Unmanned aerial vehicle flight data scheduling method based on distributed storage system

Abstract

The invention relates to the technical field of unmanned aerial vehicle flight data scheduling, and discloses an unmanned aerial vehicle flight data scheduling method based on a distributed storage system, and the method comprises the following steps: S1, an unmanned aerial vehicle node collects flight data and stores the flight data in a local memory; s2, realizing consistency and synchronous management of the data through a block chain network, and verifying the validity of the uploaded data among a plurality of nodes; and S3, realizing the consistency and synchronous management of the data through the block chain network, and verifying the validity of the uploaded data among the plurality of nodes. By introducing a block chain technology, an intelligent contract and a bandwidth optimization algorithm, the problems of consistency, redundant copy management and data transmission path optimization of the unmanned aerial vehicle flight data in a distributed storage system are effectively solved, the integrity, security and high availability of the data are ensured, the bandwidth utilization rate and transmission efficiency are improved, and the system performance is improved. Therefore, the reliability and the intelligent scheduling capability of the system are enhanced.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle flight data scheduling, and in particular to a unmanned aerial vehicle flight data scheduling method based on a distributed storage system. Background Art

[0002] With the rapid development of drone technology, more and more industries are beginning to use drones for data collection and processing. For example, in the fields of agriculture, environmental monitoring, disaster relief, logistics, etc., drones collect a large amount of flight data and environmental data in real time by carrying various sensors and shooting equipment. The amount of data is huge and has timeliness and real-time requirements. Therefore, how to efficiently and securely store, process and transmit drone data has become an important research topic.

[0003] At present, traditional drone data processing and storage usually rely on cloud platforms or local storage solutions. Although cloud platforms can provide powerful computing and storage capabilities, they also bring problems such as data transmission bandwidth and latency. At the same time, the cloud platform's dependence on data also makes the system vulnerable to factors such as network interruptions and central node failures. In addition, since drones are usually in an environment far away from the ground control center, the network connection is often unstable, resulting in frequent data loss and inconsistency problems.

[0004] On the other hand, existing distributed storage systems also have certain limitations in managing redundant copies. In order to ensure data reliability and high availability, it is often necessary to create multiple redundant copies. However, the management of redundant copies usually relies on static rules, which not only wastes storage resources, but also cannot flexibly respond to node failures or changes in storage pressure. Especially in bandwidth-constrained environments, the management of redundant copies is particularly prominent, and excessive creation and improper migration of copies will further increase the consumption of storage resources.

[0005] In addition, bandwidth and data transmission efficiency are also key issues in drone flight data management. Since drone flight missions have high real-time requirements for data transmission, especially in multi-machine collaboration scenarios, how to prioritize and select the optimal transmission path based on the network bandwidth and data importance has become the key to improving system efficiency. Most existing technologies use simple bandwidth allocation strategies or fixed transmission paths, which cannot be adaptively optimized in a dynamically changing network environment, resulting in low transmission efficiency and even affecting the execution of drone missions.

[0006] Therefore, how to efficiently manage the storage and transmission of drone flight data while ensuring data security and consistency has become an important issue that needs to be urgently solved in distributed storage systems. Summary of the invention

[0007] In view of the deficiencies of the prior art, the present invention provides a method for scheduling UAV flight data based on a distributed storage system, which solves the problems of consistency management, redundant copy optimization, and bandwidth scheduling of UAV flight data in a distributed storage system.

[0008] To achieve the above objectives, the present invention is realized through the following technical solutions: A method for scheduling UAV flight data based on a distributed storage system includes the following steps: S1. The UAV node collects flight data and stores it in the local memory; S2. Through the blockchain network, realize the consistency and synchronization management of data, and verify the validity of the uploaded data among multiple nodes; S3. Through the blockchain network, realize the consistency and synchronization management of data, and verify the validity of the uploaded data among multiple nodes. Among them, data synchronization is carried out through the decentralized consensus mechanism in the blockchain network, and a multi-level consensus protocol is adopted to ensure the consistency of data among multiple nodes; S4. Dynamically manage the redundant copies of data based on smart contracts, and manage the creation, migration, and deletion of copies; S5. Use a bandwidth optimization algorithm to adjust the data transmission path, and perform priority scheduling according to network bandwidth and data importance; S6. When a node fails, recover the data through redundant copies.

[0009] Preferably, the priority of the data in step S2 is calculated according to the following formula:

[0010] where is the priority of the data , and α, β, γ, and δ are the weight coefficients of data urgency, network bandwidth, data size, and upload delay respectively.

[0011] Preferably, the multi-level consensus protocol in step S3 includes: Single-node level, used for each node to verify the validity of the uploaded data through local consensus; Cross-node level, used for data consistency verification through smart contracts in the blockchain network; Global level, used to ensure data consistency among different UAV groups through cross-chain protocols.

[0012] Preferably, the bandwidth optimization algorithm in step S5 adopts a reinforcement learning model, and optimizes the data transmission path through the following objective function:

[0013] where is the total amount of data to be uploaded, is the current available bandwidth, R is the current network transmission rate, and L is the size of the current data packet.

[0014] Preferably, the S3 step verifies the validity of the uploaded data among multiple nodes and adopts cross-chain synchronization technology in the data transmission process. The cross-chain synchronization ensures data consistency between different drone clusters by the following methods: Data synchronization is performed through the local blockchain network within each drone cluster; Use cross-chain protocols to synchronize data between different clusters, and ensure data consistency through synchronized timestamps and on-chain information; If cross-chain data inconsistency occurs, the cross-chain synchronization strategy is adjusted through the smart contract to automatically select the optimal data synchronization path.

[0015] Preferably, the storage rules of the redundant copies in step S4 are adjusted based on the following parameters: Data importance, which is used to dynamically adjust the number of replicas based on the importance of the data, creating more replicas for important data; Node bandwidth and storage capacity, which are used for smart contracts to automatically create copies when node bandwidth and storage space meet certain conditions; Node health status, which is used to determine the migration and deletion strategies of replicas based on the node health status.

[0016] Preferably, in the S3 step, the blockchain network adopts zero-knowledge proof technology and distributed ledger technology.

[0017] Preferably, when the smart contract manages redundant copies of data in step S4, the following mechanism is further used to control the creation and migration of copies: The creation of replicas is based on the historical upload frequency of the data. When the upload frequency of a certain data is high, the number of replicas created for the data will be automatically increased; The migration of replicas is determined by the health of the node and the storage pressure. When the node health is poor or the storage space is close to the upper limit, the smart contract automatically triggers the migration or deletion of replicas. The copy deletion mechanism is based on data expiration time and node load. When the data is not accessed within the specified time and the node storage resources are tight, the smart contract will automatically delete the copy.

[0018] The present invention provides a method for scheduling UAV flight data based on a distributed storage system. It has the following beneficial effects: 1. The present invention solves the problems of data consistency and security in a distributed storage system by introducing blockchain technology and a decentralized consensus mechanism. When multiple drone nodes work collaboratively, data is verified and synchronized through a blockchain network to ensure data consistency across all nodes. The immutability and distributed ledger characteristics of the blockchain guarantee the integrity and credibility of the uploaded data, effectively avoiding potential data loss, tampering, and consistency issues in traditional distributed systems. Through this method, the system can operate without relying on a centralized server, enhancing data security and improving the reliability during drone operations.

[0019] 2. The present invention solves the efficiency problem of redundant data management by dynamically managing the creation, migration, and deletion of redundant copies through smart contracts. The system can intelligently determine whether to create, migrate, or delete redundant copies based on factors such as the importance of the data, the bandwidth, storage space, and health status of the nodes. This mechanism significantly reduces unnecessary redundant storage, improves the utilization rate of storage resources, and can quickly recover data through redundant copies in case of node failures, ensuring high availability and fast recovery capabilities of drone flight data.

[0020] 3. The present invention combines a bandwidth optimization algorithm and a reinforcement learning model to solve the intelligent scheduling problem of data transmission path selection. By dynamically adjusting the upload path and performing priority scheduling based on multiple factors such as the importance of the data, network bandwidth, and upload latency, the system can achieve the optimal data transmission path selection, thereby improving bandwidth utilization and transmission efficiency. This method not only reduces waste of network bandwidth but also ensures that high-priority data can be uploaded in a timely manner, avoiding latency issues caused by network bottlenecks during data transmission. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 is a flowchart of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0022] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0023] Please refer to the attached Figure 1 , the embodiments of the present invention provide a method for scheduling drone flight data based on a distributed storage system, including the following steps: S1. Drone nodes collect flight data and store it in the local memory; Specifically, in the UAV flight data scheduling method based on a distributed storage system, it is first necessary to ensure that UAV nodes can collect flight data in real time and effectively store it in the local memory. The types of flight data include, but are not limited to, flight trajectories, sensor data, environmental monitoring data, etc. These data are crucial for the autonomous flight of UAVs and the collaborative operation of multiple UAVs. Through local storage, the stable preservation of data can be ensured during the flight of UAVs, facilitating subsequent processing and uploading.

[0024] In this embodiment, the UAV node continuously collects various data during the flight through its built-in sensors. Specifically, various sensors of the UAV (such as GPS sensors, environmental sensors, inertial measurement units, etc.) will continuously monitor and record the flight state of the UAV and the changes in the surrounding environment. The types of flight data include, but are not limited to, information such as flight altitude, speed, position coordinates, heading angle, environmental temperature, air pressure, humidity, wind speed, etc. These data play an important role in subsequent flight decision-making and the scheduling of collaborative operation of multiple UAVs.

[0025] To ensure the timeliness and stability of data storage, in this embodiment, the data collection and storage process relies on the local memory (such as solid-state drives, memory cards, etc.) in the UAV for preliminary preservation. The data will be locally cached according to the collection frequency and data volume requirements. To avoid data loss or storage space overload, the local memory will perform data backup and cleaning operations regularly or according to certain conditions based on the importance of the data and the storage strategy.

[0026] For example, within each time period, the collected flight data will be stored in the form of data packets. Each data packet includes the specific measurement values of sensor data, the collection timestamp, and information on the current flight state. The size of the data packet can be calculated by the formula:

[0027] where represents the data size of sensor i, represents the collection time interval of each sensor data, and n is the number of all sensors.

[0028] In addition, during the process of collecting data, to address possible storage space limitations and fault recovery requirements, the management strategies of the local memory include, but are not limited to, data compression, data indexing, and redundant storage. For example, when the flight data collected within a certain time period reaches the storage threshold, the system will automatically perform data compression or upload some historical data to the edge node or blockchain network to avoid storage overflow. The data compression technology uses lossless compression algorithms to ensure the integrity of the data and subsequent use.

[0029] The design of this embodiment not only meets the basic requirements for real-time data acquisition and storage during the flight of the drone, but also provides a stable and reliable data source for subsequent data uploading and redundant copy management. Through local storage, the system ensures that data will not be lost in case of network interruption or other temporary failures, thereby improving the stability and reliability of the entire distributed storage system.

[0030] S2. Implement the consistency and synchronization management of data through the blockchain network, and verify the validity of the uploaded data among multiple nodes; The priority of the data in step S2 Calculate according to the following formula:

[0031] Where is the priority of the data , and α, β, γ, and δ are the weight coefficients of data urgency, network bandwidth, data size, and upload delay respectively.

[0032] Specifically, in the drone flight data scheduling method of the present invention, the main task of step S2 is to implement the consistency and synchronization management of data through the blockchain network. Since drone nodes are often distributed in a wide area and the storage and computing capabilities of each node may vary, an efficient and trustworthy mechanism must be adopted to verify and synchronize the uploaded data to ensure the accuracy and consistency of data among multiple drones. As a decentralized technology, the blockchain can effectively solve this problem. Through the blockchain network, the uploaded data can not only be verified by multiple nodes, but also ensure the immutability of the data, enhancing the reliability and security of data transmission.

[0033] In addition, the calculation of data priority plays a key role in this step. To ensure the efficiency of data transmission and the rational utilization of network resources, it is necessary to determine the order and timing of data uploading according to the priority of the data. The priority of the data is not only related to the urgency of the data, but also closely related to factors such as network bandwidth, data size, and upload delay. Through precise priority calculation, bandwidth usage can be optimized, network congestion can be avoided, and key data can be uploaded first.

[0034] In this embodiment, after the drone node collects data, the system will first calculate the priority of each piece of data according to factors such as data urgency, network bandwidth, data size, and upload delay.

[0035] The calculation formula for data priority is as follows:

[0036] Where: P represents the priority of the data; α, β, γ, δ are the weight coefficients of data urgency, network bandwidth, data size, and upload delay respectively, and these coefficients can be dynamically adjusted according to actual requirements; E is the data urgency, reflecting whether the data is real-time or critical data; B is the network bandwidth, indicating the available bandwidth of the current network. The larger the bandwidth, the relatively lower the priority of data upload; S is the data size. Larger data requires more bandwidth, so its priority is lower; D is the upload delay. The higher the delay, the higher the priority of data upload should be accordingly.

[0037] The priority value calculated according to this formula can be used to guide the data upload strategy. For data with high urgency and low upload delay, its priority is high and it should be uploaded as soon as possible to ensure that critical data can reach the target node in time. For those files with low urgency but large data volume, the system will make reasonable scheduling according to the bandwidth resources to avoid wasting bandwidth resources and transmission delay.

[0038] In the process of data upload, the blockchain network plays an important role. It verifies and synchronizes the uploaded data through a decentralized consensus mechanism and smart contracts. The upload of each data block requires multiple nodes to jointly verify to ensure the integrity and validity of the data. Specifically, after the data is uploaded to the blockchain network, its validity is first verified through the local consensus mechanism. Each node conducts a preliminary verification based on the timestamp, data source, and content in the data packet. Then, the cross-node level consensus protocol verifies the consistency of the data through smart contracts to ensure the consistency of the same data among different nodes and avoid the risks of data tampering and loss.

[0039] The decentralized feature of the blockchain enables multiple drone nodes to jointly maintain and update data records relying on the consensus mechanism without trusting a single central node, which greatly enhances the reliability and security of data storage.

[0040] Data Priority Calculation and Network Bandwidth Optimization: When implementing priority scheduling, the weight coefficients in the priority calculation formula can be dynamically adjusted according to the network environment, drone task type, and data type. For example, when the task requirements are relatively urgent (such as real-time processing of data when an obstacle appears during flight), α (data urgency) can be set to a higher value, while for monitoring data with low real-time requirements, its priority can be appropriately reduced and the corresponding weight coefficient can be adjusted. In addition, the usage of network bandwidth (B) is also dynamically monitored and acts together with the data size (S) and upload delay (D) to determine the final data transmission order. The dynamic allocation and optimization of network bandwidth ensure that the entire system can better handle different bandwidth conditions and data priority requirements during data upload, thereby maximizing the utilization rate of bandwidth and the stability of the system.

[0041] S3. Implement the consistency and synchronization management of data through the blockchain network, and verify the validity of the uploaded data among multiple nodes. Among them, data synchronization is carried out through the decentralized consensus mechanism in the blockchain network, and a multi-level consensus protocol is adopted to ensure the consistency of data among multiple nodes; The multi-level consensus protocol in step S3 includes: Single-node level, used for each node to verify the validity of the uploaded data through local consensus; Cross-node level, used for data consistency verification through smart contracts in the blockchain network; Global level, used to ensure data consistency among different drone groups through cross-chain protocols.

[0042] In step S3, when verifying the validity of the uploaded data among multiple nodes, cross-chain synchronization technology is adopted during the data transmission process. Cross-chain synchronization ensures data consistency between different drone clusters through the following methods: Perform data synchronization within each drone cluster through the local blockchain network; Use cross-chain protocols to synchronize data between different clusters, and ensure data consistency through synchronized timestamps and on-chain information; If cross-chain data inconsistency occurs, adjust the cross-chain synchronization strategy through smart contracts and automatically select the optimal data synchronization path.

[0043] In step S3, the blockchain network adopts zero-knowledge proof technology and distributed ledger technology.

[0044] Specifically, in the drone flight data scheduling method of the present invention, step S3 focuses on solving the problem of how to effectively ensure data consistency, synchronization and validity in a distributed environment. Due to the differences in physical location, computing power, network bandwidth, etc. among multiple drone nodes, how to ensure the consistency of the uploaded data among all nodes and prevent data from being tampered with or lost has become a core challenge in system design. To achieve this goal, this step adopts the blockchain network and its decentralized consensus mechanism to ensure data validity and consistency. Especially in the scenario of multi-drone group collaboration, blockchain technology can provide reliable decentralized verification to ensure data synchronization and consistency among multiple nodes.

[0045] The advantages of the blockchain lie in the immutability, transparency and decentralization characteristics of its data structure. Using these characteristics can ensure that all uploaded data is consistently verified among multiple nodes and eliminate the risk of single-point failure. To further enhance the guarantee of data consistency, this step introduces a multi-level consensus protocol. Through multiple levels of consensus verification mechanisms, the process of data verification and synchronization becomes more efficient and reliable.

[0046] In this embodiment, The specific implementation of step S3 includes the following key elements: 1. Single-node level: At this level, each drone node verifies the validity of the uploaded data through a local consensus mechanism. When each node receives data, it will first conduct a preliminary verification according to the local consensus rules to ensure the legality of the data source and the integrity of the data itself. Each node independently verifies the data it receives, including verifying the timestamp, source node, and integrity of the data, to avoid tampering or loss.

[0047] 2. Cross-node level: Further data consistency verification is carried out through smart contracts in the blockchain network. After the data is uploaded, the smart contract will verify the data consistency among multiple nodes to ensure that the data stored on all nodes is consistent in terms of time and content. This process can trigger data verification through the rules set by the smart contract. For example, when a node uploads data, the smart contract will automatically verify whether the same or matching data exists on other relevant nodes. If there are differences, it will trigger a further data synchronization or error handling mechanism.

[0048] 3. Global level: In the collaborative scenario across drone groups, a cross-chain protocol is used to ensure data consistency between different drone clusters. The data within each cluster is synchronized and verified by the local blockchain network, but the data between different clusters needs to be synchronized through the cross-chain protocol. The cross-chain protocol will ensure data consistency between different blockchains, guaranteeing error-free data synchronization across regions and nodes. Through the cross-chain protocol, the data between different clusters can be synchronized according to the predetermined timestamp, ensuring that the collaboration between multiple drone groups is not affected by data consistency issues.

[0049] 4. Cross-chain synchronization technology: In the scenario of multi-drone cluster collaboration, the synchronization of data is not limited to the blockchain network of a single cluster. To ensure data consistency between different drone clusters, cross-chain synchronization technology is used in this embodiment. Specifically, the data within each drone cluster is synchronized through the local blockchain network, while the cross-cluster data synchronization is achieved through the cross-chain protocol. Cross-chain synchronization uses timestamps and on-chain information as the basis for synchronization to ensure that the data between different clusters is consistent in time during the data transmission process. When data inconsistency occurs, the smart contract can automatically adjust the cross-chain synchronization strategy according to the preset rules, select the optimal data synchronization path, and achieve data consistency recovery in the shortest time.

[0050] 5. Zero - Knowledge Proof and Distributed Ledger: During the data transmission process, to ensure the privacy and security of data, zero - knowledge proof technology and distributed ledger technology are adopted in this embodiment. Zero - knowledge proof technology enables the verification of the authenticity and validity of data during the upload process without directly exposing its content. In this way, data can be verified in the blockchain network without disclosing detailed information, greatly enhancing the ability to protect data privacy. At the same time, distributed ledger technology ensures the transparency and immutability of data records. The upload and verification of each piece of data can be shared and traced by all network nodes, ensuring data consistency throughout the network.

[0051] Specific process of data synchronization: The process of cross - chain synchronization includes the following steps: Local data upload: After each drone node collects data, it first uploads the data to the local blockchain network to ensure data synchronization and verification at this node.

[0052] Cross - chain protocol trigger: When data needs to be synchronized across clusters, the cross - chain protocol is triggered to transfer the data to the blockchain network of other clusters.

[0053] Data consistency check: Through the synchronized timestamp and on - chain information, the cross - chain protocol ensures data consistency. Each node verifies whether the data matches according to the timestamp. If there are differences, the smart contract will issue a warning or perform a recovery operation.

[0054] Smart contract adjusts the synchronization path: If data inconsistency occurs, the smart contract automatically selects the optimal data synchronization path according to the set rules and resynchronizes the data through the network to ensure final consistency.

[0055] S4. Dynamically manage redundant copies of data based on smart contracts, and manage the creation, migration, and deletion of copies; The storage rules of redundant copies in step S4 are adjusted based on the following parameters: Data importance, which is used to dynamically adjust the number of copies according to the importance of data. More copies are created for important data; Node bandwidth and storage capacity, which are used to automatically create copies by the smart contract when the node bandwidth and storage space meet certain conditions; Node health status, which is used to determine the migration and deletion strategies of copies based on the health status of the node.

[0056] When the smart contract manages redundant copies of data in step S4, the following mechanisms are further adopted to control the creation and migration of copies: The creation of copies is based on the historical upload frequency of data. When the upload frequency of a certain piece of data is higher than the set frequency, the number of copies created for this data is automatically increased; The migration of replicas is jointly determined by node health and storage pressure. When the node health is poor or the storage space is approaching the upper limit, the smart contract automatically triggers the migration or deletion of replicas. The replica deletion mechanism is based on the data expiration time and node load conditions. When the data has not been accessed within the specified time and the node storage resources are tight, the smart contract will automatically delete the replica.

[0057] Specifically, in the drone flight data scheduling method of the present invention, the core objective of step S4 is to automate the management of redundant replicas of data through a smart contract, ensuring high availability, reliability, and reasonable allocation of storage resources of data in a distributed storage system. Redundant replicas are a key measure to ensure data backup and recovery in the system. Especially in the case of a large number of drone nodes and dispersed storage resources, the management of redundant replicas is crucial for ensuring data integrity and consistency.

[0058] To achieve efficient management of redundant replicas, this step introduces a smart contract automated control mechanism. By setting a series of rules and conditions, the smart contract dynamically adjusts the creation, migration, and deletion operations of replicas according to data importance, network conditions, node health, and storage requirements. This can effectively avoid excessive creation of redundant replicas and storage waste, while ensuring high availability of critical data. Especially during the flight of a drone cluster, it can make a quick response according to real-time network status and node conditions.

[0059] In this embodiment, the specific implementation of step S4 includes the following key mechanisms: 1. Creation rules for redundant replicas: The creation of redundant replicas is dynamically adjusted according to the following parameters: Data importance: Regarding the importance of data, the smart contract will dynamically adjust the number of replicas according to the priority of the data. Important data or data that needs to be accessed frequently will create more replicas to ensure its quick recovery in case of node failure or network problems.

[0060] Historical upload frequency: The creation of replicas is also related to the upload frequency of data. When the upload frequency of a certain data is higher than the set frequency, the smart contract will automatically increase the number of replicas created for this data to improve the accessibility of the data between different nodes and avoid data loss caused by frequent uploads. The specific value of the set frequency can be determined in a certain way.

[0061] Node bandwidth and storage capacity: When the bandwidth or storage space of a certain node is sufficient, the smart contract will decide whether to create replicas according to the remaining bandwidth and storage capacity of the node. If the node has sufficient storage resources, the smart contract will choose to store the replica on this node to disperse the storage load and improve data redundancy.

[0062] 2. Migration mechanism for redundant copies: The migration of copies is jointly determined by the following conditions: Node health status: When the health status of a certain node deteriorates (such as a decrease in processing capacity or poor network stability), the smart contract will automatically trigger the migration of copies, migrating the data copies on this node to other healthy nodes to ensure data security and high availability. The assessment of node health status can be achieved by monitoring indicators such as the computing load, network latency, and data loss rate of the node. For example, a node health score is performed, and the change in the score can reflect the change in the node health status.

[0063] Storage pressure: When the storage space of a node is approaching the upper limit, the smart contract will evaluate the redundant copies stored on the current node and select to migrate some copies to other nodes to free up storage space and avoid node overload. By monitoring the remaining available storage space, it can be determined whether it is approaching the storage upper limit. The available storage space can refer to the size of the storage space that can be utilized without affecting the storage performance.

[0064] Copy migration trigger condition: Once it is found that the node health status is poor or the storage capacity is approaching the upper limit, the smart contract will automatically select the copies to be migrated through preset strategies (such as the priority of the copies, the usage frequency of the data, etc.) and select the target node for storage.

[0065] 3. Deletion rules for redundant copies: The deletion of copies is controlled by the following factors: Data expiration time: Each copy will have an expiration time during the storage process, and this time is dynamically set according to the importance and usage frequency of the data. When a certain copy has not been accessed for a long time and exceeds the set expiration time, the smart contract will automatically delete this copy to free up storage space.

[0066] Node load situation: When the storage pressure of a node increases (such as when storage resources are insufficient), the smart contract will evaluate the access frequency and importance of all redundant copies on the node and, if necessary, delete the copies with lower priority, giving priority to retaining the copies of important data.

[0067] Copy deletion trigger condition: The trigger condition for copy deletion can be controlled by the preset rules of the smart contract. For example, when the data has not been accessed within the specified time or the node storage space is too tense, the copy deletion mechanism will dynamically adjust the deletion strategy according to the priority.

[0068] 4. Management strategies for smart contracts: The role of smart contracts in redundant replica management is crucial. Smart contracts are not only used to automatically create, migrate, and delete replicas, but also can adjust the storage strategy of replicas according to the real-time changes of the network and data requirements. Through an automated decision-making mechanism, smart contracts can efficiently allocate resources, avoid the abuse of redundant replicas, while ensuring the redundant storage of critical data across multiple nodes and avoiding single points of failure.

[0069] 5. Lifecycle Management of Redundant Replicas: The management of redundant replicas is actually a lifecycle management process. From creation, replica migration, update, to deletion, every link requires the intelligent control of smart contracts. In this way, the management of redundant replicas can not only improve the reliability and availability of data, but also intelligently optimize resource allocation in the case of limited storage resources, thereby improving the efficiency of the entire distributed storage system.

[0070] Through the above steps and mechanisms, step S4 realizes the intelligent and dynamic management of redundant replicas. By using the automated control of smart contracts for replica creation, migration, and deletion, the need for manual intervention is greatly reduced, improving the stability and scalability of the system. At the same time, the creation and migration of replicas are dynamically adjusted based on the importance of the data and the status of the nodes, ensuring that the data can be quickly restored in case of a failure, while the deletion mechanism of redundant replicas avoids waste of storage resources. Ultimately, this automated and intelligent replica management method not only improves the availability of data and the fault tolerance of the system, but also optimizes the utilization rate of storage resources, meeting the requirements of efficient data management in a distributed storage environment.

[0071] S5. Use a bandwidth optimization algorithm to adjust the data transmission path and perform priority scheduling according to network bandwidth and data importance; In step S5, the bandwidth optimization algorithm uses a reinforcement learning model to optimize the data transmission path through the following objective function:

[0072] Where, is the total amount of data to be uploaded, is the current available bandwidth, R is the current network transmission rate, and L is the size of the current data packet.

[0073] Specifically, in the UAV flight data scheduling method of the present invention, step S5 mainly solves the problem of how to efficiently utilize network bandwidth resources to optimize the data transmission path in a distributed storage environment and reasonably schedule the data upload process according to the priority of the data. During the transmission of UAV flight data, due to limited bandwidth resources and complex network environments (such as bandwidth differences between nodes, network congestion, etc.), how to intelligently adjust the data transmission path and perform dynamic scheduling according to the priority of the data and network conditions is the key to improving data transmission efficiency and ensuring the priority upload of critical data.

[0074] To this end, step S5 adopts a reinforcement learning model for bandwidth optimization. Reinforcement learning can automatically learn and optimize the data transmission path according to the network conditions and data upload priorities, so that when the network bandwidth is tight, high-priority data is preferentially transmitted, while ensuring the reasonable allocation of network resources and avoiding bandwidth waste.

[0075] In this embodiment, the specific implementation of step S5 includes the following aspects: 1. Bandwidth optimization algorithm: The bandwidth optimization algorithm is based on a reinforcement learning model. Through multiple interactions with the environment, the model can gradually learn the optimal transmission path and scheduling strategy. The reinforcement learning model optimizes the data transmission path through the following objective function:

[0076] Where, is the total amount of data to be uploaded, is the current available bandwidth, R is the current network transmission rate, and L is the size of the current data packet.

[0077] The meaning of the objective function is: optimize the transmission rate of each data packet under the current bandwidth conditions, ensure that the upload process of each data packet can be carried out under the optimal bandwidth resources, reduce transmission delay, avoid network congestion, and improve transmission efficiency.

[0078] 2. Application of the reinforcement learning model: The agent in the reinforcement learning algorithm learns how to dynamically adjust the data transmission path according to the real-time bandwidth of the network, the size of the data packet, and the priority of the data through multiple interactions with the environment. The reinforcement learning process includes the following steps: State space: The state space represents various factors in the current network environment, such as network bandwidth, data packet size, data priority, etc. By perceiving these factors, the reinforcement learning model can obtain the current state information.

[0079] Action Space: The action space represents all the ways of adjusting the transmission path that the model can choose. Each action corresponds to a choice of transmission path or an adjustment of the bandwidth. For example, adjusting the order of data packet transmission or selecting a path with a higher network bandwidth.

[0080] Reward Function: The reward function is used to measure the effect of each action. In the bandwidth optimization problem, the reward function can be designed based on the following factors: If the network bandwidth is effectively utilized and high-priority data is transmitted first, a positive reward is given.

[0081] If the transmission delay is high or the bandwidth is wasted, a negative reward is given.

[0082] Through such a mechanism, the agent can learn the optimal transmission strategy, so that in a complex network environment, it can intelligently adjust the transmission path and optimize the use of bandwidth resources.

[0083] 3. Data Priority Scheduling: In step S5, in addition to bandwidth optimization, data priority scheduling is also a very important factor. According to the priority of each data packet, the reinforcement learning model will automatically adjust the upload order of the data during transmission. The specific scheduling strategy is as follows: High-priority data is uploaded first: For flight data with higher urgency (such as real-time control instructions or critical status information), the reinforcement learning model will schedule the transmission of this data first when the bandwidth is low, avoiding delays in important data.

[0084] Low-priority data can be appropriately delayed: For relatively ordinary or non-urgent data (such as flight logs or analysis data), if the network bandwidth is tight, the upload of this data can be appropriately delayed until the network conditions improve or the bandwidth resources are released.

[0085] 4. Dynamic Bandwidth Adjustment Mechanism: Since the network conditions between UAV nodes may fluctuate (such as bandwidth changes or network congestion), step S5 also includes a dynamic bandwidth adjustment mechanism that can adjust the data transmission strategy according to the real-time bandwidth conditions of the network. Specifically, it includes: Monitor the changes in network bandwidth and dynamically adjust the data upload path according to the bandwidth changes.

[0086] When the network bandwidth is insufficient, select an appropriate transmission path through the reinforcement learning model to avoid bottlenecks in data transmission.

[0087] If the current bandwidth is not sufficient to support the transmission of all data, give priority to transmitting high-priority data packets, and low-priority data can be delayed for transmission to ensure the real-time nature of critical mission data.

[0088] 5. Multi-node collaboration optimization: In a multi-node collaboration environment, the bandwidth optimization algorithm also needs to consider the collaboration issues between different nodes. For example, when the bandwidth of a certain node reaches the upper limit, other nodes can temporarily provide bandwidth support, thereby optimizing the data transmission efficiency of the entire system. Through the cross-node collaboration mechanism, the reinforcement learning model can automatically adjust the data upload path, allocate tasks to nodes with relatively idle bandwidth resources, and avoid the transmission bottleneck of the system.

[0089] Through the implementation of the above bandwidth optimization algorithm, step S5 can effectively improve the transmission efficiency of the entire UAV flight data and optimize the utilization of bandwidth resources. The application of the reinforcement learning model enables the data transmission path to dynamically adapt to the changes in the network environment, intelligently schedules high-priority data, ensures the real-time nature of key tasks, and avoids bandwidth waste. In addition, the scheduling mechanism based on data priority also ensures the reasonable upload order of different types of data. When the bandwidth is tight, it can ensure that the most important data is transmitted first, thereby improving the overall performance and reliability of the system.

[0090] S6. When a node fails, recover data through redundant copies.

[0091] Specifically, in a distributed storage system, node failure is one of the common challenges. Especially during the flight of a UAV cluster, the failure of any single node may lead to data loss or inaccessibility. Therefore, how to ensure the reliability and persistence of data has become one of the key issues. The present invention solves the problem of data recovery in the case of node failure through redundant copy technology. The design of redundant copies ensures that even if some nodes fail, the data can still be recovered through the copies of other nodes, thus avoiding data loss or system downtime caused by failures.

[0092] In this embodiment, the specific implementation of step S6 includes the following aspects: 1. Storage and management of redundant copies: In the present invention, the redundant copies of data are dynamically managed through smart contracts. The creation, migration, and deletion of copies are adjusted according to multiple factors (such as data importance, node bandwidth, storage capacity, and node health status). To ensure high reliability, redundant copies of all data are created on multiple nodes. Specifically, each data copy is stored on different UAV nodes, avoiding the risk of data loss caused by the failure of a single node.

[0093] The creation of redundant copies does not solely rely on fixed policies. Instead, smart contracts adjust the number and location of copies based on real-time network conditions and the status of nodes. For example, when a node has limited storage space or poor health, the smart contract automatically migrates data copies to healthy nodes, thereby reducing the risk of single-point failures.

[0094] 2. Fault Detection and Replica Recovery: When a UAV node fails, the system needs to quickly detect the fault and initiate the data recovery mechanism. The specific operation process is as follows: Node Health Monitoring: Each node reports its own health status through periodic self-checks and health monitoring mechanisms. When a node fails or cannot work properly, the node automatically sends a fault signal to the system.

[0095] Fault Detection: The system monitors the health status of each node through a centralized monitoring system or a distributed protocol. Once a fault is detected in a certain node, the system triggers an automatic recovery mechanism and starts the process of recovering data from redundant copies.

[0096] Data Recovery Mechanism: When the system confirms a node failure, it recovers data from redundant copies in other healthy nodes. The specific steps of the recovery process are as follows: Selecting Replica Sources: The system selects healthy nodes that store redundant copies as data sources for data recovery operations. The strategy for selecting replica sources is usually based on factors such as replica synchronization status, network bandwidth, and node health status.

[0097] Replica Synchronization and Recovery: Through the blockchain network or other synchronization mechanisms, the data of the recovery source node is synchronized to the required location of the faulty node.

[0098] Notification after Data Recovery Completion: After the data recovery is completed, the system sends a notification of recovery completion to all relevant nodes in the UAV cluster to ensure that other nodes can synchronize and update the data status and resume normal operation.

[0099] 3. Replica Update and Consistency Maintenance: During the data recovery process, in order to maintain the consistency of redundant copies, the system performs replica updates and consistency maintenance according to the following mechanisms: Replica Synchronization Protocol: Through the decentralized consensus mechanism of the blockchain, ensure data consistency among various replicas. For example, use smart contracts to ensure that after data recovery, the replica versions on different nodes are consistent, avoiding data inconsistency problems.

[0100] Cross-chain Synchronization Mechanism: Among different drone clusters, a cross-chain protocol is adopted to synchronize replica data, ensuring that all nodes can obtain the same version of the replica after data recovery, and avoiding data loss or version inconsistency caused by network latency or synchronization differences.

[0101] 4. High-availability Design for Redundant Replica Recovery: To further improve the fault tolerance of the system, the design of redundant replicas follows the principle of "high availability". Specifically, it includes: Multi-node Redundancy: The replicas of each important piece of data are stored on multiple drone nodes to avoid data loss caused by a single node failure.

[0102] Geographical Redundancy: The replicas are distributed in different geographical locations to avoid the impact of regional failures (such as bad weather, network outages, etc.) on data availability.

[0103] Dynamic Replica Migration by Smart Contracts: When the storage space or bandwidth of a certain node is tight, the smart contract will automatically trigger the migration of replicas to ensure that the replicas are distributed on the optimal nodes, thus guaranteeing the fast recovery and access of data.

[0104] 5. Performance Optimization for Fault Recovery: The performance of redundant replica recovery depends on the following factors: Recovery Speed Optimization: After a node failure occurs, the system will select the best recovery path and replica source based on factors such as network bandwidth, node health status, and synchronization status of data replicas, ensuring that the fault recovery can be completed as soon as possible and minimizing the system downtime caused by node failures.

[0105] Bandwidth Resource Management: Through a bandwidth optimization algorithm, bandwidth resources are rationally allocated during the fault recovery process to ensure that the recovery process does not affect normal data transmission and avoid network overload.

[0106] 6. Deletion Mechanism for Redundant Replicas: When the node failure is repaired and the recovery is completed, subsequent processing is also required for the management of redundant replicas. Specifically: Replica Cleaning Mechanism: To avoid waste of storage resources caused by excessive redundant replicas, the smart contract will regularly check the data replica situation of each node and delete the redundant replicas that are no longer needed.

[0107] Replica Retention Mechanism: If some replicas are still in an important state (such as high-priority data or critical task data), the smart contract will still retain these replicas even after the node recovers to ensure the persistence and high availability of their data.

[0108] Through the redundant copy recovery mechanism in step S6, the present invention can effectively improve the data reliability and availability of the UAV system in the case of node failures. When a node fails, the redundant copies can ensure that data is not lost. At the same time, by using smart contracts to dynamically manage the copies, the storage and migration strategies of the copies can be flexibly adjusted according to the network environment and node conditions, ensuring the efficiency and consistency of data recovery. This mechanism effectively improves the fault tolerance of the system and the reliability of data storage, reduces the risk of service interruption caused by node failures, and enhances the stability and durability of the UAV system in complex flight environments.

[0109] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A UAV flight data scheduling method based on a distributed storage system, characterized in that: The following steps are involved: S1, the UAV node collects flight data and stores it in the local memory; S2. Realize data consistency and synchronization management through the blockchain network and verify the validity of uploaded data among multiple nodes; S3. Realize data consistency and synchronization management through the blockchain network, and verify the validity of uploaded data among multiple nodes. Data synchronization is carried out through the decentralized consensus mechanism in the blockchain network, and a multi-level consensus protocol is used to ensure the consistency of data among multiple nodes. S4. Dynamically manage redundant copies of data based on smart contracts, and manage the creation, migration and deletion of copies; S5. Use bandwidth optimization algorithm to adjust the data transmission path and perform priority scheduling based on network bandwidth and data importance; S6. When a node fails, data is restored through redundant copies.

2. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: The priority of data in step S2 Calculated according to the following formula: ； in, For data The priority of , α, β, γ, and δ are the weight coefficients of data urgency, network bandwidth, data size, and upload delay, respectively.

3. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: The S3 steps of the multi-level consensus protocol include: Single-node level, where each node verifies the validity of uploaded data through local consensus; Cross-node level, used to verify data consistency through smart contracts in the blockchain network; The global level is used to ensure data consistency between different drone groups through cross-chain protocols.

4. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: The bandwidth optimization algorithm in step S5 adopts a reinforcement learning model to optimize the data transmission path through the following objective function: ； in, is the total amount of data to be uploaded, is the current available bandwidth, R is the current network transmission rate, and L is the size of the current data packet.

5. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: The S3 step verifies the validity of the uploaded data between multiple nodes. The cross-chain synchronization technology is used in the data transmission process. The cross-chain synchronization ensures the data consistency between different drone clusters through the following methods: Data synchronization is performed through the local blockchain network within each drone cluster; Use cross-chain protocols to synchronize data between different clusters, and ensure data consistency through synchronized timestamps and on-chain information; If cross-chain data inconsistency occurs, the cross-chain synchronization strategy is adjusted through the smart contract to automatically select the optimal data synchronization path.

6. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: The storage rules of the redundant copies in step S4 are adjusted based on the following parameters: Data importance, which is used to dynamically adjust the number of replicas based on the importance of the data, creating more replicas for important data; Node bandwidth and storage capacity, which are used for smart contracts to automatically create copies when node bandwidth and storage space meet certain conditions; Node health status, which is used to determine the migration and deletion strategies of replicas based on the node health status.

7. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: In the S3 step, the blockchain network adopts zero-knowledge proof technology and distributed ledger technology.

8. The method for scheduling UAV flight data based on a distributed storage system according to claim 1, characterized in that: When the smart contract manages redundant copies of data in step S4, the following mechanism is further used to control the creation and migration of copies: The creation of copies is based on the historical upload frequency of the data. When the upload frequency of a certain data is high, the number of copies of the data created is automatically increased; when the upload frequency of a certain data is higher than the set frequency, it is determined that the upload frequency is high; The migration of replicas is determined by the health of the node and the storage pressure. When the node health is poor or the storage space is close to the upper limit, the smart contract automatically triggers the migration or deletion of replicas. When the node health score is lower than the set score, the node health is determined to be poor; when the remaining storage space is lower than the set storage capacity, the storage space is determined to be close to the upper limit; The copy deletion mechanism is based on data expiration time and node load. When the data is not accessed within the specified time and the node storage resources are tight, the smart contract will automatically delete the copy.

Images