Unmanned aerial vehicle swarm final consensus model and asynchronous consensus method, electronic device, and medium

Abstract

This invention discloses a final consensus model and asynchronous consensus method for unmanned aerial vehicle (UAV) swarms, as well as electronic devices and media. The model includes: a task cluster blockchain for storing the task execution states within the UAV swarm, and a main cluster blockchain for storing the final mixed task execution state, which aggregates the task execution states from each task cluster blockchain. The final mixed task execution state has two states: a temporary soft state and a deterministic final state. The transition module between the temporary soft state and the deterministic final state is as follows: If the task execution state of a task cluster fails to be submitted to the main cluster blockchain in period t, a temporary soft state appears and is stored in the task cluster blockchain. This temporary soft state continues until the task execution state of task cluster i in the subsequent period n is successfully submitted to the main cluster blockchain. The missing temporary soft state is then aggregated with the task execution state of the task cluster in period n and stored in the main cluster blockchain. At this point, the temporary soft state transitions to the deterministic final state.

Description

Technical Field

[0001] This invention belongs to the field of blockchain consensus technology, specifically relating to a final consensus model and asynchronous consensus method for drone swarms, electronic devices, and media. Background Technology

[0002] Currently, drone swarm collaborative operations have gradually evolved from "single-platform remote control" to "intelligent collective autonomy." Distributed systems often employ strong consistency models based on the CAP theorem or consensus algorithms from blockchains to achieve state consistency. Both of these solutions require a continuous, stable network environment with a majority of members online and rely on globally strong consistency semantics, meaning that all honest nodes must have completely consistent data at any given time. With the expansion of drone swarm application scenarios, their dynamic nature, varied network topologies, and diverse tasks place higher demands on consensus technologies.

[0003] In distributed systems, the CAP theorem states that consistency means all nodes in the system appear to the client as if they are on the same machine, and read and write operations are executed in the order they arrive. In the CAP theorem, consistency, availability, and partition tolerance form an impossible triangle. To address the availability problem in distributed consistency systems, the BASE theory was proposed, representing a trade-off between consistency and availability in the CAP theorem. With the development of blockchain, research on distributed consensus models has entered a new phase.

[0004] Distributed consensus protocols are mainly divided into proof-based consensus protocols and voting-based consensus protocols. Proof-based consensus protocols require participants to prove their capabilities to gain the right to record transactions, broadcasting their state to other nodes to achieve distributed consensus. This protocol is a typical proof-based consensus protocol and can achieve complete distributed consensus. Voting-based consensus protocols require participants to negotiate and vote; when a state receives enough approval from participants, distributed consensus is achieved. Researchers have proposed a trust-driven, multi-sharded, hierarchical state consensus protocol to improve blockchain consensus throughput. Other researchers have proposed a consensus algorithm for distributed cooperative formation trajectory planning in UAV swarms, achieving convergence and consistency in the relative distance and speed between UAVs.

[0005] The existing methods have the following drawbacks:

[0006] 1. Overly strong network assumptions: Traditional consensus protocols require strong consistency, but drone swarms are affected by factors such as channel obstruction, node movement, and electromagnetic interference, resulting in unstable network links and the number of consensus participant nodes often falling below the protocol threshold, leading to consensus failure.

[0007] 2. Strong consistency in a single chain causes interruptions: Traditional consensus protocols require strong consistency. Once a consensus participant node fails beyond a certain threshold, the entire network must wait for a view switch or a re-election of consensus participant nodes, resulting in a loss of real-time performance.

[0008] 3. Failure to utilize the inherent characteristics of task clusters: Traditional consensus protocols require that the independent states of different task clusters be agreed upon as the same distributed consistent state, which increases the amount of communication and the probability of conflicts, and loses the flexibility of task clusters.

[0009] 4. Lack of consensus intermediate state management: Traditional consensus protocols only distinguish between consistent and inconsistent states, without the concept of intermediate states. This lack of state granularity leads to reduced system availability. Summary of the Invention

[0010] To address the aforementioned problems in existing technologies, this invention provides a final consensus model for unmanned aerial vehicle (UAV) swarms, an asynchronous consensus method, electronic devices, and media. This addresses issues such as overly strong consensus network assumptions, susceptibility to strong consistency consensus interruptions, and a lack of intermediate consensus states. It also meets the higher requirements for consensus technology imposed by the dynamic nature of UAV swarms, the variability of network topology, and the diversity of tasks.

[0011] To achieve the above objectives, the present invention adopts the following technical solution:

[0012] A drone swarm final consensus model, the model includes a drone swarm task execution state hybrid blockchain storage module consisting of a main cluster blockchain and multiple task cluster blockchains, and a temporary soft state and deterministic final state transition module;

[0013] The task cluster blockchain is used to store the task execution status within the drone swarm. The main cluster blockchain is used to store the final mixed task execution status that aggregates the task execution statuses within each task cluster blockchain. The final mixed task execution status has two states: a temporary soft state and a deterministic final state. The temporary soft state is an intermediate state that is recognized only by drone nodes within a single task cluster, while the deterministic final state is a final state that is recognized by all drone nodes in the drone swarm.

[0014] The temporary soft state and deterministic final state transition module is as follows: Let the state be... Periodic task cluster The task execution status failed to be submitted to the main cluster blockchain, resulting in a temporary soft state. It appears and is stored in the task cluster blockchain until the subsequent [number]th [number]. Periodic task cluster The task execution status was successfully submitted to the main cluster blockchain; the missing temporary soft state is missing. With the Periodic task cluster The task execution status is aggregated and stored in the main cluster blockchain, at which point it is in a temporary soft state. The transition is to a deterministic final state.

[0015] This invention also discloses an asynchronous consensus method for unmanned aerial vehicle (UAV) swarms, the specific steps of which are as follows:

[0016] Step S1: Based on blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements. Within the task cluster shard, drone nodes establish network connections with each other and execute predetermined tasks, while within the main cluster shard, drone nodes establish network connections with each other.

[0017] Step S2: Randomly select a drone node as an anchor node in each task cluster shard. The anchor node establishes a network connection with the drone nodes inside the main cluster shard. The original drone nodes inside the main cluster shard and these anchor nodes together form the main cluster shard.

[0018] Step S3: Within each task cluster shard, the drone nodes execute tasks and generate blockchain transactions, which contain the task execution status of the drone node.

[0019] Step S4: Within each task cluster shard, the drone nodes collectively reach a consensus on the blockchain transactions.

[0020] Step S5: Based on the blockchain transaction consensus result, each task cluster shard generates the latest task execution status within the cluster and stores the task execution status within the cluster in the form of a block on the task cluster blockchain. At this time, the task execution status within the cluster is a temporary soft state.

[0021] Step S6: Each task cluster shard anchor node generates a blockchain transaction, which contains the latest task execution status within the task cluster shard.

[0022] Step S7: The drone nodes within the main cluster shard jointly reach a final consensus on the blockchain transactions.

[0023] Step S8: Based on the blockchain transaction consensus result, the main cluster shard generates the latest task execution status within the group and stores the task execution status within the group in the form of blocks on the task cluster blockchain. At this time, the task execution status within the group is a deterministic final state.

[0024] In the above steps, the temporary soft state is an intermediate state that is only recognized by the drone nodes within a single task cluster, while the deterministic final state is a final state that is recognized by all drone nodes in the drone swarm. The above steps include the temporary soft state and the deterministic final state, forming a hybrid state. The above steps include intra-group consensus and final consensus, forming asynchronous consensus.

[0025] When the network connection is stable, the method executes steps S1-S8 to complete consensus; wherein, each task cluster shard executes steps S2-S5 in parallel, and the task cluster shards and the main cluster shards execute steps S3-S6 and S7-S8 asynchronously.

[0026] Preferably, the present invention also discloses an asynchronous consensus method under hybrid conditions, which assumes task cluster sharding when network connectivity is unstable. In the The cycle failed to execute step S4, i.e., task cluster sharding. The network connection of most drone nodes was interrupted, and consensus within the blockchain transaction group failed. At this time, no task execution status was generated within the group until subsequent task cluster sharding. In the Successfully executing step S4 of the cycle, i.e., task cluster sharding. The network connection of most drone nodes was re-established, and consensus within the blockchain transaction group was successful; the specific steps are as follows:

[0027] Step S1: Based on blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements. Within the task cluster shard, drone nodes establish network connections with each other and execute predetermined tasks, while within the main cluster shard, drone nodes establish network connections with each other.

[0028] Step S2, sharding the task cluster A drone node is randomly selected as the anchor node. The anchor node establishes a network connection with the drone nodes inside the main cluster shard. The original drone nodes inside the main cluster shard and the anchor node of each task cluster shard together form the main cluster shard.

[0029] Step S3, in the Periodicity, task cluster sharding Internal drone nodes execute tasks and generate blockchain transactions, which contain the task execution status of the drone node.

[0030] Step S4, due to task cluster sharding The network connection of most drone nodes was interrupted, and consensus within the blockchain transaction group failed.

[0031] Step S5, Task Cluster Sharding The drone node continues to execute steps S3-S4 until the... Periodicity, task cluster sharding The network connection of most drone nodes was re-established, and consensus within the blockchain transaction group was successful;

[0032] Step S6: Based on the blockchain transaction consensus results, the task cluster is sharded. The latest task execution status within the group is generated internally and stored in the task cluster blockchain in the form of blocks. At this time, the task execution status within the group is a temporary soft state.

[0033] Step S7: Each task cluster shard anchor node generates a blockchain transaction, which contains the latest task execution status within the task cluster shard.

[0034] Step S8: The drone nodes within the main cluster shard jointly reach a final consensus on the blockchain transactions.

[0035] Step S9: Based on the blockchain transaction consensus result, the main cluster shard generates the latest task execution status within the group and stores the task execution status within the group in the form of blocks on the task cluster blockchain. At this time, the task execution status within the group is a deterministic final state.

[0036] In this process, each task cluster shard executes steps S3-S7 in parallel, while the task cluster shards and the main cluster shards execute steps S3-S7 and S8-S9 asynchronously.

[0037] Preferably, the present invention also discloses an asynchronous consensus method under hybrid conditions, which assumes task cluster sharding when network connectivity is unstable. In the Step S6, i.e., task cluster sharding, was not successfully executed during the cycle. The anchor node loses its connection with the main cluster shard network, and the latest task execution status within the group fails to be submitted to the main cluster blockchain. At this point, the task execution status within the group is in a temporary soft state until a subsequent task is submitted to the main cluster shard. In the Successfully executing step S6 of the cycle, i.e., task cluster sharding. The anchor node re-establishes its connection with the main cluster shard network, and the latest task execution status within the cluster is successfully submitted to the main cluster blockchain. At this point, in step S6, the task cluster shards... The blockchain transactions generated by the anchor node aggregate the missing temporary soft state. With task cluster sharding No. Periodic task execution status The specific steps are as follows:

[0038] Step S1: Based on blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements. Within the task cluster shard, drone nodes establish network connections with each other and execute predetermined tasks, while within the main cluster shard, drone nodes establish network connections with each other.

[0039] Step S2: Randomly select a drone node as an anchor node in each task cluster shard. The anchor node establishes a network connection with the drone nodes inside the main cluster shard. The original drone nodes inside the main cluster shard and these anchor nodes together form the main cluster shard.

[0040] Step S3: Within each task cluster shard, the drone nodes execute tasks and generate blockchain transactions, which contain the task execution status of the drone node.

[0041] Step S4: Within each task cluster shard, the drone nodes collectively reach a consensus on the blockchain transactions.

[0042] Step S5: Based on the blockchain transaction consensus result, each task cluster shard generates the latest task execution status within the cluster and stores the task execution status within the cluster in the form of a block on the task cluster blockchain. At this time, the task execution status within the cluster is a temporary soft state.

[0043] Step S6, due to task cluster sharding The anchor node loses connection with the main cluster shard network, and the latest task execution status in the group fails to be submitted to the main cluster blockchain. At this time, the task execution status in the group is a temporary soft state.

[0044] Step S7, Task Cluster Sharding The drone node continues to execute steps S3-S6 until the... Periodicity, task cluster sharding The anchor node re-establishes its connection with the main cluster shard network, and the latest task execution status within the cluster is successfully submitted to the main cluster blockchain. At this point, the blockchain transactions generated by the anchor node aggregate the temporary soft state. and task cluster sharding No. Periodic task execution status ;

[0045] Step S8: The drone nodes within the main cluster shard jointly reach a final consensus on the blockchain transactions.

[0046] Step S9: Based on the blockchain transaction consensus result, the main cluster shard generates the latest task execution status within the group and stores the task execution status within the group in the form of blocks on the task cluster blockchain. At this time, the task execution status within the group is a deterministic final state.

[0047] In this process, each task cluster shard executes steps S2-S7 in parallel, while the task cluster shards and the main cluster shards execute steps S2-S7 and S8-S9 asynchronously.

[0048] The above technical solution is applicable to multi-cluster drone swarm application scenarios.

[0049] This invention presents a hybrid-state final consensus model that combines the soft-state semantics of BASE theory with directed acyclic graphs. It formally describes the derivation and merging relationships between the initial consistent state, the multi-parallel soft states, and the final consistent state, and defines strict mathematical constraints to achieve a controllable evolution from weak consistency to final consistency.

[0050] This invention provides an asynchronous consensus method under hybrid conditions. Based on ACID theory and directed acyclic graphs, it first dynamically divides the cluster into logical sub-clusters according to the cluster tasks, and performs asynchronous consensus in parallel within each sub-cluster. Then, a specially set logical master cluster receives and processes the state submissions of each sub-cluster in an asynchronous manner to achieve cross-cluster eventual consistency negotiation.

[0051] Each sub-cluster can independently maintain and evolve its state chain during the task. After the task ends, the sub-cluster state chain is safely merged into the main state chain as a soft fork through a secondary consensus process, thereby achieving the fusion of distributed consensus across multiple states.

[0052] The above technical solution is described in a simplified manner as follows:

[0053] When the network connection is stable, the method includes the following steps: a. Using blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements; b. Randomly select a drone in the task cluster shard as an anchor node and make it interconnected with the drone nodes in the main shard; c. According to the task execution process, each node in the sub-shard generates relevant transactions; d. Consensus is reached within the shard; e. The sub-shard generates the latest state based on the consensus result and completes the accounting in the form of blocks; f. The anchor node of the sub-shard generates a transaction, the content of which is the latest state of the sub-shard; g. Consensus is reached within the main shard; h. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of accounting is the deterministic final state. Among them, steps b to e are performed synchronously within the sub-shards and do not interfere with each other, while steps b to e and steps f to h are performed asynchronously in the sub-shards and the main shard, respectively.

[0054] When the network connection is unstable, the method includes the following steps: a. Using blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements; b. Randomly select a drone in the task cluster shard as an anchor node and make it interconnected with the drone nodes in the main shard; c. According to the task execution process, each node in the sub-shard generates relevant transactions; d. Consensus is reached within the shard; e. The sub-shard generates the latest state based on the consensus result and completes the accounting in the form of blocks; f. Due to the unstable network connection, it can be assumed that any sub-shard has not completed step e and that the sub-shard completes step e within the subsequent (m-1) consensus cycles. The anchor node of the sub-shard that has completed step e generates a transaction, and the content of this transaction is the latest state of the sub-shard; g. Consensus is reached within the main shard; h. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of the accounting is a temporary soft state. i. In the next consensus cycle, repeat steps b-g above. It is assumed that all sub-shards have completed step e at this point. j. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. The latest accounting state at this point is a deterministic final state that includes the temporary soft state of the previous cycle and the latest states of all sub-shards in the current cycle. Steps b-e are performed synchronously within the sub-shards and do not interfere with each other. Steps b-e and steps f-j are performed asynchronously in the sub-shards and main shard, respectively.

[0055] When the network connection is unstable, the method may also include the following steps: a. Using blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements; b. Randomly select a drone in the task cluster shard as an anchor node and make it interconnect with the drone nodes in the main shard; c. According to the task execution process, each node in the sub-shard generates relevant transactions; d. Consensus is reached within the shard; e. The sub-shard generates the latest state based on the consensus result and completes the accounting in the form of blocks; f. Due to the unstable network connection, it can be assumed that any sub-shard has not completed step e and that the sub-shard has not completed step e in the subsequent (m-1) consensus cycles. The anchor node of the sub-shard that has completed step e generates a transaction, and the content of this transaction is the latest state of the sub-shard; g. Consensus is reached within the main shard; h. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest accounting state is a temporary soft state. i. In the subsequent (m-1)th consensus cycle, repeat steps b~g above; j. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest accounting state is a temporary soft state, and at the same time, the temporary soft state before the (m-1)th consensus cycle is confirmed as the deterministic final state. Among them, steps b~e are performed synchronously in the sub-shards and do not interfere with each other, while steps b~e and steps f~j are performed asynchronously in the sub-shards and the main shard, respectively.

[0056] When the network connection is unstable, the method may also include the following steps: a. Using blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to task requirements; b. Randomly select a drone in the task cluster shard as an anchor node and make it interconnect with the drone nodes in the main shard; c. According to the task execution process, each node in the sub-shard generates relevant transactions; d. Consensus is reached within the shard; e. The sub-shard generates the latest state based on the consensus result and completes the accounting in the form of blocks; f. Due to the unstable network connection, it can be assumed that there are any number of sub-shards that have not completed step e, and that any number of sub-shards have not completed step e in the subsequent (m-1) consensus cycles. The anchor node of the sub-shard that has completed step e generates a transaction, and the content of this transaction is the latest state of the sub-shard; g. Consensus is reached within the main shard; h. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of the accounting is a temporary soft state. i. In the subsequent (m-1)th consensus cycle, repeat steps b~g above; j. The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest accounting state is a temporary soft state, and at the same time, the temporary soft state before the (m-1)th consensus cycle is confirmed as the deterministic final state. Among them, steps b~e are performed synchronously in the sub-shards and do not interfere with each other, while steps b~e and steps f~j are performed asynchronously in the sub-shards and the main shard, respectively.

[0057] The present invention also discloses an electronic device, including a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to execute the above-described model or method.

[0058] The present invention also discloses a computer-readable storage medium, characterized in that the computer-readable storage medium stores a plurality of instructions, which are adapted for loading by a processor to execute the above-described model or method.

[0059] Compared with the prior art, the present invention has the following technical effects.

[0060] This invention solves the problems of overly strong assumptions in consensus networks, easy occurrence of strong consensus interruption, and lack of consensus intermediate states, and meets the higher requirements of consensus technology for the dynamic nature of UAV swarms, the variability of network topology, and the diversity of tasks. Attached Figure Description

[0061] Figure 1 This is a diagram of a drone swarm network model; Figure 2 This is a preferred embodiment of the UAV swarm hybrid blockchain storage module architecture diagram of the present invention; Figure 3This is a diagram of the asynchronous consensus process (nth round of consensus) for a drone swarm according to a preferred embodiment of the present invention; Figure 4 This is a diagram of the asynchronous consensus process (n+2 rounds of consensus) for a drone swarm according to a preferred embodiment of the present invention; Figure 5 This is the average time delay diagram confirming the deterministic final state in a preferred embodiment of the present invention; Figure 6 This is a consensus process diagram of the architecture under a preferred embodiment of the present invention in a static topology; Figure 7 This is a consensus process diagram of the prior art solution under the static topology of the preferred embodiment of the present invention; Figure 8 This is a consensus process diagram of the architecture under the dynamic topology in a preferred embodiment of the present invention. Detailed Implementation

[0062] The technical solutions of the present invention will now be clearly and completely described in conjunction with the accompanying drawings of the embodiments of the present invention.

[0063] This invention provides a final consensus model and asynchronous consensus method for unmanned aerial vehicle (UAV) swarms, addressing issues such as overly strong consensus network assumptions, susceptibility to strong consistency consensus interruptions, and a lack of intermediate consensus states. It also meets the higher requirements for consensus technology imposed by the dynamic nature of UAV swarms, the variability of network topology, and the diversity of tasks. The embodiments of this invention are described in detail below.

[0064] A preferred embodiment of the present invention provides a final consensus model for a drone swarm, comprising a hybrid blockchain storage module for the drone swarm task execution state, consisting of a main cluster blockchain and multiple task cluster blockchains, and a module for transitioning between a temporary soft state and a deterministic final state. The task cluster blockchain stores the task execution state within the drone swarm, while the main cluster blockchain stores the final hybrid task execution state, which aggregates the task execution states from each task cluster blockchain. This final hybrid task execution state has two states: a temporary soft state and a deterministic final state. The temporary soft state is an intermediate state recognized only by drone nodes within a single task cluster, while the deterministic final state is a final state recognized by all drone nodes in the drone swarm. The module for transitioning between the temporary soft state and the deterministic final state is as follows: Let the... Periodic task cluster The task execution status failed to be submitted to the main cluster blockchain, resulting in a temporary soft state. It appears and is stored in the task cluster blockchain until the subsequent [number]th [number]. Periodic task cluster The task execution status was successfully submitted to the main cluster blockchain; the missing temporary soft state is missing. With the Periodic task cluster The task execution status is aggregated and stored in the main cluster blockchain, at which point it is in a temporary soft state. The transition is to a deterministic final state.

[0065] The preferred embodiment of the present invention provides a method for asynchronous consensus among unmanned aerial vehicle (UAV) swarms, which is described in detail below.

[0066] Figure 1 This invention provides a preferred embodiment of a drone swarm network model, applied to multi-task cluster drone swarm collaborative operation scenarios. In this embodiment, the model consists of a swarm of 15 drones, assigned to monitor a rectangular area and track moving targets within that area. The network model construction specifically includes the following steps:

[0067] 1. Based on the task cluster division of the drone swarm, 12 out of the 15 drones are evenly divided into 3 task clusters, and the remaining 3 drones are in an idle state.

[0068] 2. Based on blockchain sharding technology, each task cluster is divided into a sub-shard;

[0069] 3. Randomly select one drone from each task cluster as the anchor node;

[0070] 4. All anchor node drones in the task cluster and the 3 drones in an idle state together form the main fragment;

[0071] 5. Each drone in a sub-segment establishes a point-to-point communication link with the anchor node drone in the cluster, and all drones in the main segment establish point-to-point communication links. At this point, the drone swarm network model is completed.

[0072] Figure 2 This invention provides a hybrid blockchain storage module for drone swarms, applicable to a multi-task cluster drone swarm collaborative operation network model.

[0073] This module is a hybrid storage model consisting of four blockchains: one main blockchain and three sub-blockchains.

[0074] according to Figure 1 In the network model described, each shard maintains a blockchain, where the blockchain maintained by the main shard is called the main blockchain, and the blockchain maintained by the sub-shards is called the sub-blockchain.

[0075] Its construction specifically includes the following steps:

[0076] 1. Within a consensus cycle, transactions that can change the state generated in a sub-shard are processed through consensus and packaged into blocks, and the block records are stored on the sub-blockchain;

[0077] 2. Within a consensus cycle, the main shard consensus process packages all sub-shards' recorded blocks and stores the new blocks generated by the main shard on the main blockchain.

[0078] The main blockchain and sub-blockchains are interconnected through anchor nodes of each sub-shard. That is, each anchor node stores one main blockchain and one sub-blockchain of its sub-shard.

[0079] In a sub-shard, a non-anchor node drone stores only one sub-blockchain of its sub-shard; in a main shard, an idle drone stores only one main blockchain.

[0080] Figure 3 An asynchronous consensus process for unmanned aerial vehicle (UAV) swarms is provided for embodiments of the present invention. Assuming the consensus network connection is stable in the nth round, the asynchronous consensus process specifically includes the following steps: ① Based on the task execution process, each node in the three sub-shards generates relevant transactions; ② Each of the three shards reaches a consensus internally; ③ Each of the three sub-shards generates the latest state based on the consensus result and completes the accounting in the form of blocks; ④ The drone generates a transaction at the anchor node of each of the three sub-shards. The transaction content is the latest status of each sub-shard. ⑤ Consensus is reached within the main shard; ⑥ The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of accounting is the deterministic final state. Steps ① to ③ are performed synchronously within the three sub-shards without interfering with each other, while steps ① to ③ and steps ④ to ⑥ are performed asynchronously within the sub-shards and the main shard, respectively.

[0081] Figure 4 An asynchronous consensus process for unmanned aerial vehicle (UAV) swarms is provided for embodiments of the present invention. This process includes (n+2) rounds of consensus, where network connectivity is stable, network connectivity is unstable, and the task cluster is cancelled. The asynchronous consensus process for each case is as follows:

[0082] In the first consensus cycle, the network connection is stable, and the asynchronous consensus process specifically includes the following steps: ① Based on the task execution process, each node in the three sub-shards generates relevant transactions; ② Each of the three shards reaches a consensus internally; ③ Each of the three sub-shards generates the latest state based on the consensus result and completes the accounting in the form of blocks; ④ The drone generates a transaction at the anchor node of each of the three sub-shards. The transaction content is the latest status of each sub-shard. ⑤ Consensus is reached within the main shard; ⑥ The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of accounting is the deterministic final state.

[0083] Steps ① to ③ are performed synchronously within the three sub-shards without interfering with each other, while steps ① to ③ and steps ④ to ⑥ are performed asynchronously within the sub-shards and the main shard, respectively.

[0084] During the second consensus cycle, network connectivity was unstable, and the asynchronous consensus process specifically included the following steps:

[0085] ① Based on the task execution process, each node in the three sub-shards generates relevant transactions; ② Each of the three shards reaches a consensus internally; ③ Subshard 1 has a stable network connection, generates the latest state based on the consensus result, and completes the accounting in the form of blocks; Subshard 2 has a slightly unstable network connection, but still generates the latest state based on the consensus result and completes the accounting in the form of blocks thanks to the fault tolerance capability of the consensus protocol; Subshard 3 has a very unstable network connection, and cannot complete the consensus process thanks to the fault tolerance capability of the consensus protocol. ④ The anchor nodes of sub-shard 1 and sub-shard 2 generate drone transactions, and the content of these transactions is the latest status of their respective sub-shards; ⑤ Consensus is reached within the main shard; ⑥ The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of the accounting is a temporary soft state. Steps ① to ③ are performed synchronously within the three sub-shards without interfering with each other, while steps ① to ③ and steps ④ to ⑥ are performed asynchronously within the sub-shards and the main shard, respectively.

[0086] In the third consensus cycle, network connectivity stabilizes, and the asynchronous consensus process includes the following steps: ① Based on the task execution process, each node in the three sub-shards generates relevant transactions; ② Each of the three shards reaches a consensus internally; ③ Each of the three sub-shards generates the latest state based on the consensus result and completes the accounting in the form of blocks; ④ The drone generates a transaction at the anchor node of each of the three sub-shards. The transaction content is the latest status of each sub-shard. ⑤ Consensus is reached within the main shard; ⑥ The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest accounting state is a deterministic final state that includes the temporary soft state of the previous period and the latest state of all sub-shards in the current period.

[0087] Steps ① to ③ are performed synchronously within the three sub-shards without interfering with each other, while steps ① to ③ and steps ④ to ⑥ are performed asynchronously within the sub-shards and the main shard, respectively.

[0088] In the (n+2)th consensus cycle, the network connection is stable, and the asynchronous consensus process specifically includes the following steps: ① Based on the task execution process, each node in the three sub-shards generates relevant transactions; ② Each of the three shards reaches a consensus internally; ③ Subshard 1 and Subshard 2 have stable network connections, generate the latest state based on the consensus result, and complete the accounting in the form of blocks; Subshard 3 task cluster is cancelled and will no longer update the state. ④ The anchor nodes of the two sub-shards generate transactions using drones, and the transaction content is the latest status of each sub-shard; ⑤ Consensus is reached within the main shard; ⑥ The main shard generates the latest state based on the consensus result and completes the accounting in the form of blocks. At this time, the latest state of accounting is the deterministic final state.

[0089] Steps ① to ③ are performed synchronously within the three sub-shards without interfering with each other, while steps ① to ③ and steps ④ to ⑥ are performed asynchronously within the sub-shards and the main shard, respectively.

[0090] To demonstrate the significant technical effects of this invention, numerous experiments were conducted using the ATLAB & Simulink platform and the UAV Toolbox tool.

[0091] Figure 5 The invention demonstrates the average latency for confirming the deterministic final state in a highly dynamic UAV swarm: Assuming 5000 transactions leading to state transitions are generated evenly across sub-shards, the average latency for each master shard node to confirm the deterministic final state decreases almost linearly with the increase in the number of task clusters; conversely, assuming 5000 transactions leading to state transitions are generated in each sub-shard, the average latency for each master shard node to confirm the deterministic final state remains almost constant. This demonstrates the technical advantages of the invention: 1) The average latency for confirming the deterministic final state is in the millisecond range, which is acceptable for UAV swarm applications; 2) The average latency for confirming the deterministic final state varies little with the number of task clusters, meeting the needs of large-scale UAV swarm applications.

[0092] Figure 6 and Figure 7 The results of comparing the consensus process of this invention with existing technical solutions in a static scenario are shown:

[0093] The delay in confirming the deterministic final state of this invention is significantly better than that of existing technical solutions.

[0094] Figure 6 and Figure 8 This invention demonstrates a comparison of the consensus process of drone swarms in static and dynamic scenarios:

[0095] Mobility has a significant impact on reaching a consensus on a deterministic final state. The average latency for drone swarms to confirm a deterministic final state in dynamic scenarios is significantly higher than that in static scenarios.

[0096] The embodiments described above are merely some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

Claims

1. A final consensus model for unmanned aerial vehicle (UAV) swarms, characterized in that: The model includes a hybrid blockchain storage module for the execution state of a drone swarm, consisting of a main cluster blockchain and multiple task cluster blockchains, as well as a module for transitioning between temporary soft states and deterministic final states. The task cluster blockchain is used to store the task execution status within the drone swarm, and the main cluster blockchain is used to store the final mixed task execution status that aggregates the task execution status within each task cluster blockchain. The final mixed task execution status has two states: a temporary soft state and a deterministic final state. The temporary soft state is an intermediate state that is recognized only by drone nodes within a single task cluster, while the deterministic final state is a final state that is recognized by all drone nodes in the drone swarm. The temporary soft state and deterministic final state transition module is as follows: Let the state be... Periodic task cluster The task execution status failed to be submitted to the main cluster blockchain, resulting in a temporary soft state. It appears and is stored in the task cluster blockchain until the subsequent [number]th [number]. Periodic task cluster The task execution status was successfully submitted to the main cluster blockchain; the missing temporary soft state is missing. With the Periodic task cluster The task execution status is aggregated and stored in the main cluster blockchain, at which point it is in a temporary soft state. The transition is to a deterministic final state.

2. An asynchronous consensus method for unmanned aerial vehicle (UAV) swarms, characterized in that, The specific steps are as follows: Step S1: Based on blockchain sharding technology, the drone swarm is divided into a main cluster shard (main shard) and multiple task cluster shards (sub-shards) according to the tasks. Within the task cluster shard, drone nodes establish network connections with each other and execute predetermined tasks. Within the main cluster shard, drone nodes establish network connections with each other. Step S2: Randomly select a drone node as an anchor node in each task cluster shard. The anchor node establishes a network connection with the drone nodes inside the main cluster shard. The original drone nodes inside the main cluster shard and these anchor nodes together form the main cluster shard. Step S3: Within each task cluster shard, the drone nodes execute tasks and generate blockchain transactions, which contain the task execution status of the drone node. Step S4: Within each task cluster shard, the drone nodes collectively reach a consensus on the blockchain transactions. Step S5: Based on the blockchain transaction consensus result, each task cluster shard generates the latest task execution status within the cluster and stores the task execution status within the cluster in the form of a block on the task cluster blockchain. At this time, the task execution status within the cluster is a temporary soft state. Step S6: Each task cluster shard anchor node generates a blockchain transaction, which contains the latest task execution status within the task cluster shard. Step S7: The drone nodes within the main cluster shard jointly reach a final consensus on the blockchain transactions. Step S8: Based on the blockchain transaction consensus result, the main cluster shard generates the latest task execution status within the group and stores the task execution status within the group in the form of blocks on the task cluster blockchain. At this time, the task execution status within the group is a deterministic final state.

3. The asynchronous consensus method for unmanned aerial vehicle swarms as described in claim 2, characterized in that, Set up task cluster sharding In the The cycle failed to execute step S4, i.e., task cluster sharding. The network connection of most drone nodes was interrupted, and consensus within the blockchain transaction group failed. At this time, no task execution status was generated within the group until the task cluster was sharded. In the Successfully executing step S4 of the cycle, i.e., task cluster sharding. The network connection of most drone nodes was re-established, and consensus was successfully achieved within the blockchain transaction group.

4. The asynchronous consensus method for unmanned aerial vehicle swarms as described in claim 2, characterized in that: Set up task cluster sharding In the Step S6, i.e., task cluster sharding, was not successfully executed during the cycle. If the anchor node loses its connection with the main cluster shard network, the latest task execution status within the group fails to be submitted to the main cluster blockchain. At this point, the task execution status within the group is in a temporary soft state until the task cluster shards... In the The cycle successfully executed step S6, and the task cluster was sharded. The anchor node re-establishes its connection with the main cluster shard network, and the latest task execution status within the cluster is successfully submitted to the main cluster blockchain. At this point, in step S6, the task cluster shards... The blockchain transactions generated by the anchor node aggregate the missing temporary soft state. } and task cluster sharding No. Periodic task execution status .

5. An electronic device, characterized in that, It includes a processor and a memory, the memory storing multiple instructions; the processor loads instructions from the memory to execute the model as described in claim 1 or the method as described in any one of claims 2-4.

6. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a plurality of instructions adapted for loading by a processor to perform the model as described in claim 1 or the method as described in any one of claims 2-4.

Images