A method for designing a cooperative underwater acoustic communication network reliable transmission protocol based on random linear network coding

Abstract

This invention relates to a design method for a reliable transmission protocol in cooperative underwater acoustic communication networks based on random linear network coding. It is applicable to multi-hop reliable transmission scenarios of data originating from the same source and destined for two nodes. The method includes the following steps: 1. Network initialization; 2. Encoding vector dictionary configuration; 3. Encoding data generation; 4. Candidate transmission structure generation; 5. Transmission budget and structure benefit evaluation; 6. Encoding forwarding and data recovery. This protocol generates encoded data packets based on random linear network coding and identifies the corresponding encoding vectors using encoding indices. During hop-by-hop transmission, based on local topology, link quality status, and reliability objectives, three types of candidate transmission structures—shared forwarding, cooperative forwarding, and branch forwarding—are generated. The transmission mode with the optimal structure benefit is selected to complete the encoding forwarding. This invention can improve the reliability of multi-hop underwater acoustic transmission for two-terminal purposes, control redundant transmission overhead, and is suitable for underwater acoustic communication network environments with high packet error rates, long latency, and topology changes.

Description

Technical Field

[0001] This invention relates to the fields of underwater acoustic communication networks and reliable underwater acoustic transmission protocols, and particularly to a design method for a cooperative reliable transmission protocol for underwater acoustic communication networks based on random linear network coding. It is applicable to multi-hop reliable transmission scenarios of data from the same source to dual-destination nodes in underwater acoustic communication networks. Background Technology

[0002] Underwater acoustic communication networks are crucial for applications such as marine environmental monitoring, underwater platform coordination, distributed detection, target identification, and mission information distribution. Compared to terrestrial wireless communication networks, underwater acoustic communication networks are characterized by longer propagation times, limited available bandwidth, high packet error rates, significant multipath effects, and time-varying link states. Therefore, achieving reliable data transmission under conditions of low data rates, high packet error rates, and long latency is a key challenge in the design of underwater acoustic communication network protocols.

[0003] Existing reliable transmission methods in underwater acoustic communication networks mostly revolve around data delivery from a single source to a single destination node. These methods typically improve transmission reliability through single-path routing, multi-path routing, automatic repeat requests, fixed redundancy transmission, or forward error correction. However, under conditions of long propagation delays in underwater acoustic links, automatic repeat requests rely on hop-by-hop acknowledgments and feedback, easily introducing significant delays and control overhead. While fixed redundancy transmission is simple to implement, its recovery effectiveness depends on the successful arrival of specific packets, making it highly sensitive to continuous packet loss. Forward error correction trades extra redundancy for error correction capability, but on already low-speed, low-bit-rate underwater acoustic links, it often further compresses the limited effective information carrying capacity.

[0004] Network coding offers a novel technical approach to improving transmission reliability and resource utilization efficiency in complex networks. Random linear network coding, by linearly combining data packets of the same generation, allows the receiver to recover the original data by receiving only a sufficient number of linearly independent coded packets, regardless of whether a specific original packet was successfully received. This mechanism improves the overall generation recovery probability under random packet loss conditions and is suitable for integration with multi-hop relay forwarding.

[0005] However, most existing network coding transmission mechanisms for underwater acoustic communication networks are based on fixed routing structures or predetermined multipath frameworks. In multi-hop transmission of data from the same source to two destination nodes, the following problems still exist: First, the path relationships between different destination directions lack an adaptive organization mechanism, making it difficult to flexibly adjust the transmission structure according to local topology changes; second, the coded information received by multiple relay nodes is often processed independently, lacking an effective collaborative utilization mechanism; third, there is a lack of joint decision-making among path advancement, coding recovery requirements, and redundant transmission overhead, easily leading to insufficient reliability improvement or excessive redundant transmission. Therefore, how to adaptively organize the forwarding structure according to the path relationships in multi-destination transmission and collaboratively utilize the coded information distributed across different relay nodes is a key issue in improving joint recovery capabilities and controlling transmission overhead.

[0006] To address this, this invention proposes a design method for a reliable transmission protocol in cooperative underwater acoustic communication networks based on random linear network coding. This protocol is designed for bi-purpose, multi-hop transmission scenarios. Based on the generation of local candidate sets, it calculates the transmission budget under different transmission modes and selects a path structure that facilitates the complementary use of coded information through structural benefit evaluation. This unifies path advancement, coding recovery, and redundancy control into a hop-by-hop decision-making process, thereby improving joint recovery capabilities and controlling transmission overhead. Summary of the Invention

[0007] The purpose of this invention is to overcome the problems of high packet error rate, long latency, and difficulty in adaptively adjusting path structure in multi-hop reliable transmission in underwater acoustic communication networks, thereby providing a design method for a cooperative reliable transmission protocol for underwater acoustic communication networks based on random linear network coding. This protocol combines random linear network coding, candidate transmission structure generation, transmission budget calculation, and structure benefit evaluation. During hop-by-hop forwarding, it adaptively selects shared forwarding, cooperative forwarding, or branch forwarding transmission structures to improve the joint recovery capability of data from the same source transmitted to dual-destination nodes and control redundant transmission overhead.

[0008] In view of this, the present invention provides a design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding. The method for designing a reliable transmission protocol for a cooperative underwater acoustic communication network based on random linear network coding includes: Step 1: Each node in the network obtains its own one-hop neighbor relationship, the remaining hop count for bi-destination nodes, the connectivity relationship between nodes, and the initial estimate of link quality. Step 2: Configure the same encoding vector dictionary for each node; Step 3: The source node generates an encoded data packet and sets it as the current sending unit. The current sending unit is the node or node pair that sends or forwards the encoded data packet. Step 4: The current sending unit constructs a forward candidate set based on the current one-hop neighbor relationship and the remaining hop count information, filters candidate sending units based on the forward candidate set, and forms them into three types of candidate transmission structures: shared forwarding, cooperative forwarding, or branch forwarding. Step 5: For each candidate transmission structure, the current sending unit calculates the corresponding transmission budget and selects the candidate transmission structure with the greatest benefit as the current forwarding decision. Step 6: Based on the forwarding decision in Step 5, the current sending unit generates and sends the encoded data packet. The receiving unit receives the encoded data packet, determines whether the current node is the destination node, and then recovers the encoded data packet according to the encoded vector dictionary. If it is the destination node, the data transmission is completed; otherwise, the current node becomes the current sending unit in the subsequent hop-by-hop transmission process and proceeds to Step 4. The receiving unit is the node or node pair that received the encoded data packet.

[0009] As an improvement to the above method, step 2 specifically involves: each node in the network pre-configuring the same encoding vector dictionary and establishing a mapping relationship between encoding vectors and encoding indices, so that the current sending node can identify the encoding vector corresponding to the random linear network encoded data packet through the encoding index, and the receiving node can recover the corresponding encoding vector according to the encoding index.

[0010] As an improvement to the above method, the encoded data packet in step 3 includes: a protocol header and an encoded payload. The protocol header includes: GenID, NextHopID, Transmission Mode, and CodeID corresponding to the encoded payload. The encoded payload is constructed by the current sending unit dividing the data to be transmitted into several generations and then performing random linear network coding on the data groups within the same generation.

[0011] As an improvement to the above method, the specific steps for constructing the forward candidate set in step 4 are as follows: If the current sending unit is a single node Then construct the target-oriented node respectively. and Forward candidate set: ; in, For nodes A set of neighbors that jumps over time. and These represent the current node's orientation towards the destination node. and Reference remaining hops For a single node The forward candidate set, For a single node The forward candidate set; If the current sending unit is a node pair Then construct the target-oriented node respectively. and Forward candidate set: ; in, For nodes The forward candidate set, For nodes The forward candidate set.

[0012] As an improvement to the above method, in step 4, the conditions for selecting candidate sending units and classifying them into three candidate transmission structures—shared forwarding, cooperative forwarding, or branch forwarding—are as follows: 1) The current sending unit is a single node. At the same time, the shared forwarding candidate node belongs to and Forward candidate set; Or the current sending unit is a node pair At the same time, the shared forwarding candidate node simultaneously satisfies the condition relative to the node. Target node The conditions for advancement, and the conditions relative to the node Target node Forward propulsion conditions; Then the candidate node constitutes a shared forwarding candidate transmission structure and is added to the shared forwarding candidate transmission structure set; 2) The current sending unit is a single node. Then first from and Select candidate node pairs The current jump is formed only by a single node. To candidate node pairs The broadcast is forwarded; at this time, if the candidate node... In the next hop, it can pair with its subsequent candidate nodes. satisfy: ; Or the current sending unit is a node pair At that time, respectively from and Select candidate nodes and When candidate nodes pair Simultaneously satisfy: ; Then the candidate node constitutes a cooperative forwarding candidate transmission structure and is added to the cooperative forwarding candidate transmission structure set; 3) The current sending unit is a single node. The branch forwarding candidate transmission structure consists of those belonging to... and candidate node pairs constitute; Or the current sending unit is a node pair The branch forwarding candidate transmission structure consists of those belonging to... and The candidate node pairs constitute; Then the candidate node constitutes a branch forwarding candidate transmission structure and is added to the branch forwarding candidate transmission structure set; The three sets of candidate transmission structures serve as inputs for subsequent transmission budget calculations, structure benefit assessments, and transmission mode selection.

[0013] As an improvement to the above method, in step 5, the candidate transmission structure with the greatest transmission structure benefit is... Satisfy the following formula: ; in, For the benefit of transmission structure, These represent shared forwarding, cooperative forwarding, and branch forwarding transmission structures, respectively.

[0014] As an improvement to the above method, in step 6, the receiving unit recovers the received encoded data packet, specifically as follows: The receiving node determines the data generation based on the generation number in the encoded data packet and recovers the corresponding encoded vector from the encoding index according to the encoded vector dictionary. The receiving node is a single-node receiving unit, and it performs a linear correlation judgment between the encoded vector and the already received encoded vectors of the current generation. If the encoded packet can increase the rank of the current generation's receiving matrix, it is accumulated as a linearly independent encoded packet. When the receiving node obtains no less than a preset decoding threshold of linearly independent encoded packets, that is, when the receiving matrix satisfies: ; in, The number of original data groups contained in the current generation. The receiver matrix is ​​composed of the encoding vectors corresponding to the currently received encoded packets at the receiving node. To perform the rank operation on a matrix; After performing Gaussian elimination, the corresponding encoded vector can be recovered. ; ; in, For the r-th candidate encoding vector group, GF is the received encoded vector, N is the number of generated or received encoded packets, GF is the finite field, and m is the exponent of the size of the finite field.

[0015] Compared with the prior art, the advantages of the present invention are: 1. This invention proposes a cooperative reliable transmission protocol for single-source dual-purpose multi-hop underwater acoustic communication networks, which combines random linear network coding with hop-by-hop transmission mode selection to improve the data recovery probability under high packet error underwater acoustic link conditions.

[0016] 2. Design three candidate transmission structures: shared forwarding, cooperative forwarding, and branch forwarding. This enables the protocol to adaptively organize the path advancement method based on local topology and remaining hop count, avoiding the problem of insufficient adaptability of fixed routing structures in bi-purpose transmission scenarios.

[0017] 3. Calculate the transmission budget based on link quality estimation, remaining hop count, and end-to-end reliability target, and select the transmission mode in combination with structural benefits, so that the protocol can control redundant transmission overhead while improving transmission reliability.

[0018] 4. By adopting the encoding vector dictionary and encoding index representation method, the encoded data packet does not need to explicitly carry the complete encoding vector, which reduces the protocol header overhead and is more suitable for low-speed, bandwidth-limited underwater acoustic communication environments.

[0019] 5. Supports dynamic switching between single node and node pair transmission units, enabling data transmission to flexibly switch between shared propulsion, cooperative propulsion, and branch propulsion based on local network conditions, thus improving the protocol's adaptability to complex underwater acoustic topologies. Attached Figure Description

[0020] Figure 1 This is a flowchart illustrating a reliable transmission protocol design method for a cooperative underwater acoustic communication network based on random linear network coding, as described in this embodiment. Figure 2 This is a schematic diagram of the encoded data packet structure in the embodiment; Figure 3 This is a schematic diagram of a binocular transmission scenario in the embodiment, where the left diagram shows a fixed topology and the right diagram shows a random topology; Figure 4 This is a schematic diagram of the shared forwarding mode path structure in the embodiment; Figure 5 This is a schematic diagram of the path structure in the cooperative forwarding mode of the embodiment; Figure 6 This is a schematic diagram of the branch forwarding mode path structure in the embodiment; Figure 7This is a schematic diagram comparing the delivery performance of different transmission mechanisms under different packet error rates in the embodiments; wherein Figure 7 (a) shows the change in the Combined Delivery Rate (JPDR) as a function of the error rate. Figure 7 (b) shows the change in single-destination delivery rate (PDR) with packet error rate; Figure 8 This is a schematic diagram comparing the delivery performance of different transmission mechanisms under different hop count conditions in the embodiments; wherein Figure 8 (a) shows the change in the combined delivery rate (JPDR) with the number of hops. Figure 8 (b) shows the change in single-destination delivery rate (PDR) with the number of hops; Figure 9 This is a performance comparison of different protocol combinations under different packet error rates in the embodiments; wherein Figure 9 (a) shows the change in the Combined Delivery Rate (JPDR) as a function of the error rate. Figure 9 (b) shows the change in average number of transmissions as a function of packet error rate; Figure 10 This is a performance comparison of different protocol combinations under different numbers of nodes in the embodiments; wherein Figure 10 (a) shows the change in Joint Delivery Rate (JPDR) as a function of the number of nodes. Figure 10 (b) shows the change in the average number of transmissions as the number of nodes increases. Detailed Implementation

[0021] The technical solutions provided by the present invention will be further illustrated below with reference to the embodiments.

[0022] Example 1 For reference Figure 1 As shown, this embodiment provides a design method for a reliable transmission protocol in a cooperative underwater acoustic communication network based on random linear network coding, including: 1. Network initialization and state acquisition: Each node in the network obtains its own one-hop neighbor relationship based on the communication radius, and obtains the remaining hop count information and the initial estimate of the link quality for its two destination nodes, providing basic network state information for subsequent candidate structure generation and transmission budget calculation; 2. Encoding Vector Dictionary Configuration: Each node in the network is pre-configured with the same encoding vector dictionary and a mapping relationship between encoding vectors and encoding indices is established, so that the sending node can identify the encoding vector corresponding to the random linear network encoded data packet through the encoding index, and the receiving node can recover the corresponding encoding vector according to the encoding index. 3. Encoded data generation: The sending node / node pair divides the data to be transmitted into several generations, and performs random linear network coding on the original data packets within the same generation to generate encoded data packets carrying the coding index; 4. Candidate transmission structure generation: During the hop-by-hop forwarding process, the current sending unit constructs a forward candidate set for the two destination nodes based on the one-hop neighbor relationship, the remaining hop count information, and the communication reachability relationship between nodes, and generates three types of candidate transmission structures based on the forward candidate set: shared forwarding, cooperative forwarding, and branch forwarding. 5. Transmission Budget and Structure Benefit Assessment: Based on link quality estimation, remaining hop count, and end-to-end reliability targets, the protocol calculates the transmission budget required for each candidate transmission structure to meet local reliability requirements; simultaneously, it calculates the structure benefit of each candidate transmission structure based on the subsequent transmission states formed within a two-hop range. 6. Encoding, Forwarding and Data Recovery: The current sending unit performs encoding and forwarding of encoded data packets according to the selected transmission mode. When the receiving unit receives the encoded data packet, it determines whether the current node is the destination node, and then recovers the encoded data packet according to the encoding index. If it is the destination node, the data transmission is completed; otherwise, the current node becomes the current sending unit in the subsequent hop-by-hop transmission process.

[0023] Specifically: 1. The network model used in this embodiment is as follows: Figure 3 As shown, the left figure represents a fixed topology, and the right figure represents a random topology, constructing two simulation scenarios: regular topology and random topology. Regular topology is used to analyze the contribution of cooperative structures and RLNC to transmission performance, while random topology is used to evaluate the overall performance of the protocol relative to typical comparative protocols in a general network environment. This model is oriented towards a single-source, dual-destination, multi-hop underwater acoustic communication scenario, including one source node, two destination nodes, and several relay nodes. The source node sends the same generation of data to both destination nodes, and the relay nodes participate in hop-by-hop forwarding based on local topology relationships, remaining hop count, and link quality status. In each forwarding step, the protocol uses the current sending unit as the decision object, completing candidate structure generation, transmission budget calculation, and transmission mode selection. In this embodiment, the network coverage area is set as a 30 km × 30 km two-dimensional planar area. To verify the applicability of the protocol under different topology conditions, two network scenarios are set: regular topology and random topology. In the regular topology, network nodes are deployed in a uniform grid pattern, with an adjacent node spacing of 5 km; ensuring that nodes can establish single-hop connections with adjacent nodes and diagonally opposite nodes, the single-hop communication range is set to approximately 7.15 km. In the random topology, relay nodes are randomly distributed within the same network coverage area, and connectivity between nodes is generated using the same communication range constraint. Simulation parameter settings are shown in Table 1. Table 1 Simulation parameter settings

[0024] 2. Network Initialization and Encoding Vector Dictionary Configuration: Before the protocol runs, each node in the network first obtains local network state information and completes the encoding vector dictionary configuration required for random linear network encoding. The specific process is as follows: (2.1) Each node in the network obtains its own one-hop neighbor relationship based on the communication radius. Let node... The communication radius is Then the set of one-hop neighbors of a node is: ; in, Represents a node With nodes The distance.

[0025] (2.2) Each node in the network obtains its remaining hop count information for each destination node. For a node... Target node Let its remaining number of jumps be . Based on the aforementioned one-hop neighbor set, the node Target node The forward candidate set is: ; The forward candidate set is used to generate candidate transmission structures for subsequent shared forwarding, cooperative forwarding, and branch forwarding.

[0026] (2.3) For any link Based on the estimated initial packet error rate of the link Initialize the link success rate estimate: ; in, Used for subsequent transmission budget calculations and transmission mode selection.

[0027] (2.4) For a generation including The random linear network encoding process for each original group, where each node in the network is pre-configured with the same encoding vector dictionary: ; in Indicates the first 10 candidate encoding vector groups This is the size of a dictionary. Each node further establishes a mapping relationship between the encoding index and the encoding vector: ; and satisfy ; Therefore, the sending node can identify the encoding vector corresponding to the random linear network encoded data packet through the encoding index, and the receiving node can recover the corresponding encoding vector according to the encoding index.

[0028] 3. Encoded Data Generation: The protocol uses a unified encoded data packet format, such as... Figure 2 As shown, it consists of a protocol header and a coded payload. The protocol header carries control information such as the current generation number, next-hop information, transmission mode, and coding index. The coded payload is the encoded result of the current generation data packets after random linear network coding. The specific process is as follows: (3.1) The source node divides the data to be transmitted into several generations, each generation containing The original groups are denoted as: ; in, Indicates the first in the current generation The original groups.

[0029] (3.2) The source node performs a random linear combination of the original packets within the same generation over a finite field to generate the coded payload: ; in, For the generated first Group coding payload, For the first The first original group corresponding to the first The coding coefficients in the candidate coding vector group. The sending node determines the coding index corresponding to the global coding vector according to the pre-configured coding vector dictionary, and carries the coding index CodeID in the coded data packet to identify the coding vector corresponding to the current coded data packet through CodeID.

[0030] (3.3) The source node constructs an encoded data packet based on the encoded payload and the corresponding encoded index; the encoded data packet includes a protocol header and an encoded payload, the protocol header including at least a GenID, a NextHopID, a Transmission Mode, and a CodeID; wherein, GenID is used to organize the encoding, reception and recovery process by generation, NextHopID and Transmission Mode are used to describe the subsequent advancement mode of the current encoded data packet, and CodeID is used to recover the encoding relationship and perform linear independence determination.

[0031] 4. Candidate Transmission Structure Generation: The three transmission path structures—shared forwarding, cooperative forwarding, and branch forwarding—are respectively as follows: Figure 4 , Figure 5 and Figure 6As shown. During hop-by-hop transmission, the protocol uses the current sending unit... Based on the one-hop neighbor relationship, remaining hop count information, and inter-node communication reachability, three candidate transmission structures—shared forwarding, cooperative forwarding, and branch forwarding—are generated for decision-making. The specific process is as follows: (4.1) In each forwarding step, the current sending unit is determined according to the transmission mode of the previous hop. If the previous hop is a shared forwarding mode, then the current sending unit is a single node, denoted as . If the previous hop is a cooperative forwarding mode or a branch forwarding mode, then the current sending unit is a node pair consisting of two nodes, denoted as . , where nodes and nodes They advance towards different target nodes.

[0032] (4.2) The current sending unit generates a forward candidate set for different destination nodes based on the one-hop neighbor set and the remaining hop count information for the dual-destination node.

[0033] If the current sending unit is a single node Then construct the target-oriented node respectively. and forward candidate set ; in, For nodes A set of neighbors that jumps over time. and These represent the current node's orientation towards the destination node. and Reference remaining jumps.

[0034] If the current sending unit is a node pair Then construct the target-oriented node respectively. and forward candidate set ; (4.3) The current sending unit filters the candidates that can simultaneously face the destination node based on the forward candidate set. and Maintain a single candidate node that moves forward and form it into a shared forwarding candidate transmission structure.

[0035] The current sending unit is a single node. At the same time, the shared forwarding candidate node belongs to and The current sending unit is a node pair. At the same time, the shared forwarding candidate node simultaneously satisfies the condition relative to the node. Target node The conditions for advancement, and the conditions relative to the node Target node The forward propagation condition. The shared forwarding candidate transmission structure is used to enable the same generation of coded data to continue along a single path towards both destination nodes.

[0036] (4.4) The current sending unit filters candidate node pairs that can form a cooperative advancement relationship based on the forward candidate set and the inter-node communication reachability relationship, and constructs them into a cooperative forwarding candidate structure.

[0037] If the current sending unit is a node pair Then from respectively and Select candidate nodes and When candidate nodes pair Simultaneously satisfy ; At that time, it is considered that the current sending node is... Pairs with candidate nodes The conditions for collaborative advancement must be met between them, and As a candidate structure for cooperative forwarding, this condition indicates the candidate receiving node. and All can receive data from the current sending node. and The encoded data packets enable two candidate receiving nodes to jointly accumulate, store, and utilize the same generation of encoded information.

[0038] If the current sending unit is a single node Then first from and Select candidate node pairs Since the current sending unit is a single node, the current hop is only formed by a single node. To candidate node pairs The broadcast is forwarded; at this time, if the candidate node... In the next hop, it can pair with its subsequent candidate nodes. satisfy: ; Then it is considered that the candidate node pair It can form a cooperative forwarding structure in the next hop and use it as a candidate structure for cooperative forwarding in the current single-node sending scenario.

[0039] The cooperative forwarding candidate structure is used to enable the distributed accumulation of the same generation of encoded information between two receiving nodes, and to improve the utilization efficiency and recovery reliability of encoded information by the joint forwarding of the two receiving nodes by the two sending nodes in the subsequent transmission process.

[0040] (4.5) The current sending unit filters the candidates for each target node according to the forward candidate set. and Advance candidate node pairs; when the candidate node pairs cannot form a cooperative forwarding candidate transmission structure and a shared forwarding transmission structure, they are configured as a branch forwarding candidate transmission structure.

[0041] If the current sending unit is a single node Then the branch forwarding candidate structure consists of those belonging to... and candidate node pairs Composition; the current sending unit is a node pair Then the branch forwarding candidate structure consists of those belonging to... and The candidate node pairs constitute the branch forwarding candidate structure. The branch forwarding candidate structure is used to differentiate the same generation of coded data from the current sending unit into transmission branches that are respectively directed to two destination nodes.

[0042] (4.6) The current sending unit obtains the shared forwarding candidate structure set, the cooperative forwarding candidate transmission structure set, and the branch forwarding candidate structure set respectively, and uses the three types of candidate structure sets and their two-hop subsequent structure states as inputs for subsequent sending budget calculation, structure benefit evaluation, and transmission mode selection.

[0043] 5. Transmission Budget and Structure Benefit Assessment: After obtaining three candidate transmission structures—shared forwarding, cooperative forwarding, and branch forwarding—the protocol further calculates the transmission budget for each candidate structure based on link quality estimation, remaining hop count, and end-to-end reliability targets. It then calculates the structure benefit by considering the subsequent states formed within a two-hop range. The specific process is as follows: (5.1) During transmission, the current sending unit updates the success rate of the reachable link based on the initial link success rate estimate, the cumulative number of packets sent, and the cumulative number of packets successfully received. For the link... , its first The estimated link success rate after round 10 is: ; in, For link The initial success rate estimate, This is the equivalent prior sample size. As of the date wheel node To the node The cumulative number of packets sent. As of the date wheel node To the node The cumulative number of packets successfully sent.

[0044] (5.2) The current transmitting unit calculates the local reliability threshold that must be met in the current step based on the end-to-end reliability target and the remaining transmission hops. Let the end-to-end reliability target be... The current sending unit is Its target node The remaining transmission hops are Then the local reliability threshold is: ; The current sending unit is a single node. hour, The current sending unit is a node pair. hour, The number of remaining hops in the direction of the corresponding destination node in the node pair is determined.

[0045] For each generation, it includes The current transmitting unit calculates, based on the link success rate estimate and the local reliability threshold, the encoded data of the original packets, such that the receiving node or the receiving node pair can obtain at least [amount] with a reliability not lower than the local reliability threshold. The minimum number of transmissions required to encode linearly independent packets is used as the basic transmission budget.

[0046] The transmission budget from a single sending node to a single receiving node is: ; in, Represents a node To the node send Each encoded group, node The number of linearly independent coded blocks successfully obtained.

[0047] The transmission budget from a single transmitting node to two receiving nodes is: ; in, and Representing the receiving nodes and The number of linearly independent coded blocks successfully obtained.

[0048] The transmission budget from two transmitting nodes to a single receiving node is: ; in, Represents a node and Send separately and After each encoded group, the receiving node The number of valid encoded blocks successfully obtained.

[0049] The transmission budget from the dual transmitting node to the dual receiving node is: ; in, and Representing the receiving nodes and The number of linearly independent coded blocks successfully obtained.

[0050] The four types of transmission budgets mentioned above are used to calculate the transmission cost during shared forwarding, cooperative forwarding, branch forwarding, and the conversion between different transmission structures, respectively. For any transmission mode... Let the cost of its candidate structure be... ,in , These represent shared forwarding, cooperative forwarding, and branch forwarding transmission structures, respectively.

[0051] (5.3) For candidate structures in transmission mode, the current transmitting unit calculates its expected structure utility based on the state transition results within a two-hop range. , in , These represent shared forwarding, cooperative forwarding, and branch forwarding transmission structures, respectively.

[0052] ; in, and Representing the patterns respectively The transmission status after the first and second hops. This represents the structural utility corresponding to the two-hop termination state.

[0053] If a single-node state is formed after two hops ,in Given the linearly independent encoded blocks currently held by this node, then: ; in The original number of data groups in the first generation, if a node pair state is formed after two hops. ,in The number of linearly independent coded blocks shared by both sides. and Given the number of valid coded blocks unique to each side, the number of linearly independent blocks currently held by the nodes on both sides is: ; For cooperative node pairs, the structural utility is: ; For branching node pairs of states, its structural utility is: ; in The number of original groups in the first generation of data. The maximum transmission budget for one round of transmission, This is an indicator function.

[0054] (5.6) The current transmitting unit calculates the structural benefit under the corresponding transmission mode based on the candidate structure utility and candidate structure cost: ; in, For pattern The structural utility of the next candidate structure in the two-hop subsequent states. For pattern The transmission budget cost corresponding to the next candidate structure.

[0055] 6. Transmission Mode Selection and Encoded Forwarding: After completing the transmission budget and structural benefit evaluation, the protocol determines the transmission mode of the current sending unit based on the structural benefits of three candidate transmission structures: shared forwarding, cooperative forwarding, and branch forwarding, and then executes the transmission or forwarding of encoded data packets. The specific process is as follows: (6.1) Current Transmission Unit Feasibility assessments are performed on three candidate transmission structures: shared forwarding, cooperative forwarding, and branch forwarding. If the set of candidate structures corresponding to a certain transmission mode is empty, the structural benefit of that transmission mode is denoted as... And excluded in subsequent transmission mode comparisons.

[0056] (6.2) For feasible candidate transmission modes The current sending unit compares the structural benefits of three candidate transmission structures: shared forwarding, cooperative forwarding, and branch forwarding, and selects the transmission mode with the highest structural benefit as the current forwarding decision: ; in, These represent shared forwarding, cooperative forwarding, and branch forwarding transmission structures, respectively. For pattern The corresponding structural benefits.

[0057] (6.3) The current transmitting unit, according to the selected transmission mode The corresponding next-hop node or node pair and the current hop sending budget are used to generate and forward encoded data packets.

[0058] (6.4) After receiving the encoded data packet, the receiving node determines the data generation according to the generation number in the encoded data packet and recovers the corresponding encoded vector according to the encoding index. The receiving node performs a linear correlation judgment between the encoded vector and the received encoded vector of the current generation. If the encoded group can improve the rank of the receiving matrix of the current generation, it is accumulated as a linearly independent encoded group.

[0059] (6.5) The receiving node updates the receiving matrix for the accumulated linearly independent coded packets of the current generation. Assume the current generation contains... One original data group, The receiving matrix is ​​composed of the encoding vectors corresponding to the currently received encoded packets of the receiving node; when the receiving matrix satisfies: ; At this time, the receiving node completes the recovery of the current generation of data; by performing Gaussian elimination, the corresponding encoded vector can be recovered. ; ; in, For the r-th candidate encoding vector group, Let N be the received encoded vector, GF be the number of generated or received encoded packets, m be the exponent of the size of the finite field, and m be the number of the finite field. If the receiving node is a relay node, the recovered data is re-randomized using linear network coding based on the current transmission mode selection result and forwarded. If the receiving node is a destination node, the reception of the current generation of data at the destination node is completed. Transmission ends after the current generation of data has been recovered at all destination nodes.

[0060] Figure 7 This is a schematic diagram comparing the delivery performance of different transmission mechanisms under different packet error rates in the embodiments. Figure 7 (a) shows the change in the Combined Delivery Rate (JPDR) as a function of the error rate. Figure 7 (b) shows the change in single-destination delivery rate (PDR) with packet error rate; The figure compares four transmission mechanisms: the first is the protocol described in this invention, which employs cooperative path organization combined with random linear network coding; the second is the cooperative normal transmission mechanism (CU), which uses a cooperative path structure but does not perform random linear network coding; the third is the single-branch non-coded transmission mechanism, which performs normal packet forwarding (SU) along a single path; and the fourth is the single-branch network-coded transmission mechanism (SC), which performs random linear network coding forwarding along a single path. Simulation results show that as the link packet error rate increases, the delivery performance of each transmission mechanism decreases. Compared to other mechanisms, the protocol described in this invention maintains a high level in both joint delivery rate and single-destination delivery rate, indicating that the combination of cooperative path organization and random linear network coding can improve the joint recovery capability in bi-directional multi-hop transmission.

[0061] Figure 8 This is a schematic diagram comparing the delivery performance of different transmission mechanisms under different hop count conditions in the embodiments. Figure 8 (a) shows the change in the combined delivery rate (JPDR) with the number of hops. Figure 8 (b) shows the change in single-destination delivery rate (PDR) with the number of hops; This figure also compares the protocol described in this invention, the cooperative ordinary transmission mechanism, the single-branch non-coded transmission mechanism, and the single-branch network-coded transmission mechanism to examine the reliability changes of different transmission mechanisms as the end-to-end hop count increases. Simulation results show that as the transmission hop count increases, the delivery rate of each mechanism generally decreases; among them, the protocol described in this invention shows a smaller decrease, indicating that it can mitigate the impact of increased multi-hop depth on end-to-end reliability through information complementarity on the cooperative path and the accumulation of degrees of freedom in network-coded packets.

[0062] Figure 9 This is a schematic diagram comparing the performance of different protocol combination schemes under random topology and different packet error rates in the embodiments. Figure 9 (a) shows the change in the Combined Delivery Rate (JPDR) as a function of the error rate. Figure 9(b) shows the change in average number of transmissions with the packet error rate. This figure compares the protocol described in this invention, the Repeated Redundancy Transmission scheme based on vector geographic forwarding path (VBF+REP), the Random Linear Network Coding scheme based on vector geographic forwarding path (VBF+RLNC), the Repeated Redundancy Transmission scheme based on focused beam routing path (FBR+REP), and the Random Linear Network Coding scheme based on focused beam routing path (FBR+RLNC). The figure shows the results of the joint delivery rate and average number of transmissions changing with the link packet error rate. The simulation results show that as the packet error rate increases, the joint delivery rate of each scheme decreases overall, while the average number of transmissions increases overall. In contrast, the protocol described in this invention can maintain a higher joint delivery rate under higher packet error rate conditions and control the transmission overhead required for a single successful delivery.

[0063] Figure 10 This is a schematic diagram comparing the performance of different protocol combinations under random topology and different numbers of nodes in the embodiments. Figure 10 (a) shows the change in Joint Delivery Rate (JPDR) as a function of the number of nodes. Figure 10 (b) shows the change in average number of transmissions with the number of nodes. This figure also compares the protocol described in this invention, the redundant transmission scheme based on vector geographic forwarding paths, the random linear network coding scheme based on vector geographic forwarding paths, the redundant transmission scheme based on focused beam routing paths, and the random linear network coding scheme based on focused beam routing paths to examine the impact of changes in the number of network nodes on the reliable transmission performance and transmission overhead of bi-target networks. Simulation results show that as the number of nodes increases, the number of available relay nodes and optional paths in the network increases. The protocol described in this invention can more fully utilize the complementary relationship between local topology and coding information, improve the joint delivery performance of bi-target networks, and maintain a low average number of transmissions.

[0064] Example 2 This embodiment provides a design method for a reliable transmission protocol in a cooperative underwater acoustic communication network based on random linear network coding, including: 1. Network initialization and encoding vector dictionary configuration phase: (1.1) Each node in the network obtains its own one-hop neighbor relationship based on the communication radius. Let node... The communication radius is Then the set of one-hop neighbors of a node is: ; in, Represents a node With nodes The distance.

[0065] (1.2) Each node in the network obtains its remaining hop count information and initial link quality estimate for the two destination nodes; for any link Based on the estimated initial packet error rate of the link Initialize the link success rate estimate: ; in, Used for subsequent transmission budget calculations and transmission mode selection.

[0066] (1.3) Each node in the network is pre-configured with the same encoding vector dictionary and establishes a mapping relationship between the encoding vector and the encoding index, so that the sending node can identify the encoding vector corresponding to the random linear network encoded data packet through the encoding index, and the receiving node can recover the corresponding encoding vector according to the encoding index.

[0067] 2. Encoded data generation stage: (2.1) The sending node divides the data to be transmitted into several generations, each generation containing The original groups are denoted as: ; in, Indicates the first in the current generation The original groups.

[0068] (2.2) The transmitting node performs random linear combination of the original packets within the same generation over a finite field to generate the coded payload: ; in, For the generated first Group coding payload, For the first The first original group corresponding to the first The coding coefficients in the candidate coding vector group.

[0069] (2.3) The sending node determines the encoding vector and its encoding index corresponding to the encoding payload according to the pre-configured encoding vector dictionary, and generates an encoding data packet carrying the encoding index; the encoding index is used to identify the encoding vector corresponding to the current encoding data packet.

[0070] 3. Candidate structure evaluation and transmission mode selection stage: (3.1) In each forwarding step, the protocol first determines the current sending unit. The current sending unit can be a single node or a node pair consisting of two nodes.

[0071] (3.2) Current Transmission Unit Based on the one-hop neighbor relationship and the remaining hop count information, construct a forward candidate set for nodes with dual destinations.

[0072] If the current sending unit is a single node Then construct the target-oriented node respectively. and forward candidate set ; in, For nodes A set of neighbors that jumps over time. and These represent the current node's orientation towards the destination node. and Reference remaining jumps.

[0073] If the current sending unit is a node pair Then construct the target-oriented node respectively. and forward candidate set ; (3.3) The protocol generates three types of candidate transmission structures based on the aforementioned forward candidate set: shared forwarding, cooperative forwarding, and branch forwarding. The shared forwarding transmission structure indicates that the current sending unit selects a candidate node that can simultaneously advance towards two destination nodes, allowing the same generation of encoded data to continue forwarding along the shared advancement direction. The cooperative forwarding structure indicates that the current sending unit selects a pair of candidate nodes that advance towards two destination nodes respectively and can form a cooperative advancement relationship in the current hop or the next hop. When the current sending unit is a pair of nodes, the cooperative advancement relationship is determined by the dual-transmission and dual-reception reachability relationship between the current sending node pair and the candidate receiving node pair. When the current sending unit is a single node, the cooperative advancement relationship is determined by whether the current candidate node pair can form a cooperative node pair in the next hop. The branch forwarding structure indicates that the current sending unit selects a pair of candidate nodes that advance towards two destination nodes respectively and have not formed a cooperative forwarding structure, allowing the same generation of encoded data to continue advancing towards two destination directions respectively.

[0074] (3.4) For the three candidate transmission structures, the protocol calculates the transmission budget based on the link quality estimate, remaining hop count, and end-to-end reliability target, and calculates the structure benefit based on the subsequent transmission states formed within a two-hop range. For any transmission mode... Its structural returns are expressed as: ; in, For pattern The structural utility of the next candidate structure in the two-hop subsequent states. For pattern The transmission budget cost corresponding to the next candidate structure.

[0075] (3.5) The current sending unit selects the transmission mode with the greatest structural benefit as the current forwarding decision: ; 4. Encoding, forwarding, and data recovery phase: (4.1) The current sending unit performs encoded transmission or forwarding according to the optimal transmission mode determined in step 3. When the shared forwarding mode is selected, the encoded data packet is sent to a single next-hop node; when the cooperative forwarding mode or the branch forwarding mode is selected, the encoded data packet is sent to the corresponding next-hop node pair.

[0076] (4.2) After receiving the encoded data packet, the next hop node or node pair restores the corresponding encoding vector according to the encoding index and updates the current generation encoding state; if the newly received encoded packet can improve the rank of the current generation receiving matrix, it is accumulated as a linearly independent encoded packet.

[0077] (4.3) When the current generation of the receiving matrix of the receiving node satisfies: ; At that time, the receiving node completes the recovery of the current generation of data; among which, The number of original data groups contained in the current generation. It is the reception matrix composed of the encoding vectors corresponding to the currently received encoded packets of the receiving node.

[0078] (4.4) If the node that has completed the recovery is a relay node, then the node re-encodes the recovered current generation data using random linear network coding and uses it as the current sending unit in the subsequent hop-by-hop transmission process. Continue participating in step 3, candidate structure evaluation and transmission mode selection; if the node that has completed the recovery is the destination node, then record that the destination node has successfully recovered the current generation of data.

[0079] (4.5) The current generation data transmission ends when both destination nodes have completed the current generation data recovery; if the recovery of both destination nodes has not been completed, the subsequent hop-by-hop forwarding process continues according to the current receiving status.

[0080] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent substitutions to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A design method for a reliable transmission protocol in a cooperative underwater acoustic communication network based on random linear network coding, comprising: Step 1: Each node in the network obtains its own one-hop neighbor relationship, the remaining hop count for bi-destination nodes, the connectivity relationship between nodes, and the initial estimate of link quality. Step 2: Configure the same encoding vector dictionary for each node; Step 3: The source node generates an encoded data packet and sets it as the current sending unit. The current sending unit is the node or node pair that sends or forwards the encoded data packet. Step 4: The current sending unit constructs a forward candidate set based on the current one-hop neighbor relationship and the remaining hop count information, filters candidate sending units based on the forward candidate set, and forms them into three types of candidate transmission structures: shared forwarding, cooperative forwarding, or branch forwarding. Step 5: For each candidate transmission structure, the current sending unit calculates the corresponding transmission budget and selects the candidate transmission structure with the greatest benefit as the current forwarding decision. Step 6: Based on the forwarding decision in Step 5, the current sending unit generates and sends the encoded data packet. The receiving unit receives the encoded data packet, determines whether the current node is the destination node, and then recovers the encoded data packet according to the encoded vector dictionary. If it is the destination node, the data transmission is completed; otherwise, the current node becomes the current sending unit in the subsequent hop-by-hop transmission process and proceeds to Step 4. The receiving unit is the node or node pair that received the encoded data packet.

2. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 1, characterized in that, Step 2 specifically involves: each node in the network pre-configuring the same encoding vector dictionary and establishing a mapping relationship between encoding vectors and encoding indices, so that the current sending node can identify the encoding vector corresponding to the random linear network encoded data packet through the encoding index, and the receiving node can recover the corresponding encoding vector according to the encoding index.

3. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 2, characterized in that, The encoded data packet in step 3 includes a protocol header and an encoded payload. The protocol header includes a GenID, a NextHopID, a Transmission Mode, and a CodeID corresponding to the encoded payload. The encoded payload is constructed by the current sending unit dividing the data to be transmitted into several generations and then performing random linear network coding on the data groups within the same generation.

4. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 1, characterized in that, The specific steps for constructing the forward candidate set in step 4 are as follows: If the current sending unit is a single node Then construct the target-oriented node respectively. and Forward candidate set: ； in, For nodes A set of neighbors that jumps over time. and These represent the current node's orientation towards the destination node. and Reference remaining hops For a single node The forward candidate set, For a single node The forward candidate set; If the current sending unit is a node pair Then construct the target-oriented node respectively. and Forward candidate set: ； in, For nodes The forward candidate set, For nodes The forward candidate set.

5. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 4, characterized in that, In step 4, the conditions for selecting candidate sending units and classifying them into three candidate transmission structures—shared forwarding, cooperative forwarding, or branch forwarding—are as follows: 1) The current sending unit is a single node. At the same time, the shared forwarding candidate node belongs to and Forward candidate set; Or the current sending unit is a node pair At the same time, the shared forwarding candidate node simultaneously satisfies the condition relative to the node. Target node The conditions for advancement, and the conditions relative to the node Target node Forward propulsion conditions; Then the candidate node constitutes a shared forwarding candidate transmission structure and is added to the shared forwarding candidate transmission structure set; 2) The current sending unit is a single node. Then first from and Select candidate node pairs The current jump is formed only by a single node. To candidate node pairs The broadcast is forwarded; at this time, if the candidate node... In the next hop, it can pair with its subsequent candidate nodes. satisfy: ； Or the current sending unit is a node pair At that time, respectively from and Select candidate nodes and When candidate nodes pair Simultaneously satisfy: ； Then the candidate node constitutes a cooperative forwarding candidate transmission structure and is added to the cooperative forwarding candidate transmission structure set; 3) The current sending unit is a single node. The branch forwarding candidate transmission structure consists of those belonging to... and candidate node pairs constitute; Or the current sending unit is a node pair The branch forwarding candidate transmission structure consists of those belonging to... and The candidate node pairs constitute; Then the candidate node constitutes a branch forwarding candidate transmission structure and is added to the branch forwarding candidate transmission structure set; The three sets of candidate transmission structures serve as inputs for subsequent transmission budget calculations, structure benefit assessments, and transmission mode selection.

6. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 1, characterized in that, In step 5, the candidate transmission structure that yields the greatest benefit is selected. Satisfy the following formula: ； in, For the benefit of transmission structure, These represent shared forwarding, cooperative forwarding, and branch forwarding transmission structures, respectively.

7. The design method for a reliable transmission protocol of a cooperative underwater acoustic communication network based on random linear network coding as described in claim 1, characterized in that, In step 6, the receiving unit recovers the received encoded data packet, specifically as follows: The receiving node determines the data generation based on the generation number in the encoded data packet and recovers the corresponding encoded vector from the encoding index according to the encoded vector dictionary. The receiving node is a single-node receiving unit, and it performs a linear correlation judgment between the encoded vector and the already received encoded vectors of the current generation. If the encoded packet can increase the rank of the current generation's receiving matrix, it is accumulated as a linearly independent encoded packet. When the receiving node obtains no less than a preset decoding threshold of linearly independent encoded packets, that is, when the receiving matrix satisfies: ； in, The number of original data groups contained in the current generation. The receiving matrix is ​​composed of the encoding vectors corresponding to the currently received encoded packets by the receiving node. To perform the rank operation on a matrix; After performing Gaussian elimination, the corresponding encoded vector can be recovered. ； ； in, For the r-th candidate encoding vector group, GF is the received encoded vector, N is the number of generated or received encoded packets, GF is the finite field, and m is the exponent of the size of the finite field.

Images