Deep space network link adaptive coding transmission method based on hop-by-hop redundancy confirmation

Abstract

The invention relates to the technical field of computer networks, and discloses a deep space network link adaptive coding transmission method based on hop-by-hop redundancy confirmation, which comprises a data transmission architecture. Wherein the link sensing module dynamically estimates the link packet loss rate between the nodes by periodically sending topology learning packets between the nodes; the redundancy calculation module calculates the optimal redundancy at the source node according to the estimated link packet loss rate, and calculates the required minimum number of coded data packets; the coding module outputs a coding data packet at the source node according to the optimal redundancy obtained through calculation; the decoding module decodes the received coded data packet at the destination node and outputs original data; the data cache temporarily stores the coded data packet at each node, generates a corresponding feedback packet according to a check result, and notifies the previous hop node to continue to transmit or clean the data cache; the transceiving queue comprises a receiving queue and a sending queue and is used for receiving and sending data packets and feedback packets at each node; the method has the advantages of ensuring hop-by-hop redundancy reliability, reducing network transmission resource waste, reducing decoding overhead and the like.

Description

Technical Field

[0001] The present invention relates to the field of computer network technology, and in particular to a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation. Background Art

[0002] With the rapid development of space exploration technology and the growing demand for deep space network communications, the IP addresses in the traditional Internet are designed for static network addresses and the transmission control protocol relies on a fixed path from terminal to terminal. They are not suitable for deep space network environments and have great limitations in terms of resource efficiency, reliability, and network performance. In this context, the research on innovative deep space network transmission mechanisms has become a hot topic of global concern.

[0003] The traditional network architecture is centered on IP addresses and uses a TCP / IP architecture designed according to the end-to-end principle, which is difficult to adapt to the deep space network environment. First, the transmission control protocol of the traditional Internet is based on a fixed path from terminal to terminal, which is not applicable to deep space networks with large delays and topological changes. Secondly, the channel conditions of deep space networks change dynamically, and key indicators such as packet loss rate are difficult to predict, which undermines the reliability of upper-layer application services and affects the network's service quality. Finally, deep space networks place higher requirements on the resource efficiency of data transmission, and traditional coding transmission protocols cannot take into account high channel utilization, high reliability, and low encoding and decoding overhead, and cannot achieve efficient and reliable data transmission.

[0004] In response to the above problems, existing solutions such as linear network coding have proposed a transmission method for reliability. By sending redundant coded data packets to increase the probability of successful data transmission, the original data can be decoded from a certain number of coded data packets, thus being able to tolerate a certain degree of data packet loss. This transmission method can solve the reliability problem to a certain extent, but since the link quality of the deep space network changes dynamically with time and space, a fixed number of redundant coded data packets are not competent for data transmission in such cases. If the redundancy is too small, it may not guarantee the successful delivery of data, and if it is too large, it will cause a waste of network resources. In addition, dynamic network topology and link interruption will also reduce transmission reliability. The existing end-to-end reliable transmission mechanism will trigger retransmission when the link is disconnected, resulting in a large retransmission delay and additional redundant transmission, thereby reducing transmission efficiency.

[0005] In response to the above problems, the present invention proposes a new deep space network transmission interaction mode based on the deep space network's requirements for resource efficiency, reliability and encoding and decoding overhead to achieve reliable data transmission and ensure high resource utilization. Summary of the invention

[0006] Since the existing TCP / IP architecture cannot solve the fundamental problems of reliability and efficiency of data transmission in deep space networks, and the existing data coding mechanism is not efficient enough and cannot adapt to the dynamic changes in link quality, the present invention proposes a new deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation. The deep space network node updates its coding redundancy rate in real time according to the dynamic link quality, reducing the waste of network transmission resources, and using the hop-by-hop feedback mechanism to reduce the retransmission delay caused by link interruption. In addition, by designing a unique coding mechanism based on the Cauchy matrix, the destination node does not need to repeatedly try to calculate the decoding matrix in real time when preparing for decoding, thereby reducing the decoding overhead and calculation delay. This significantly improves the efficiency of data transmission and ensures the reliability of the data transmission process.

[0007] In order to solve the above technical problems, the present invention adopts the following technical solutions:

[0008] A deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation, comprising:

[0009] The data transmission architecture has a link sensing module, a redundant computing module, an encoding module, a decoding module, a data cache and a transceiver queue, wherein each node of the deep space network uses the link sensing module, the data cache and the transceiver queue; the source node uses the redundant computing module and the encoding module; the intermediate node uses the redundant computing module; and the destination node uses the decoding module;

[0010] The link awareness module dynamically estimates the link packet loss rate between nodes by periodically sending topology learning packets between nodes;

[0011] The redundancy calculation module calculates the optimal redundancy according to the estimated link packet loss rate at the source node, and calculates the minimum number of required coded data packets according to the estimated link packet loss rate at the intermediate node;

[0012] The encoding module encodes the original data at the source node according to the calculated optimal redundancy using a coding matrix based on the Cauchy matrix and outputs a coded data packet;

[0013] The decoding module decodes the received coded data packet at the destination node and outputs the original data;

[0014] The data cache temporarily stores the encoded data packets at each node and performs batch checks when the transmission is not completed. According to the check results, the corresponding feedback packets are generated to notify the previous hop node to continue the transmission or clear the data cache;

[0015] The transceiver queue includes a receiving queue and a sending queue, which are used to perform receiving operations and sending operations of data packets and feedback packets at each node;

[0016] Among them, the single-hop transmission process of the encoded data packet is as follows: the current node sends the encoded data packet to the next-hop node, and the next-hop node temporarily stores the encoded data packet in the data cache after receiving it, checks whether the current batch of encoded data packets meets the redundancy requirements of subsequent transmission or decoding, and notifies the current node to continue or stop transmission through the feedback packet based on the inspection results.

[0017] As a further preferred technical solution of the present invention, the checking whether the current batch of encoded data packets meets the redundancy requirements of subsequent transmission or decoding, and notifying the current node to continue transmission or stop transmission through a feedback packet according to the checking result, specifically includes:

[0018] If the number of coded data packets received by the next-hop node has not yet met the redundancy requirement, the next-hop node will send a feedback packet containing information about the need to continue transmitting the coded data packets to the current node. After receiving the feedback packet, the current node will continue to transmit the current batch of coded data packets to the next-hop node; if the number of coded data packets received by the next-hop node has met the redundancy requirement for subsequent transmission or decoding, the next-hop node will send a feedback packet containing information that the number of coded data packets is sufficient to the current node. After receiving the feedback packet, the current node will stop subsequent transmission of the current batch of coded data packets and remove the current batch of coded data packets from the data cache of the current node.

[0019] As a further preferred technical solution of the present invention, the link perception module dynamically estimates the link packet loss rate between nodes by regularly sending topology learning packets between nodes, specifically including:

[0020] Link 1 i,j The packet loss rate is denoted as p i,j ; From node U i To the destination node U D The link packet loss rate is denoted as p i→D ; For node U i and node U i The next hop node U j , link packet loss rate p i→D satisfy:

[0021] p i→D =1-(1-p j→D )×(1-p i,j ); (1)

[0022] When a node U i Receive the destination node U d After the topology learning packet is generated, the link packet loss rate is calculated according to formula (1), and the updated topology learning packet is forwarded to the neighboring nodes; if multiple topology learning packets generated by the same node are received, the topology learning packet with the least forwarding hops is selected, thereby realizing the regular update of the link packet loss rate.

[0023] As a further preferred technical solution of the present invention, the redundancy calculation module calculates the optimal redundancy at the source node according to the estimated link packet loss rate, and calculates the minimum number of encoded data packets required at the intermediate node according to the estimated link packet loss rate, specifically including:

[0024] The number of original data packets in a batch is recorded as N S , the number of coded data packets generated by the source node is recorded as N C , the coding redundancy is t, then:

[0025] N C =(t+1)×N s ;

[0026] The required successful decoding probability threshold is denoted as Γ * ; N generated and sent by the source node C There are no less than N coded packets. S The probability that different coded packets arrive at the destination node, so that the original data can be decoded L for:

[0027]

[0028] Where N C ≥N S ; then the smallest N C It is obtained by solving the following minimization problem:

[0029]

[0030] Among them, redundancy t is the optimization variable; if N C The minimum value of Then the optimal redundancy t * for:

[0031]

[0032] The redundancy calculation module of the source node uses the calculation result of the optimal redundancy to generate a coded data packet in the subsequent coding module for data transmission;

[0033] The redundancy calculation module of the intermediate node calculates the link packet loss rate and batch size of the subsequent path to determine the minimum number of encoded data packets Y required from the previous hop node. i , so that the destination node can reach a value not less than the preset threshold Γ * The probability of receiving enough coded data packets; the intermediate node U i Forward Y i The probability that the coded data packets can enable the destination node to complete the decoding of the original data is Γi for:

[0034]

[0035] The minimum number of encoded packets required by the intermediate node Y i It is obtained by solving the following minimization problem:

[0036]

[0037] The redundant computing module of the intermediate node will i Compare with the number of different coded packets in the current data cache. If the number of coded packets currently received is less than Y i , the generated feedback packet is sent back to the previous hop node, requiring it to continue transmitting the coded data packet until the number of different coded data packets received exceeds Y i , and then generate a feedback packet to the previous hop node, notifying the previous hop node to stop sending subsequent encoded data packets and clean up the data cache.

[0038] As a further preferred technical solution of the present invention, the encoding module encodes the original data at the source node according to the calculated optimal redundancy using a coding matrix based on a Cauchy matrix and outputs a coded data packet, specifically comprising:

[0039] Assume that a batch of data contains n original data packets, and the coding redundancy is set to t*. Then the coding module will generate k = c + n coded data packets, where c = t* × n, and the coding matrix used is Φ k×n It is expressed as:

[0040]

[0041] Among them I n is the n-order unit matrix; E c×n The encoding matrix is ​​c×n, which is the following Cauchy matrix C n×n Submatrix of :

[0042]

[0043] Among them, x i ,y j ∈GF(2 q ),i,j=1,2,...,n, and x i ≠y j GF(2 q ) represents a finite field; using the encoding matrix Φ based on the Cauchy matrix k×n After encoding, the source node will obtain k encoded data packets for subsequent transmission.

[0044] Compared with the prior art, the beneficial technical effects of the present invention are:

[0045] The present invention has the advantages of hop-by-hop redundant reliability guarantee, reduced network transmission resource waste, and reduced decoding overhead. It breaks the limitations of the traditional end-to-end design of network coding transmission, and decomposes the end-to-end reliability into hop-by-hop redundant check through hop-by-hop redundant guarantee. Each intermediate node on the transmission path will cache the received coded data packets, thereby effectively improving reliability. The hop-by-hop redundant guarantee design also reduces a large number of full-path retransmissions under the end-to-end design, and the transmission feedback design enables the transmission of redundant coded data packets to be stopped in time, thereby reducing the waste of network transmission resources. In addition, by designing a unique encoding and decoding mechanism based on the Cauchy matrix, the destination node does not need to repeatedly try to calculate the decoding matrix in real time when preparing for decoding, thereby reducing the decoding overhead and calculation delay, and is suitable for network data transmission such as deep space networks that need to control energy delay overhead and have reliable transmission requirements. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A data transmission architecture diagram of the present invention;

[0047] Figure 2 This is a timing diagram of the single-hop data transmission and reception operation of the present invention;

[0048] Figure 3 It is a flow chart of the link sensing module of the present invention;

[0049] Figure 4 It is a flow chart of the redundant calculation module of the present invention;

[0050] Figure 5 It is a flow chart of the encoding module of the present invention;

[0051] Figure 6 It is a flow chart of the decoding module of the present invention;

[0052] Figure 7 It is a flow chart of the data caching and transceiver module of the present invention;

[0053] Figure 8 It is a schematic diagram of the hop-by-hop transmission and feedback process of the present invention. DETAILED DESCRIPTION

[0054] A preferred embodiment of the present invention is described in detail below with reference to the accompanying drawings.

[0055] The present invention proposes a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation. Figure 1As shown in the figure, the node mainly includes link sensing module, redundant computing module, encoding module, decoding module, data cache and transceiver queue. Each node uses link sensing module, data cache and transceiver queue. The source node in the transmission architecture uses redundant computing module and encoding module; the intermediate node in the transmission architecture uses redundant computing module; the destination node in the transmission architecture uses decoding module.

[0056] The link awareness module is responsible for dynamically acquiring the link packet loss rate information between network nodes by periodically sending topology learning packets (TLP) between nodes.

[0057] The redundancy calculation module is responsible for calculating the optimal redundancy for subsequent encoding at the source node of the transmission according to the estimated link packet loss rate, and calculating the minimum number of encoded data packets required for subsequent transmission at the intermediate node of the transmission according to the estimated link packet loss rate.

[0058] The encoding module is responsible for encoding the original data using the Cauchy matrix according to the optimal redundancy at the source node of the transmission and outputting the encoded data packet.

[0059] The decoding module is responsible for decoding the encoded data packet information received in the cache at the destination node of the transmission and outputting the original data.

[0060] The data cache is responsible for temporarily storing the encoded data packets at each node and performing batch checks when the transmission is not completed. Based on the check results, the corresponding feedback packets are generated to notify the previous hop node to continue transmission or clear the data cache.

[0061] The send and receive queue is responsible for performing basic data packet and feedback packet sending and receiving operations at each node.

[0062] The single-hop transmission process of the encoded data is as follows: Figure 2 As shown. The current node sends a coded data packet to the next-hop node. After receiving the coded data packet, the next-hop node temporarily stores it in the data cache and checks whether the batch of coded data packets is sufficient to meet the redundancy requirements of subsequent transmission or decoding. If the number of coded data packets received by the next-hop node does not meet the redundancy requirements, the next-hop node will send a feedback packet containing information about the need to continue transmitting coded data packets to the current node. After receiving the feedback packet, the current node will continue to transmit the batch of coded data packets to the next-hop node; if the number of coded data packets received by the next-hop node has met the redundancy requirements of subsequent transmission or decoding, the next-hop node will send a feedback packet containing information that the number of coded data packets is sufficient to the current node. After receiving the feedback packet, the current node will stop the subsequent transmission of the batch of coded data packets and remove the batch of coded data packets from the data cache of the current node.

[0063] The specific workflow of each major module is as follows.

[0064] 1. Link perception module

[0065] The link sensing module periodically sends topology learning packets (TLPs) between nodes to continuously update the path packet loss rate. i,j The packet loss rate is denoted as p i,j From node U i To the destination node U D The path packet loss rate is expressed as p i→D For node U i and its next hop node U j , the path packet loss rate satisfies:

[0066] p i→D =1-(1-p j→D )×(1-p i,j ); (1)

[0067] When a node U i Receive the destination node U d After the TLP is generated, the path packet loss rate is calculated according to formula (1), and the updated TLP is forwarded to the neighboring node. If multiple TLPs generated by the same node are received, the TLP with the least forwarding hops is selected to achieve regular update of the path packet loss rate. The process is as follows Figure 3 shown.

[0068] 2. Redundant computing modules

[0069] The redundancy calculation module calculates the required redundancy and number of coded packets based on the link quality and batch size. The goal of the redundancy calculation is to find the minimum number of coded packets generated or forwarded so that the destination node can receive enough coded packets with a probability not less than a preset threshold.

[0070] Consider a batch of raw data packets, the number of which is N S , the number of coded data packets generated by the source node is recorded as N C , the coding redundancy is t, then:

[0071] N C =(t+1)×N S ; (2)

[0072] The required successful decoding probability threshold is denoted as Γ * .

[0073] For the source node, it generates and sends N C There are no less than N coded packets. L The probability that different coded packets arrive at the destination node, so that the original data can be decoded L for:

[0074]

[0075] Where N C ≥N S . Then the smallest N C It can be obtained by solving the following minimization problem:

[0076]

[0077] The redundancy t is the optimization variable.

[0078] If the minimum value is Then the optimal redundancy can be expressed as:

[0079]

[0080] The redundant calculation module of the source node uses the result in the subsequent encoding module to generate an encoded data packet for data transmission.

[0081] For the intermediate nodes, the redundancy calculation module performs similar calculations based on the packet loss rate of the subsequent path and the batch size to determine the minimum number of encoded data packets Y required from the previous hop node. i , so that the destination node can reach a value not less than the preset threshold Γ * The probability of receiving enough coded data packets is i , which forwards Y i The probability that the coded data packets can enable the destination node to complete the decoding of the original data is Γ i for:

[0082]

[0083] The minimum number of encoded packets required by the intermediate node Y i It can be obtained by solving the following minimization problem:

[0084]

[0085] The redundant computing module of the intermediate node will i Compare with the number of different coded packets in the current data cache. If the number of coded packets currently received is less than Y i , then the generated feedback packet will be sent back to the previous hop node to request the continued transmission of the coded data packet until the number of different coded data packets received exceeds Y i , and then generate a feedback packet to the previous hop node to notify it to stop subsequent transmission and clean up the data cache. The process of the redundant calculation module is as follows Figure 4 shown.

[0086] 3. Encoding module

[0087] The encoding module encodes a batch of original data packets at the source node and outputs the encoded data packets. In order to avoid the complex matrix operation overhead, the encoding module uses a coding matrix based on the Cauchy matrix. Assuming that a batch of data contains n original data packets and the coding redundancy is set to t*, k = c + n encoded data packets will be generated, where c = t* × n. The encoding matrix used is Φ k×n It can be expressed as:

[0088]

[0089] Among them I n is the n-order identity matrix. c×n is a c×n encoding matrix, which is a submatrix of the following Cauchy matrix:

[0090]

[0091] Among them, x i ,y j ∈GF(2 q ),i,j=1,2,...,n, and x i ≠y j A simple way to get the value is to let x i =2i-1,y i =2i, which depends on the row number i, and the value range depends on the finite field GF(2 q ). Using the encoding matrix Φ based on the Cauchy matrix k×n After encoding, the source node will obtain k encoded data packets for subsequent transmission. The process of the encoding module is as follows: Figure 5 shown.

[0092] 4. Decoding module

[0093] It can be proved that from Φ k×n The n-order matrix Φ is established by randomly selecting n rows from n is reversible. Therefore, as long as the destination node receives n different coded data packets in this batch, it can complete the decoding operation and finally output the complete original data of this batch. Thanks to the unique coding matrix design based on the Cauchy matrix, once the value function of the matrix elements is determined, the node can calculate all possible decoding matrices offline and store them in the memory space. If the node receives enough coded data packets, it can directly select the corresponding decoding matrix from the memory to quickly restore the original data. The process of the decoding module is as follows Figure 6 shown.

[0094] 5.Data storage and transceiver module

[0095] The data storage and transceiver module includes the node's data cache and transceiver queue, which is responsible for temporarily storing the encoded data packets that have not yet been completed and performing basic underlying data transceiver operations. Figure 7 shown.

[0096] Example

[0097] This embodiment uses a feedback-based retransmission process as an example. This embodiment optimizes the data coding transmission protocol in the deep space network, greatly improving the reliability of data transmission and the utilization rate of the channel. The following introduces the feedback-based hop-by-hop transmission mechanism used in this embodiment.

[0098] like Figure 8 As shown in the middle right figure (2): ① Node A forwards a batch of coded data packets to node B, A is the sender and B is the receiver. ② The redundancy check module of node B checks the number of different coded data packets received in the data cache, and finds that the number of coded data packets currently received does not meet the redundancy guarantee requirements, and generates a feedback packet containing information that the batch of coded data packets needs to be resent and sends it to node A. ③ After receiving the feedback packet, node A continues to send the batch of coded data packets from the data cache. ④ The redundancy check module of node B checks the number of different coded data packets received in the data cache, and finds that the number of coded data packets currently received meets the redundancy guarantee requirements, and generates a feedback packet containing information to stop sending the batch of coded data packets and sends it to node A. Node A receives the information to stop forwarding the batch of coded data packets and clears the corresponding cached data from the data cache.

[0099] It is obvious to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential features of the present invention. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting, and the scope of the present invention is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention, and any reference numerals in the claims should not be regarded as limiting the claims involved.

[0100] In addition, it should be understood that although the present specification is described in terms of implementation modes, not every implementation mode contains only one independent technical solution. This narrative method of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation, characterized in that: include: The data transmission architecture has a link sensing module, a redundant computing module, an encoding module, a decoding module, a data cache and a transceiver queue, wherein each node of the deep space network uses the link sensing module, the data cache and the transceiver queue; the source node uses the redundant computing module and the encoding module; the intermediate node uses the redundant computing module; and the destination node uses the decoding module; The link awareness module dynamically estimates the packet loss rate of links between nodes by periodically sending topology learning packets between nodes; The redundancy calculation module calculates the optimal redundancy according to the estimated link packet loss rate at the source node, and calculates the minimum number of required coded data packets according to the estimated link packet loss rate at the intermediate node; The encoding module encodes the original data at the source node according to the calculated optimal redundancy using a coding matrix based on the Cauchy matrix and outputs a coded data packet; The decoding module decodes the received coded data packet at the destination node and outputs the original data; The data cache temporarily stores the encoded data packets at each node and performs batch checks when the transmission is not completed. According to the check results, the corresponding feedback packets are generated to notify the previous hop node to continue the transmission or clear the data cache; The transceiver queue includes a receiving queue and a sending queue, which are used to perform receiving operations and sending operations of data packets and feedback packets at each node; Among them, the single-hop transmission process of the encoded data packet is as follows: the current node sends the encoded data packet to the next-hop node, and the next-hop node temporarily stores the encoded data packet in the data cache after receiving it, checks whether the current batch of encoded data packets meets the redundancy requirements of subsequent transmission or decoding, and notifies the current node to continue or stop transmission through the feedback packet based on the inspection results.

2. According to claim 1, a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation is characterized in that: The checking whether the current batch of encoded data packets meets the redundancy requirements of subsequent transmission or decoding, and notifying the current node to continue transmission or stop transmission through a feedback packet according to the checking result, specifically includes: If the number of coded data packets received by the next-hop node has not yet met the redundancy requirement, the next-hop node will send a feedback packet containing information about the need to continue transmitting the coded data packets to the current node. After receiving the feedback packet, the current node will continue to transmit the current batch of coded data packets to the next-hop node; if the number of coded data packets received by the next-hop node has met the redundancy requirement for subsequent transmission or decoding, the next-hop node will send a feedback packet containing information that the number of coded data packets is sufficient to the current node. After receiving the feedback packet, the current node will stop subsequent transmission of the current batch of coded data packets and remove the current batch of coded data packets from the data cache of the current node.

3. According to claim 1, a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation is characterized in that: The link perception module dynamically estimates the link packet loss rate between nodes by regularly sending topology learning packets between nodes, specifically including: Link 1 i,j The packet loss rate is denoted as p i,j ; From node U i To the destination node U D The link packet loss rate is denoted as p i→D ; For node U i and node U i The next hop node U j , link packet loss rate p i→D satisfy: p i→D =1-(1-p j→D )×(1-p i,j ); (1) When a node U i Receive the destination node U d After the topology learning packet is generated, the link packet loss rate is calculated according to formula (1), and the updated topology learning packet is forwarded to the neighboring nodes; if multiple topology learning packets generated by the same node are received, the topology learning packet with the least forwarding hops is selected, thereby realizing the regular update of the link packet loss rate.

4. According to claim 1, a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation is characterized in that: The redundancy calculation module calculates the optimal redundancy at the source node according to the estimated link packet loss rate, and calculates the minimum number of required encoded data packets at the intermediate node according to the estimated link packet loss rate, specifically including: The number of original data packets in a batch is recorded as N S , the number of coded data packets generated by the source node is recorded as N C , the coding redundancy is t, then: N C =(t+1)×N S ; The required successful decoding probability threshold is denoted as Γ * ; N generated and sent by the source node C There are no less than N coded packets. S The probability that different coded packets arrive at the destination node, so that the original data can be decoded, is L for: Where N C ≥N S ; then the smallest N C It is obtained by solving the following minimization problem: Among them, redundancy t is the optimization variable; if N C The minimum value of Then the optimal redundancy t * for: The redundancy calculation module of the source node uses the calculation result of the optimal redundancy to generate a coded data packet in the subsequent coding module for data transmission; The redundancy calculation module of the intermediate node calculates the link packet loss rate and batch size of the subsequent path to determine the minimum number of encoded data packets Y required from the previous hop node. i , so that the destination node can reach a value not less than the preset threshold Γ * The probability of receiving enough coded data packets; the intermediate node U i Forward Y i The probability that the coded data packets can enable the destination node to complete the decoding of the original data is Γ i for: The minimum number of encoded packets required by the intermediate node Y i It is obtained by solving the following minimization problem: The redundant computing module of the intermediate node will i Compare with the number of different coded packets in the current data cache. If the number of coded packets currently received is less than Y i , the generated feedback packet is sent back to the previous hop node, requiring it to continue transmitting the coded data packet until the number of different coded data packets received exceeds Y i , and then generate a feedback packet to the previous hop node, notifying the previous hop node to stop sending subsequent encoded data packets and clean up the data cache.

5. According to claim 4, a deep space network link adaptive coding transmission method based on hop-by-hop redundant confirmation is characterized in that: The encoding module encodes the original data at the source node according to the calculated optimal redundancy using a coding matrix based on a Cauchy matrix and outputs a coded data packet, specifically including: Assume that a batch of data contains n original data packets, and the coding redundancy is set to t*. Then the coding module will generate k = c + n coded data packets, where c = t* × n, and the coding matrix used is Φ k×n It is expressed as: Among them I n is the n-order unit matrix; E c×n The encoding matrix is ​​c×n, which is the following Cauchy matrix C n×n Submatrix of : Among them, x i ,y j ∈GF(2 q ),i,j=1,2,...,n, and x i ≠y j GF(2 q ) represents a finite field; using the encoding matrix Φ based on the Cauchy matrix k×n After encoding, the source node will obtain k encoded data packets for subsequent transmission.

Images