNCP packet obtained by the DNCP encapsulation module 1540 to obtain a ninth IP data packet and send the same.

In an application scenario, the second encoding coefficient selection module 1520 is further configured to: if the second receiving module 1510 receives the tenth IP data packet within a set duration after receiving the fifth IP data packet, and the tenth IP The destination address of the data packet and the fifth IP data packet are the same, and the IP layer payloads of the tenth IP data packet and the fifth IP data packet both contain the DNCP data packet, and one of the tenth IP data packet and the fifth IP data packet The DNCP header of the DNCP packet in the IP layer payload contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains the two source identifiers. And one of the two IP data packet identifiers, and the coding coefficient vector is randomly selected for the payload of the DNCP data packet in the fifth IP data packet and the tenth IP data packet IP layer payload in the finite field;

The encoding module 1530 is further configured to use the second encoding coefficient selection module 1520 to load the DNCP data packet in the fifth IP data packet and the tenth IP data packet IP layer payload in the finite field respectively. Randomly selecting a coding coefficient vector, performing random linear coding on the payload of the PNCP data packet in the IP packet of the fifth IP data packet and the foregoing tenth IP data packet to obtain a fourth linearly encoded data packet;

The DNCP encapsulating module 1540 is further configured to perform DNCP encapsulation on the fourth linearly encoded data packet obtained by the encoding module 1530 to obtain a fifth DNCP data packet, where the DNCP packet header of the fifth DNCP data packet includes the foregoing two source identifiers, and the foregoing Two IP data packet identifiers are associated with a coding coefficient vector randomly selected by the second coding coefficient selection module 1520 in the finite field for the load of the DNCP data packet in the fifth IP data packet and the tenth IP data packet IP layer payload, respectively. Coding coefficient vector information;

The IP encapsulation module 1550 is further configured to perform IP encapsulation on the fifth DNCP packet obtained by the DNCP encapsulation module 1540 to obtain an eleventh IP data packet and send the IP packet.

In practical applications, the source address recorded in the IP header of the eleventh IP packet is, for example, the same as the source address recorded in the IP header of the fifth IP packet or the tenth IP packet, and the IP of the eleventh IP packet. The destination address recorded in the header may be, for example, the same as the destination address recorded in the IP header of the fifth IP packet or the tenth IP packet (where the IP header of the fifth IP packet and the tenth IP packet are recorded in the IP header) The same destination address); the IP header of the eleventh IP packet may also include a DNCP packet identifier (the DNCP packet identifier, for example, a protocol field that may be carried in the IP header of the eleventh IP packet, eg, the protocol field may be padded For FE or other specific values to indicate that its IP layer payload contains DNCP packets, so that other network nodes can directly know the eleventh IP packet according to the DNCP packet identifier contained in the IP header of the eleventh IP packet. The IP layer payload contains DNCP packets.

The finite field is, for example, a Galois field GF2 ⁿ , wherein the range of n is, for example, an integer between 4 and 16 (for example, an integer between 8 and 16), or n may be greater than 16. Integer.

It can be understood that the encoding device 1500 in this embodiment may be the encoding device in the foregoing method embodiment, and the functions of the respective functional modules may be specifically determined according to the method in the foregoing method embodiment. For the specific implementation process, reference may be made to the related description of the foregoing method embodiments, and details are not described herein again.

Referring to FIG. 16, a decoding device 1600 according to an embodiment of the present invention may include: a third receiving module 1610, configured to receive an IP data packet;

The first storage module 1620 is configured to: when the IP layer payload of the IP data packet received by the third receiving module 1610 includes the DNCP data packet, store the DNCP data packet into the buffer area;

The first reading module 1630 is configured to: when the DNCP data packet stored in the buffer area reaches a preset number, read the DNCP data packet from the buffer area;

The second reading module 1640 is configured to: if the DNCP packet of the DNCP packet read by the first reading module 1630 includes two source identifiers, two IP packet identifiers, and a set of encoding coefficients related to the encoding coefficient vector Vector information, then it is found in the buffer area that another DNCP packet whose DNCP header contains the above two source identifiers and two IP packet identifiers, if any, reads the other DNCP packet from the buffer area. The DNCP header of the another DNCP packet further includes another set of coding coefficient vector information related to the coding coefficient vector.

The decoding module 1650 is configured to perform random linear decoding on the load of the read DNCP data packet by using two sets of coding coefficient vector information.

In a practical application, the decoding module 1650 can perform random linear decoding on the load of the read DNCP packet by, for example, matrix multiplication, Gaussian elimination, or Clemme's law.

If the decoding module 1650 obtains the random linear decoding successfully, the decoding device 1600 can obtain two original IP data packets from different sources and can send the original IP data packet. If the decoding module 1650 fails to perform random linear decoding, the device can be lost. Discard the read DNCP packet.

In an application scenario, the decoding device 1600 may further include:

a second storage module, configured to store the index of the DNCP data packet and the address of the DNCP data packet in the buffer area into the content addressable after the first storage module 1620 stores the DNCP data packet into the buffer area In the memory CAM, the index includes: at least one IP packet identifier included in a DNCP header of the DNCP packet. In an application scenario, the third receiving module 1610 is specifically configured to receive IP data packets from the first link and the second link, respectively.

The first storage module 1620 is specifically configured to: if the IP layer payload of the IP data packet received by the third receiving module 1610 from the first link includes a DNCP data packet, store the DNCP data packet in the first buffer area; The IP layer payload of the IP data packet received by the third receiving module 1610 from the second link includes a DNCP data packet, and the DNCP data packet is stored in the second buffer area;

The first reading module 1630 is specifically configured to: when the DNCP data packet stored in the first buffer area reaches a preset number and/or the DNCP data packet stored in the second buffer area reaches a preset number, or When the sum of the DNCP packets entering the first buffer area and the second buffer area reaches a preset number, the DNCP data packet is read from the first buffer area in a predetermined reading order.

In an application scenario, the second reading module 1640 may be specifically configured to: if the first read mode, two IP data packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, according to the At least one of the two source identifiers and at least one of the two IP packet identifiers, or, according to at least one of the two IP packet identifiers, finding whether the DNCP header exists in the second buffer includes the above Two source identifiers and another DNCP packet identified by two IP packets, if present, the other DNCP packet can be read from the second buffer, wherein the DNCP of the other DNCP packet The header also contains another set of coding coefficient vector information associated with the coding coefficient vector.

It can be understood that the decoding device 1600 in this embodiment may be the decoding device in the foregoing method embodiment, and the functions of the respective functional modules may be specifically implemented according to the method in the foregoing method embodiment, and the specific implementation process may refer to the foregoing method. The related description of the embodiment is not described herein again.

Referring to FIG. 17, an encoding device 1700 according to an embodiment of the present invention may include: an IP layer module 1710 and a DNCP layer module 1720.

The IP layer module 1710 is configured to receive a first Internet Protocol IP data packet from the first source. The DNCP layer module 1720 is configured to receive, by the IP layer module, a second IP data packet from the second source, and the first IP data packet and the second IP data, within a set duration after receiving the first IP data packet. If the destination address of the packet is the same, the coding coefficient vector is randomly selected for the first IP data packet and the second IP data packet in the finite field; the first IP data packet and the second IP data packet are randomly linearized by using the coding coefficient vector. Encoding to obtain a first linearly encoded data packet; the first linearly encoded data packet is encapsulated by a dual source network coding protocol DNCP to obtain a first DNCP data packet, where a DNCP header of the first DNCP data packet includes an identifier of the first source, An identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and coding coefficient vector information related to the coding coefficient vector; and performing IP encapsulation on the first DNCP data packet to obtain a third IP data packet The above IP layer module is further configured to send the third IP data packet obtained by the DNCP layer module. In an application scenario, the IP layer module 1710 is further configured to receive a fifth Internet Protocol IP data packet.

The DNCP layer module 1720 is further configured to: if the IP layer module 1710 receives the sixth IP data packet within a set duration after receiving the fifth IP data packet, and the destination addresses of the sixth IP data packet and the fifth IP data packet Similarly, the IP layer payloads of the sixth IP data packet and the fifth IP data packet both contain a dual source network coding protocol DNCP data packet, and the DNCP header of the DNCP data packet in the IP layer payload of the sixth IP data packet includes two The source identifier and the two IP packet identifiers are the same as the two source identifiers and the two IP packet identifiers contained in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet, and then in the finite field The coding coefficient vector is randomly selected for the load of the DNCP data packet in the IP packet of the fifth IP data packet and the sixth IP data packet respectively; and is used in the IP layer payload of the fifth IP data packet and the sixth IP data packet in the finite field respectively. The coded coefficient vector randomly selected by the payload of the DNCP data packet is subjected to random linear coding of the payload of the DNCP data packet in the IP layer payload of the fifth IP data packet and the sixth IP data packet to obtain a second linear coded data packet; Linear coded packet The DNCP encapsulation obtains a third DNCP packet, and the DNCP header of the third DNCP packet includes the above two source identifiers, the above two IP packet identifiers, and the fifth IP packet and the sixth IP in the finite field respectively. DNCP number in packet IP layer payload Coding coefficient vector information related to the coding coefficient vector randomly selected according to the payload of the packet;

The DNCP data packet is IP-encapsulated to obtain a seventh IP data packet;

The IP layer module 1710 is further configured to send the seventh IP data packet obtained by the DNCP layer module 1720. In an application scenario, the DNCP layer module 1720 is further configured to: if the IP layer module 1710 receives the eighth IP data packet within a set duration after receiving the fifth IP data packet, and the eighth IP data packet and the first The destination addresses of the five IP data packets are the same, and the IP layer payload of one of the eighth IP data packet and the fifth IP data packet includes the DNCP data packet, and the other IP data packet is obtained by random linear coding to obtain the DNCP data packet payload. One of the original IP data packets, in the finite field, respectively, the load of the DNCP data packet and the other IP data packet are randomly selected from the coding coefficient vector; using the load of the DNCP data packet in the finite field and the above The other coding data vector randomly selected by the IP data packet is subjected to random linear coding of the payload of the DNCP data packet and the another IP data packet to obtain a third linearly encoded data packet; and the third linearly encoded data packet is encapsulated by DNCP. The fourth DNCP packet, the DNCP header of the fourth DNCP packet contains two source identifiers and two IP packet identifiers. And coding coefficient vector information related to the coding coefficient vector, where the two source identifiers and the two IP packet identifiers and the two source identifiers included in the DNCP header of the DNCP packet in the IP layer payload are The two IP data packets are identified by the same identifier; the fourth DNCP data packet is IP-encapsulated to obtain the ninth IP data packet.

The IP layer module 1710 is further configured to send the ninth IP data packet obtained by the DNCP layer module 1720. In an application scenario, the DNCP layer module 1720 is further configured to: if the IP layer module 1710 receives the tenth IP data packet within a set duration after receiving the fifth IP data packet, and the tenth IP data packet and the first The destination addresses of the five IP data packets are the same, and the IP layer payloads of the tenth IP data packet and the fifth IP data packet all contain the DNCP data packet, and the IP layer payload of one of the tenth IP data packet and the fifth IP data packet is included. The DNCP header of the DNCP packet contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains one of the above two source identifiers and the above two Among the IP packet identifiers One, the coding coefficient vector is randomly selected for the load of the DNCP data packet in the fifth IP data packet and the tenth IP data packet IP layer payload in the finite field; respectively, using the fifth IP data packet and the first in the finite field The above-mentioned coding coefficient vector randomly selected from the payload of the DNCP packet in the IP packet payload of the IP packet is randomly linearly encoded by the payload of the DNCP packet in the IP layer payload of the fifth IP packet and the tenth IP packet. a fourth linearly encoded data packet; the fourth linearly encoded data packet is DNCP-encapsulated to obtain a fifth DNCP data packet, where the DNCP header of the fifth DNCP data packet includes the two source identifiers, the two IP packet identifiers, and Encoding coefficient vector information related to the coding coefficient vector; performing IP encapsulation on the fifth DNCP packet to obtain an eleventh IP data packet;

The IP layer module 1710 is further configured to send the eleventh IP data packet obtained by the DNCP layer module 1720.

It can be understood that the encoding device 1700 in this embodiment may be the encoding device in the foregoing method embodiment (for example, the hardware structure shown in FIG. 11), and the functions of the respective functional modules may be specifically implemented according to the method in the foregoing method embodiment. For a specific implementation process, reference may be made to the related description of the foregoing method embodiments, and details are not described herein again.

Referring to FIG. 17, a decoding device 1800 according to an embodiment of the present invention may include: an IP layer module 1810 and a DNCP layer module 1820.

The IP layer module 1810 is configured to receive an Internet Protocol IP data packet.

a DNCP layer module, configured to store the DNCP data packet into the buffer area if the IP layer payload of the IP data packet received by the IP layer module 1810 includes the dual source network coding protocol DNCP data packet; When the DNCP data packet of the buffer area reaches a preset number, the DNCP data packet is read out from the buffer area; if the DNCP header of the DNCP data packet read above includes two source identifiers, two IP data packet identifiers, and one Grouping the coding coefficient vector information related to the coding coefficient vector, and searching whether there is another DNCP packet whose DNCP header contains the above two source identifiers and two IP packet identifiers in the buffer area, and if present, the slave buffer area Reading out another DNCP packet, where the DNCP header of the other DNCP packet Another set of coding coefficient vector information related to the coding coefficient vector is further included; the load of the read DNCP data packet is randomly linearly decoded by using the two sets of coding coefficient vector information.

The IP layer module 1810 is further configured to send the seventh IP packet that is randomly and linearly decoded by the DNCP layer module 1820.

It can be understood that the decoding device 1800 in this embodiment may be the decoding device in the foregoing method embodiment (for example, the hardware structure shown in FIG. 12), and the functions of the respective functional modules may be specifically implemented according to the method in the foregoing method embodiment. For a specific implementation process, reference may be made to the related description of the foregoing method embodiments, and details are not described herein again.

The embodiment of the invention further provides a communication system, including:

The encoding device and/or the decoding device in the above embodiments.

In the above embodiments, the descriptions of the various embodiments are different, and the details are not described in detail in an embodiment, and the related descriptions of other embodiments can be referred to.

In summary, the embodiment of the present invention proposes DNCP to support codec between two sources. Since the encoding device performs random linear coding on two IP data packets whose arrival interval is less than the set duration and the destination address is the same, which is beneficial to To some extent, the effect of mismatching problems such as transmission delay and transmission rate in the actual network is eliminated. Meanwhile, the DNCP header of the DNCP packet carries two source identifiers, two IP packet identifiers, and coding coefficients related to the coding coefficient vector. Vector information, which lays the foundation for the encoding device to continue encoding multiple times or for decoding the decoding device quickly, and the decoding device can quickly decode according to this, the mechanism based on DNCP is beneficial to improve the versatility of the encoding in the actual network; It is implemented based on network coding technology, so it can fully utilize network resources, significantly improve network throughput and achieve higher transmission efficiency.

A person skilled in the art can understand that all or part of the steps of the foregoing embodiments can be completed by a program to instruct related hardware. The program can be stored in a computer readable storage medium. The storage medium can include: Read-only memory, random access memory, disk or optical disk, etc.

Network codec method and related device and communication system provided by embodiment of the present invention It is to be noted that the description of the above embodiments is only for helping to understand the method of the present invention and its core ideas; at the same time, for those skilled in the art, according to the idea of the present invention, there will be changes in specific embodiments and applications. In conclusion, the contents of this specification are not to be construed as limiting the invention.

Claims

Rights request

1. An encoding method, comprising:

Receiving a first internet protocol IP data packet from the first source;

The first linearly encoded data packet is encapsulated by a dual source network coding protocol (DNCP) to obtain a first DNCP data packet, where the DNCP header of the first DNCP data packet includes an identifier of the first source, An identifier of the second source, an identifier of the first IP data packet, an identifier of the second IP data packet, and coding coefficient vector information related to the coding coefficient vector; performing IP encapsulation on the first DNCP data packet The third IP packet is obtained and sent.

2. The method according to claim 1, wherein the method further comprises: if the received time from the second source is not received within the set duration after receiving the first IP packet Transmitting the first IP data packet by using an IP data packet with the same destination address of the first IP data packet; or performing DNCP encapsulation on the first IP data packet to obtain a second DNCP data packet, where the second DNCP packet is used The data packet is IP-encapsulated to obtain a fourth IP data packet, and the DNCP packet header of the second DNCP data packet includes an identifier of the first source and an identifier of the first IP data packet.

The method according to claim 1, wherein the randomly selecting the coding coefficient vector for the first IP data packet and the second IP data packet in the finite field, respectively: A coding coefficient vector is randomly selected in the domain GF2n for the first IP data packet and the second IP data packet, where the value of the n is an integer between 8 and 16.

4. Method according to claim 1 or 3, characterized in that The source address recorded in the IP header of the third IP data packet is the same as the source address recorded in the IP header of the first IP data packet or the second IP data packet, where the third IP data packet is The destination address recorded in the IP header is the same as the destination address recorded in the IP header of the first IP packet or the second IP packet;

The IP header of the third IP packet further includes a DNCP packet identifier.

5. An encoding method, comprising:

Receiving a fifth internet protocol IP data packet;

And performing random linear coding on the payload of the DNCP data packet in the IP layer payload of the fifth IP data packet and the sixth IP data packet by using the coding coefficient vector to obtain a second linear coding data packet;

DNCP encapsulating the second linearly encoded data packet to obtain a third DNCP data packet, where the DNCP header of the third DNCP data packet includes the two source identifiers, the two IP packet identifiers, and Encoding coefficient vector information related to the coding coefficient vector; IP encapsulating the third DNCP data packet to obtain a seventh IP data packet and transmitting.

The method according to claim 5, wherein the method further comprises: receiving an eighth IP data packet within a set duration after receiving the fifth IP data packet, and the method The eight IP data packets and the fifth IP data packet have the same destination address, and the IP layer payload of one of the eighth IP data packet and the fifth IP data packet includes a DNCP data packet, and And another IP data packet is one of original IP data packets obtained by random linear coding to obtain the DNCP data packet payload, and the load of the DNCP data packet and the other IP data packet are respectively random in a finite field. Select a coding coefficient vector;

Randomly linearizing the load of the DNCP data packet and the another IP data packet by using the coded coefficient vector randomly selected by the payload of the DNCP data packet and the another IP data packet in a finite field Encoding to obtain a third linearly encoded data packet;

The third linearly encoded data packet is DNCP-encapsulated to obtain a fourth DNCP data packet, where the DNCP header of the fourth DNCP data packet includes two source identifiers, two IP data packet identifiers, and is associated with the coding coefficient vector. Coding coefficient vector information, wherein the two source identifiers and two IP packet identifiers and two source identifiers and two IP packets included in the DNCP header of the DNCP packet in the IP layer payload The same identifier;

The fourth DNCP data packet is IP-encapsulated to obtain a ninth IP data packet and sent.

The method according to claim 5, wherein the method further comprises: receiving a tenth IP data packet within a set duration after receiving the fifth IP data packet, and the method The ten IP data packet and the fifth IP data packet have the same destination address, and the tenth IP data packet and the fifth IP data packet IP layer payload both contain a DNCP data packet, and the tenth IP data packet The DNCP header of the DNCP packet in the IP layer payload of one of the packet and the fifth IP packet contains two source identifiers and two IP packet identifiers, and the other DNCP packet in the IP layer payload The DNCP header includes one of the two source identifiers and one of the two IP packet identifiers, and the fifth IP data packet and the tenth IP data packet are respectively in a finite field. The code of the DNCP packet in the IP layer payload is randomly selected from the coding coefficient vector;

Using the coding coefficient vector randomly selected for the load of the DNCP packet in the fifth IP data packet and the tenth IP data packet IP layer payload in the finite field, respectively, for the fifth IP data packet and The load of the DNCP data packet in the IP layer payload of the tenth IP data packet is randomly linearly encoded to obtain a fourth linearly encoded data packet; Performing DNCP encapsulation on the fourth linearly encoded data packet to obtain a fifth DNCP data packet, where the DNCP header of the fifth DNCP data packet includes the two source identifiers, the two IP packet identifiers, and the encoding Coefficient vector related coding coefficient vector information;

The fifth DNCP data packet is IP-encapsulated to obtain an eleventh IP data packet and sent.

8. A method according to any one of claims 5 to 7, characterized in that

The finite field is the Galois field GF2n, and the value of n ranges from 8 to 16 integers.

9. A decoding method, comprising:

Receiving Internet Protocol IP data packets;

The method according to claim 9, wherein the method further comprises: after the storing the DNCP data packet into the buffer area, further comprising:

And storing the index of the DNCP data packet and the address of the DNCP data packet in the buffer area into the content addressable memory CAM, wherein the index comprises: at least one included in a DNCP packet header of the DNCP data packet IP packet identification.

The method according to claim 9 or 10, comprising: Receiving the IP data packet includes:

Receiving IP data packets from the first link and the second link, respectively;

If the IP layer payload of the received IP data packet includes a DNCP data packet, the DNCP data packet is stored in the buffer area, including:

If the IP layer payload of the IP data packet received from the first link includes a DNCP data packet, the DNCP data packet is stored in the first buffer area; if the IP data packet is received from the second link The IP layer payload includes a DNCP data packet, and the DNCP data packet is stored in the second buffer area. When the DNCP data packet stored in the buffer area reaches a preset number, the DNCP data packet is read from the buffer area. Out DNCP packets, including:

When the DNCP data packet stored in the first buffer area reaches a preset number and/or the DNCP data packet stored in the second buffer area reaches a preset number, or is stored in the first When the sum of the DNCP data packets of the buffer area and the second buffer area reaches a preset number, the DNCP data packet is read from the first buffer area according to a predetermined reading order.

12. The method according to claim 11, comprising:

If the DNCP header of the read DNCP data packet includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, it is found in the buffer area whether it exists. The DNCP header contains the two source identifiers and another DNCP packet identified by two IP packets, including:

If the DNCP header of the DNCP packet read from the first buffer area includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, according to the two At least one of the source identifiers and at least one of the two IP packet identifiers, or, according to at least one of the two IP packet identifiers, finding whether the DNCP header includes the second buffer area Two source identifiers and another DNCP packet identified by two IP packets.

The method according to claim 9, wherein the method further comprises: removing the load of the read DNCP data packet without random linear coding The DNCP header of the DNCP packet gets the IP packet.

14. An encoding device, comprising:

And an encoding module, configured to perform random linear coding on the first IP data packet and the second IP data packet by using a coding coefficient vector randomly selected by the first coding coefficient selection module to obtain a first linearly encoded data packet;

15. The encoding device according to claim 14, wherein:

The IP encapsulating module is further configured to: if the first receiving module receives the first IP data packet from the second source, within a set duration after receiving the first IP data packet Sending the first IP data packet with the same IP data packet of the destination address;

Or,

The DNCP encapsulating module is further configured to: if the first receiving module receives the first IP data packet from the second source, within a set duration after receiving the first IP data packet The IP data packet with the same destination address, the DNCP encapsulation of the first IP data packet to obtain the second DNCP data packet; The IP encapsulation module is further configured to perform IP encapsulation on the second DNCP data packet to obtain a fourth IP data packet, where the DNCP packet header of the second DNCP data packet includes an identifier and a location of the first source The identifier of the first IP data packet.

16. An encoding device, comprising:

a DNCP encapsulating module, configured to perform DNCP encapsulation on a second linearly encoded data packet obtained by the encoding module to obtain a third DNCP data packet, where a DNCP packet header of the third DNCP data packet includes the two source identifiers, The two IP data packet identifiers and coding coefficient vector information related to the coding coefficient vector;

17. The encoding device according to claim 16, wherein:

The second coding coefficient selection module is further configured to: if the second receiving module receives the eighth IP data packet within a set duration after receiving the fifth IP data packet, and the eighth IP number And the destination address of the fifth IP data packet is the same, and the IP layer payload of one of the eighth IP data packet and the fifth IP data packet includes a DNCP data packet, and another IP data packet is passed Obtaining one of the original IP data packets of the DNCP data packet payload by random linear coding, and randomly selecting a coding coefficient vector for the payload of the DNCP data packet and the another IP data packet in the finite field;

The encoding module is further configured to: use the second coding coefficient selection module to respectively obtain, in the finite field, a load of the DNCP data packet and a coding coefficient vector randomly selected by the another IP data packet, and the DNCP data The payload of the packet and the other IP data packet are randomly linearly encoded to obtain a third linearly encoded data packet;

The DNCP encapsulation module is further configured to: perform DNCP encapsulation on the third linearly encoded data packet obtained by the coding module to obtain a fourth DNCP data packet, where the DNCP packet header of the fourth DNCP data packet includes two source identifiers, and two And an IP data packet identifier and coding coefficient vector information related to the coding coefficient vector, where the two source identifiers and the two IP data packet identifiers are the same as the two IP data packet identifiers;

The IP encapsulation module is further configured to perform IP encapsulation on the fourth DNCP data packet obtained by the DNCP encapsulation module to obtain a ninth IP data packet and send the same.

18. The encoding device according to claim 16, wherein:

The second coding coefficient selection module is further configured to: if the second receiving module receives the tenth IP data packet within a set duration after receiving the fifth IP data packet, and the tenth IP data packet and The destination address of the fifth IP data packet is the same, and the IP layer payload of the tenth IP data packet and the fifth IP data packet both contain a DNCP data packet, and the tenth IP data packet and the first The DNCP header of the DNCP packet in the IP layer payload of one of the five IP packets contains two source identifiers and two IP packet identifiers, and the DNCP header of the DNCP packet in the other IP layer payload contains the One of the two source identifiers and one of the two IP packet identifiers are respectively the fifth IP data packet and the tenth in the finite field The encoding of the DNCP data packet in the IP packet payload of the IP packet randomly selects a coding coefficient vector; the encoding module is further configured to: use the second encoding coefficient selection module to respectively be the fifth IP data packet in a finite field a randomly selected coding coefficient vector of a load of the DNCP data packet in the IP layer payload of the tenth IP data packet, and a DNCP data packet in the IP layer payload of the fifth IP data packet and the tenth IP data packet The load is subjected to random linear coding to obtain a fourth linearly encoded data packet;

The DNCP encapsulation module is further configured to perform DNCP encapsulation on the fourth linearly encoded data packet obtained by the coding module to obtain a fifth DNCP data packet, where the DNCP packet header of the fifth DNCP data packet includes the two sources Identification, the two IP data packet identifiers, and coding coefficient vector information related to the coding coefficient vector;

The IP encapsulation module is further configured to perform IP encapsulation on the fifth DNCP data packet obtained by the DNCP encapsulation module to obtain an eleventh IP data packet and send the same.

19. An encoding device, comprising:

The IP layer module is further configured to send the third IP data packet obtained by the DNCP layer module.

20. The encoding device according to claim 19, characterized in that The IP layer module is further configured to receive a fifth internet protocol IP data packet;

The DNCP layer module is further configured to: if the IP layer module receives a sixth IP data packet within a set duration after receiving the fifth IP data packet, and the sixth IP data packet and the The destination address of the fifth IP data packet is the same, and the IP layer payload of the sixth IP data packet and the fifth IP data packet both include a dual source network coding protocol DNCP data packet, and the sixth IP data packet The IP address of the DNCP packet in the IP layer payload contains two source identifiers and two IP packet identifiers, and two letters contained in the DNCP header of the DNCP packet in the IP layer payload of the fifth IP packet. If the source identifier and the two IP data packet identifiers are the same, the coding coefficient vector is randomly selected for the load of the DNCP data packet in the fifth IP data packet and the sixth IP data packet IP layer payload in the finite field; The coding domain vector randomly selected in the finite field for the load of the DNCP packet in the fifth IP data packet and the sixth IP data packet IP layer payload, respectively, for the fifth IP data packet and the first IP layer payload of six IP packets The payload of the DNCP data packet is randomly linearly encoded to obtain a second linearly encoded data packet; the second linearly encoded data packet is DNCP-encapsulated to obtain a third DNCP data packet, and the DNCP header of the third DNCP data packet includes the Two source identifiers, the two IP packet identifiers, and a random selection of loads of the DNCP packets in the fifth IP data packet and the sixth IP data packet IP layer payload in the finite field, respectively Coding coefficient vector correlation coding coefficient vector information; performing IP encapsulation on the third DNCP data packet to obtain a seventh IP data packet;

The IP layer module is further configured to send the seventh IP data packet obtained by the DNCP layer module.

A decoding device, comprising:

a second reading module, configured to: if the first reading module reads the DNCP data packet The DNCP header contains two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector. Then, if there is a DNCP header in the buffer, the two source identifiers and the two are included. Another DNCP packet identified by the IP packet, if present, the other DNCP packet is read from the buffer, wherein the DNCP header of the other DNCP packet further contains another set of coding coefficient vectors. Coding coefficient vector information; a decoding module, configured to perform random linear decoding on the load of the read DNCP data packet by using two sets of coding coefficient vector information.

The decoding device according to claim 21, further comprising: a second storage module, configured to: after the first storage module stores the DNCP data packet into the buffer area, index the DNCP data packet The address of the DNCP packet in the buffer is stored in the content addressable memory CAM, wherein the index comprises: at least one IP packet identifier contained in the DNCP header of the DNCP packet.

A decoding device according to claim 21 or 22, characterized in that

The third receiving module is specifically configured to receive an IP data packet from the first link and the second link, respectively;

The first storage module is specifically configured to: if the IP layer payload of the IP data packet received by the third receiving module from the first link includes a DNCP data packet, store the DNCP data packet into the first a buffer area; if the IP layer payload of the IP data packet received by the third receiving module from the second link includes a DNCP data packet, storing the DNCP data packet into the second buffer area; The reading module is specifically configured to: when the DNCP data packet stored in the first buffer area reaches a preset number and/or the DNCP data packet stored in the second buffer area reaches a preset number, or When the sum of the DNCP data packets stored in the first buffer area and the second buffer area reaches a preset number, the DNCP data packet is read from the first buffer area according to a predetermined reading order.

24. The decoding device according to claim 23, wherein

The second reading module is specifically configured to: if the first reading module is from the first buffer area The DNCP header of the read DNCP packet includes two source identifiers, two IP packet identifiers, and a set of coding coefficient vector information related to the coding coefficient vector, according to at least one of the two source identifiers and At least one of the two IP packet identifiers, or, according to at least one of the two IP packet identifiers, finding whether the DNCP header exists in the second buffer area, including the two source identifiers and two Another DNCP data packet identified by the IP data packet, if present, the other DNCP data packet is read from the second buffer area, wherein the DNCP header of the other DNCP data packet further includes another group Coding coefficient vector correlation coding coefficient vector information.

25. A decoding device, comprising:

An IP layer module, configured to receive an Internet Protocol IP data packet;

26. A communication system, comprising:

The apparatus according to any one of claims 14 to 25.