CN108966259B

CN108966259B - Anti-interference transmission method based on network coding

Info

Publication number: CN108966259B
Application number: CN201810789578.8A
Authority: CN
Inventors: 许道峰; 田少鹏; 樊县林
Original assignee: CETC 28 Research Institute
Current assignee: CETC 28 Research Institute
Priority date: 2018-07-18
Filing date: 2018-07-18
Publication date: 2021-07-16
Anticipated expiration: 2038-07-18
Also published as: CN108966259A

Abstract

The invention discloses an anti-interference transmission method based on network coding, which comprises the following steps: (1) acquiring original data information from an Ethernet frame to be transmitted; (2) splitting original data information into a plurality of slices, and carrying out network coding on the plurality of slices to form a plurality of network coding slices; (3) adding a network layer protocol header, a transport layer protocol header, a network coding identification header, a total number of slices, a current network coding slice serial number and a network coding serial number to the head of a network coding slice, and adding frame check data to the tail of the network coding slice to form a network coding data frame; (4) selecting an optimal link according to a minimum broadband criterion or a minimum delay criterion; (5) and transmitting the network coding data frame by adopting the preferred link. The invention can greatly improve the correct information receiving probability under the condition of high link frame error rate and ensure the reliable information receiving.

Description

Anti-interference transmission method based on network coding

Technical Field

The invention relates to the technical field of communication, in particular to an anti-interference transmission method based on network coding.

Background

The tactical communication environment is mainly wireless transmission, and the DIL (dynamic Intermitent Link) characteristics of high dynamic, weak connection, openness and the like of the channel environment cause unstable and unreliable information transmission and are easy to be interfered and intercepted by enemies. The anti-interference transmission technology of the tactical communication environment comprises the processing of a plurality of layers such as a signal layer, a network layer, an information layer and the like, and achieves the purpose of reliably transmitting the combat information in the tactical communication environment through systematic and systematic comprehensive application. The processing means of the signal layer comprises the processing technologies of space, time, frequency and code domains such as frequency hopping, time hopping, directional beam, code division multiple access and the like; the processing means of the network layer comprises processing technologies such as message retransmission, breakpoint continuous transmission, rerouting, multipath and the like; the processing means of the information layer mainly processes the messages, and increases the success rate of information delivery by adding certain redundancy, such as the network coding technology mentioned herein. At present, anti-interference transmission in a tactical communication environment is mainly concentrated on a signal layer and a network layer, and the processing on an information layer is rarely involved.

Disclosure of Invention

The purpose of the invention is as follows: the invention provides an anti-interference transmission method based on network coding, which is suitable for a tactical communication environment and aims to solve the problems in the prior art.

The technical scheme is as follows: the anti-interference transmission method based on the network coding comprises the following steps:

(1) acquiring original data information from an Ethernet frame to be transmitted;

(2) splitting original data information into a plurality of slices, and carrying out network coding on the plurality of slices to form a plurality of network coding slices;

(3) adding a network layer protocol header, a transport layer protocol header, a network coding identification header, a total number of slices, a current network coding slice serial number and a network coding serial number to the head of a network coding slice, and adding frame check data to the tail of the network coding slice to form a network coding data frame;

(4) selecting an optimal link according to a minimum broadband criterion or a minimum delay criterion;

(5) and transmitting the network coding data frame by adopting the preferred link.

Further, the method further comprises:

(6) the network coding data is transmitted by adopting a redundant transmission mode and a multipath transmission mode while the optimal link transmission mode is adopted.

Further, after the network coded data frame is formed in the step (3), the network coded data frame is processed by HARQ and then transmitted.

Further, the step (2) specifically comprises:

(2-1) splitting the original data information into K slices, labeled d₁,d₂,…,d_K；

(2-2) forK slices are coded by adopting a network coding matrix C to obtain M network coding slices y₁,y₂,…,y_MWherein, the network coding formula is as follows:

Y＝DC

wherein Y is (Y)₁ y₂ … y_M)，D＝(d₁ d₂ … d_K) The network coding matrix C meets the following requirements: element C_ijAll are integers, and C_ij∈GF(2⁸) { 01 … 255}, and C is at GF (2)⁸) The above satisfies that any K columns form matrix uncorrelation.

Further, the network coding matrix C is specifically a matrix V obtained by elementary transformation of a matrix V with K rows and M columns or a matrix V obtained by elementary transformation, where the matrix V is specifically:

in the formula, q₁,q₂,...,q_MAre real numbers different from each other.

Further, the frame format of the network coded data frame in step (3) is specifically:

the narrowband network encoded data frame format is as follows:

CH

NCH

FN

Fseq

Nseq

Coded FDU

FCS

the format of the encoded data frame of the broadband network is as follows:

IPH

UDPH

NCH

FN

Fseq

Nseq

Coded FDU

FCS

wherein CH represents a frame header in a narrowband communication environment; NCH represents a network coding identification head with the length of 1 byte; FN denotes the total number of slices, length 1 byte; FSeq represents the serial number of the current network coding slice and the length is 1 byte; the NSeq represents a network coding sequence number, namely a message sequence number of original data information, and the length of the network coding sequence number is 1 byte; the Coded FDU represents specific data of the network coding slice, and the length of the Coded FDU is S bytes; IPH denotes an IP header in a broadband communication environment; UDPH denotes a UDP header in a broadband communication environment; FCS denotes a frame check, and is 2bytes in length.

Further, the specific step of selecting the preferred link by using the minimum bandwidth criterion in the step (4) is as follows:

and sequencing all exit links from large to small according to the utility metric E, and distributing the bandwidth requirement of the transmission data from the 1 st link until the m links are distributed, wherein the sum of the utility metrics of the first m links is greater than the sum of the bandwidths of the network coding data frames.

Further, the specific step of selecting the preferred link by using the minimum delay criterion in the step (4) is as follows:

and sequencing all the egress links from large to small according to the utility metric E, and proportionally distributing the bandwidth requirement of the transmission data to all the egress links.

Wherein, the utility metric E of the egress link is calculated according to the following formula:

wherein p is 1-gammaⁿγ represents a frame error rate, n represents the number of retransmissions, and R represents a link bandwidth.

Has the advantages that: compared with the prior art, the invention has the following remarkable advantages: (1) the application layer anti-interference processing technology facing the combat information is complementary with the traditional anti-interference technology; (2) the data frame is taken as a processing unit and is suitable for different transmission links and different network types; (3) the system has the comprehensive utilization capacity of multiple links, and is particularly suitable for application scenes with multiple exit links, such as fixed command posts, maneuvering command posts and the like; (4) the coding efficiency and the coding efficiency are flexible and variable, and network coding schemes with different efficiencies in the range of [ 1/2-1) can be conveniently constructed.

Drawings

Fig. 1 is a schematic flow chart of an anti-interference transmission method based on network coding according to the present invention;

FIG. 2 is a diagram illustrating a format of a network encoded data frame;

FIG. 3 is a graph comparing network coding system performance;

FIG. 4 is a diagram illustrating the probability of correct reception when the network coding efficiency is 50%;

fig. 5 is a schematic diagram of network coding average delay.

Detailed Description

The embodiment provides an anti-interference transmission method based on network coding, as shown in fig. 1, including the following steps:

(1) and acquiring original data information from the Ethernet frame to be transmitted.

(2) The original data information is divided into a plurality of slices, and the plurality of slices are subjected to network coding to form a plurality of network coding slices.

When network coding is carried out, the length of original data to be transmitted is set to be L₀If the network coding supports data slices of length S, an original data may be split into K ═ L₀(S) data slices of d₁,d₂,…,d_K. Suppose that M data slices of length S are formed by redundant coding, y₁,y₂,…,y_MThen the coding efficiency is K/M. The coding scheme is as follows: y is_i＝c_i1d₁+c_i2d₂+…+c_iKd_K＝(d₁ d₂ … d_K)(c_i1 c_i2 … c_iK)^TI-1, 2, …, K, i.e., each encoded slice of the data frame is linearly combined from all the original slices. Ideal network coding is expected to achieve the following performance:

a coding efficiency is as high as possible;

the receiving end B correctly receives any k coded data slices and can correctly decode the original data frame;

and C, the transmission delay of the original data is minimum.

Let Y be (Y)₁ y₂ … y_M) Then, then

Wherein D ═ D₁ d₂ … d_K) Is S × K dimensional matrix, C is K × M dimensional (K < M) network coding matrix, and each column corresponds to one timeAnd linearly combining the weight vectors of the network coding. Obviously, if the original data frame is to be completely restored, any K columns of matrix C are required to form a matrix

Is not relevant. At this time

Wherein

And forming a matrix for K columns corresponding to Y. Without any constraints, C may be taken from M uniformly distributed vectors spanned into a K-dimensional space, similar to codebook space vectors. The operation of the data frame is limited to GF (2)⁸) Thus, the elements of C must all be integers satisfying C_ij∈GF(2⁸) { 01 … 255}, and C must be at GF (2)⁸) The above satisfies that any K columns form matrix uncorrelation.

The Vandermonde matrix meets the above requirements, given a vector q ═ q (q)₁ q₂ … q_M) The construction method comprises the following steps:

when the vector q is taken (12 … 16), the 8 × 16 dimensional network coding matrix is shown in table 1:

TABLE 1 GF (2)⁸) Domain encoding matrix

1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
																1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1	4	5	16	17	20	21	64	65	68	69	80	81	84	85	29
																1	8	15	64	85	120	107	58	115	146	221	231	186	127	36	205
1	16	17	29	28	13	12	205	204	221	220	208	209	192	193	76
																1	32	51	116	108	46	36	38	226	1	215	169	116	244	59	180
1	64	85	205	193	228	252	45	161	10	146	191	62	241	100	143
																1	128	255	19	226	98	206	117	192	68	79	87	43	199	38	24

From this, a generic network coding scheme can be obtained, which is formed by Vandermonde matrices of different Galois fields, as shown in table 2.

TABLE 2 different Galois field network coding schemes

The operation units corresponding to the black adding part in the table are just half bytes, 1 byte and 2bytes, and the operation units are relatively good for network coding and network decoding; the M-value and K-value (corresponding to the dimension of the Vandermonde matrix and a specific column) corresponding to the simultaneously darkened portion can be just expressed in integer byte number, and the content extraction of the sliced data frame is relatively good.

Performing corresponding elementary transformation on the Vandermonde matrix, wherein the obtained matrix can also be used as a network coding matrix; in addition, the network coding matrix may also be generated by other construction methods.

(3) Adding a network layer protocol header, a transport layer protocol header, a network coding identification header, the total number of the slices, the serial number of the current network coding slice and the serial number of the network coding to the head of the network coding slice, and adding frame check data to the tail of the network coding slice to form a network coding data frame. The specific format is shown in fig. 2.

NCH: network Coding Header, 1 byte;

FN：Fragment Number，GF(2⁸) 1 byte in time, slice number K, maximum 256;

FSeq：Fragment Sequence，GF(2⁸) 1 byte, slice serial number (value 1-M), representing a specific network coding slice, corresponding to a column of Vandermonde matrix, maximum 256;

NSeq: network Coding Sequence, 1 byte, Network Coding Sequence number, namely original data message Sequence number, maximum 256, and recycling. Referring to the original message lengths listed in the table, the message length supported by 1-byte NSeq reaches 256 × 25.6Kbytes to 6.55Mbytes and 256 × 64Kbytes to 16.4Mbytes, and the corresponding transmission delay reaches 5461Sec and 13653Sec (link bandwidth 19.2Kbps), so that 1 byte can meet the transmission requirement in a cycle use;

coded FDUs: the length of the network coded slice data is S Bytes;

CH: channel Header, frame Header under narrow-band tactical communication;

UDPH: UDP Header, UDP Header in a broadband communication environment;

IPH: IP Header, IP Header in broadband communication environment;

FCS: and checking the frame, wherein the length of the frame is 2 Bytes.

In addition, hybrid ARQ (hybrid automatic repeat request), namely HARQ, can be selected for processing after the network coded data frame is obtained.

(4) The preferred link is selected according to a minimum wideband criterion or a minimum delay criterion.

Lsa (link Select algorithm) is an algorithm for choosing one or more specific links for transmission when multiple egress links are available, and is generally closely related to the transmission optimization protocol. For an egress link, the most common attributes include the link bandwidth R and the frame error rate γ, where the former is associated with the transmission delay; the latter is related to the anti-interference capability, i.e. if the anti-interference capability is strong, the frame error rate is low; if the anti-interference capability is weak, the frame error rate is high. Thus, for a fixed/mobile command post, the egress link it possesses can be described by the following mathematical set:

{(R_i,γ_i)},i＝1,2,...,L

the goal of LSA is to reasonably distribute network-coded slice data packets to corresponding egress links under the constraint of each egress link index, so as to maximize the success rate of information transmission, minimize the time delay of information transmission, minimize the possibility of interception and tampering of information transmission, and so on.

For multivariable constrained optimization problems, it is relatively difficult to solve directly, and it is common practice to convert them into univariate optimization problems. The utility metric (Efficiency measure) of the link is proposed, namely the actual available bandwidth of the link under a specific frame error rate in consideration of factors such as retransmission and frame error.

In the frame error rate γ, assuming that p (e.g. 99%) is reached, i.e. it is considered that information can be reliably transmitted, under the retransmission policy, the retransmission number n conforms to the following formula:

p＝1-γⁿ

at this time, the utility metric E of the link is defined as:

based on the utility metric E, a link selection may be made for the network encoded data frame. In the anti-interference transmission mode, the accuracy of information transmission is the primary optimization target, and the time delay is an additional condition, so the following several formulas can be adopted.

A、LSA-B

Minimum bandwidth criteria LSA. And sequencing all exit links from large to small according to the utility metric E, and distributing the bandwidth requirement of the transmission data from the 1 st link until the m links are distributed, wherein the sum of the utility metrics of the first m links is greater than the sum of the bandwidths of the network coding data frames. That is, if the maximum 1 satisfies the requirement, only the transmission is performed through the link; and if the utility metric is not enough, calculating the difference value between the bandwidth of the network coding data frame and the utility metric, if the 2 nd link meets the requirement, transmitting the slice data of the difference value part on the 2 nd link, and the like.

B、LSA-T

The minimum delay criterion LSA. All the exit links are sorted from large to small according to the utility metric E, and the bandwidth requirement of the transmission data is proportionally distributed to all the exit links, which is similar to a water filling algorithm in the communication theory.

The following performance analysis was performed for the above method.

(1) Probability of correct reception

Assuming that the frame error rate of the transmission system is γ, the interference rejection (correct reception probability) of the uncoded system, the network coded system and the simple retransmission system are:

pr₀＝(1-γ)^K

pr_d＝1-(1-pr₀)²＝pr₀(2-pr₀)

fig. 3 shows a comparison of the probability of correct reception for the above three transmission schemes at the same frame error rate. Therefore, the network coding scheme provided by the invention has larger performance gain. Fig. 4 shows the performance difference caused by different network coding parameters under the same coding efficiency. When (K, M) is (256,512), the frame error rate is still about 99% correct transmission at 45%, and when (K, M) is (128,256), the transmission correct rate is about 96% at the same frame error rate. From the simulation curves the following conclusions can be drawn:

A. using GF (2)⁸) The coding scheme over the domain outperforms the uncoded and repeated transmission schemes when the frame error rate is less than 0.5;

B. under the same coding efficiency, the performance gain of network coding is improved along with the increase of K and M, and the gain is gradually reduced;

B. when the frame error rate is greater than 0.5, the performance of repeated transmission is better;

D. it is assumed that when K, M → ∞ is reached, the correct reception probability curve approaches a step function, and the transition point is γ ═ 0.5, that is, when K (M) is increased, the correct reception probability tends to 1, in the case where the same coding efficiency and the frame error rate are less than 50%.

(2) Time delay analysis

Table 3 shows the change in data length before and after network coding at 50% network coding efficiency, typical slice length. Because of the presence of GF (2)⁸) The total number of 256 possible code vectors is one byte, so that at 50% efficiency, 128 data slices can be supported at most, and the supported original data length is divided into 25.6Kbytes and 64Kbytes, which is much larger than the MTU length of the Ethernet.

TABLE 3 message Length before and after network coding

Assuming that the transmission rate of the link is bKbps in a tactical communication environment, the average delay caused by network coding in two typical slice lengths in the above table can be estimated according to the following formula:

τ_i＝i×S×8/b

the average value of the network coding time delay can be obtained by arranging the three formulas

Fig. 5 shows the average delay when the link transmission rate is 19.2Kbps, the data slice K is 128, the network coding number M is 256, and the short frame length S is 200bytes and 500bytes, respectively. As can be seen, the average delay increases with the length of the short frame.

While the invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not to be limited to the disclosed embodiment, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. An anti-interference transmission method based on network coding is characterized by comprising the following steps:

2. The network coding based interference rejection transmission method of claim 1, wherein: further comprising:

3. The network coding-based interference rejection transmission method according to claim 1 or 2, wherein: and (3) after the network coding data frame is formed, processing the network coding data frame through HARQ and then transmitting the network coding data frame.

4. The network coding based interference rejection transmission method of claim 1, wherein: the step (2) specifically comprises the following steps:

(2-2) for K slices, coding by adopting a network coding matrix C to obtain M network coding slices y₁,y₂,…,y_MWherein, the network coding formula is as follows:

Y＝DC

wherein Y is (Y)₁ y₂ … y_M)，D＝(d₁ d₂ … d_K) Network ofThe coding matrix C meets the following requirements: element C_ijAll are integers, and C_ij∈GF(2⁸) { 01 … 255}, and C is at GF (2)⁸) The above satisfies that any K columns form matrix uncorrelation.

5. The network coding based interference rejection transmission method of claim 4, wherein: the network coding matrix C is specifically a matrix V of K rows and M columns or a matrix obtained by elementary transformation of V, where the matrix V is specifically:

in the formula, q₁,q₂,...,q_MAre integers different from each other.

6. The network coding based interference rejection transmission method of claim 1, wherein: the frame format of the network coding data frame in the step (3) is specifically as follows:

the narrowband network encoded data frame format is as follows:

CH NCH FN Fseq Nseq Coded FDU FCS

the format of the encoded data frame of the broadband network is as follows:

IPH UDPH NCH FN Fseq Nseq Coded FDU FCS

7. The network coding based interference rejection transmission method of claim 1, wherein: the specific steps of selecting the preferred link by adopting the minimum bandwidth criterion in the step (4) are as follows:

8. The network coding based interference rejection transmission method of claim 1, wherein: the specific steps of selecting the preferred link by adopting the minimum delay criterion in the step (4) are as follows:

9. The network coding-based interference rejection transmission method according to claim 7 or 8, wherein: the utility metric E of the egress link is calculated according to the following formula: