Unmanned aerial vehicle group swarding communication control method and system based on link life perception

Abstract

The invention provides an unmanned aerial vehicle group swarding communication control method and system based on link life perception. The method comprises the following steps: S1, constructing an unmanned aerial vehicle ad hoc network; setting a state field for each unmanned aerial vehicle node; setting a task field for each task; calculating the life weighted link cost between any two unmanned aerial vehicle nodes based on the unmanned aerial vehicle self-organizing network; a multi-task preemptive scheduling strategy is adopted, tasks are sequenced according to task priorities and generation time, a service life weighted Dijkstra algorithm is used, paths are distributed for the tasks on the basis of the service life weighted link cost between any two nodes, unmanned aerial vehicle nodes are locked, state fields and task fields of the unmanned aerial vehicle nodes are updated, and the unmanned aerial vehicle nodes are distributed according to the state fields and the task fields of the unmanned aerial vehicle nodes. The priority of the tasks which are not successfully distributed is dynamically adjusted based on a dynamic adjustment rule; and after the task is completed, releasing the nodes and triggering the next round of scheduling. According to the method, the timely delivery rate of the key tasks is remarkably improved, the link retransmission and congestion risks are reduced, and the service fairness of the low-priority tasks is ensured.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of unmanned aerial vehicle communication, and particularly relates to a method and system for controlling swarming communication of a group of unmanned aerial vehicles based on link lifetime perception. BACKGROUND

[0002] In recent years, unmanned aerial vehicles have been widely used in environmental monitoring, post-disaster emergency, intelligent agriculture, and space-ground integration communication. Compared with traditional ground wireless networks, self-organizing networks composed of multiple unmanned aerial vehicles have the advantages of rapid deployment and rich line-of-sight links, but they also face natural constraints such as highly dynamic topology, frequent link interruption, and limited node computing power and energy. In real-world tasks, there is often a need for multiple unmanned aerial vehicles to simultaneously upload large amounts of data (such as high-definition videos or batch sensor information) to a central base station / server, and reliable single-hop direct links do not always exist, which inevitably requires the use of other unmanned aerial vehicle nodes for multi-hop relaying.

[0003] In typical applications, an unmanned aerial vehicle network may simultaneously have several heterogeneous tasks: emergency image backhaul, ordinary inspection data, and low-priority log upload. The wireless transceiver, buffer, and energy of a node can only be occupied by one data stream; when multiple tasks compete, if there is a lack of fine-grained scheduling, high-priority tasks may be blocked or low-priority tasks may be starved for a long time, leading to network congestion and global performance collapse.

[0004] Existing research on the instability of unmanned aerial vehicle links still mainly focuses on single-task and single-routing perspectives, and lacks systematic consideration of global resource scheduling for multiple tasks. In addition, mainstream methods usually assume the existence of a centralized controller, which is difficult to adapt to the distributed constraints in unmanned aerial vehicle clusters, where each unmanned aerial vehicle needs to make independent decisions and can only obtain neighbor information.

[0005] Swarming control is inspired by group behavior in nature and emphasizes achieving global coordination through local rules. When introduced into communication networks, it can adjust the sending rate and forwarding direction of nodes without central coordination, naturally fitting the distributed, dynamic, and adaptive characteristics of unmanned aerial vehicle clusters. However, existing swarming communication control mainly focuses on traffic balancing and congestion avoidance, and has not yet incorporated link lifetime and multiple task priorities into a unified framework.

[0006] In summary, for a multiple unmanned aerial vehicle self-organizing network, a system is needed that can meet the following requirements: Real-time evaluation of link remaining lifetime, considering both latency and failure risk during routing to avoid disconnection halfway due to short-lived links; Support for multiple task priorities, preemption, and aging mechanisms to ensure that critical tasks are completed in a timely manner and that low-priority tasks can eventually be serviced; Run under the condition of complete distribution, only rely on neighbor broadcast, to adapt to the reality that unmanned aerial vehicle is constantly moving, temporarily enters the network / retires from the network; Control network congestion and node conflict, avoid multi-task simultaneously occupying the same relay node. SUMMARY

[0007] In view of the problems that the link residual life of the unmanned aerial vehicle cluster is insufficient under high-speed dynamic topology, and the node resources are competed by concurrent tasks, causing congestion and task starvation, the application provides a link life perception-based unmanned aerial vehicle swarm swarming communication control method and system.

[0008] To solve the above technical problems, the application provides the following technical solutions: A link life perception-based unmanned aerial vehicle swarm swarming communication control method, comprising the following steps: Step S1. Constructing an unmanned aerial vehicle self-organizing network, a plurality of collection unmanned aerial vehicles perform a plurality of data collection tasks according to a preset trajectory, and transmit data to a central base station / server through multi-hop relay; Step S2. Setting a state field for each unmanned aerial vehicle node, including: node occupation mark, current occupation task number and node locking remaining time length; Step S3. Setting a task field for each task, including: unique identification, source node, destination node, data volume, dynamic priority and task state; Step S4. Based on the unmanned aerial vehicle self-organizing network, calculating the life-weighted link cost between any two unmanned aerial vehicle nodes; Step S5. Adopting a multi-task preemption scheduling strategy, sorting tasks according to task priority and generation time, using life-weighted Dijkstra algorithm, and based on the life-weighted link cost between the two nodes, assigning paths to tasks, locking unmanned aerial vehicle nodes and updating the state field and task field of the unmanned aerial vehicle nodes, and dynamically adjusting the priority of the tasks that are not successfully assigned based on dynamic adjustment rules; Step S6. Releasing the node after the task is completed and triggering the next round of scheduling.

[0009] Further, the unmanned aerial vehicle self-organizing network in step S1 comprises: A plurality of collection unmanned aerial vehicles perform a plurality of data collection tasks according to a preset trajectory, and transmit data to a designated central base station / server through a wireless link in a multi-hop manner; each task is defined as transmitting a data volume of Data packets from a source node to a destination node, when the distance between the source node and the destination node exceeds the maximum reliable transmission distance threshold , multi-hop relay forwarding through other unmanned aerial vehicle nodes in the network is needed.

[0010] Further, the life-weighted link cost between any two UAV nodes in step S4 is:

[0011] wherein, denotes the life-weighted link cost between neighbor nodes at time ; denotes the required delay for stable transmission of data packets between neighbor nodes at time ; is a risk-averse coefficient; denotes the remaining life of the link between nodes at time ; is a link failure risk penalty term, reflecting the ratio of the required time for transmission to the surplus time ; when the ratio increases, it means that the risk of the link remaining life being insufficient to support the completion of the task intensifies, thereby significantly increasing the link cost.

[0012] Further, the remaining life of the link between UAV node and node at time is:

[0013] wherein, denotes the maximum value in the set; denotes the distance between the two nodes after seconds from time ; denotes the maximum distance threshold that UAVs can maintain reliable communication; denotes the length of time that the link can remain continuously effective; therefore, the physical meaning of is that, from the current time , to the time when the link is first terminated in the future, the link can remain continuously effective for the longest time .

[0014] Further, the mechanism for locking the UAV node in step S5 includes: When the node is occupied by a task, set its occupation marker to the occupied state, write the task number, set the lock countdown to the estimated transmission time of the task, and broadcast the state to the neighbor nodes; ​​After the lock countdown reaches zero, the node is automatically reset to an idle state and the task number is cleared, and the status update is broadcast again; a new task is allowed to occupy the node only when it is in an idle state.

[0015] Furthermore, the multi-task preemptive scheduling strategy includes: Filter all tasks in the waiting state, sort them from highest to lowest priority value, and sort tasks with the same priority from earliest to latest creation time; sequentially search for the source-to-destination path with the lowest lifetime-weighted link cost in the set of idle nodes for each task; if the path is found successfully, mark the path node as occupied and associate it with the current task, and update the task to running status; for tasks that are not successfully assigned, increase their priority value by one level.

[0016] Furthermore, the dynamic adjustment rule for priority is as follows: The initial priority of a task is determined by its urgency. The priority value is increased by one level after each scheduling failure. High-priority tasks can preempt node resources already occupied by low-priority tasks.

[0017] Furthermore, the lifetime-weighted Dijkstra algorithm performs the following operations: Construct a network topology graph based on the set of idle nodes; Assign a lifetime-weighted link cost value to each edge in the network topology graph; The shortest path algorithm is used to iteratively calculate the shortest path from the source node to the destination node with the lifetime-weighted link cost. A complete transmission path is generated by tracing back the predecessor node.

[0018] Furthermore, it also includes a link failure handling mechanism: If insufficient remaining link lifetime is detected during data transmission, the task will be immediately terminated and a node release command will be broadcast. The status of the interrupted task will be reset to the waiting state and added back to the scheduling queue.

[0019] On the other hand, the present invention provides a swarm communication control system for unmanned aerial vehicles (UAVs) based on link lifetime awareness, comprising: Self-organizing network building module. It is used to build a drone self-organizing network in which multiple data collection drones transmit data to a central base station / server through multi-hop relays; Status field setting module. It is used to set status fields for each drone node, including: node occupancy flag, current occupancy task number, and remaining node lock duration; The task field setting module is used to set task fields for each task, including: unique identifier, source node, destination node, data volume, dynamic priority, and task status. Link cost calculation module. It is used to calculate the lifetime-weighted link cost between any two UAV nodes based on the UAV self-organizing network; The path calculation module is used to employ a multi-task preemptive scheduling strategy, sort tasks by task priority and generation time, use the lifetime-weighted Dijkstra algorithm, and allocate paths for tasks based on the lifetime-weighted link cost between any two nodes, lock UAV nodes, and update the status field and task field of the UAV nodes. Tasks that are not successfully allocated will have their priorities dynamically adjusted based on dynamic adjustment rules. The scheduling module is used to release nodes and trigger the next round of scheduling after a task is completed.

[0020] Compared with the prior art, the present invention has the following beneficial effects: In a fully distributed UAV self-organizing network, this invention realizes link lifetime-aware path selection, task priority scheduling, and node resource mutual exclusion control, significantly improving the timely delivery rate of critical tasks, reducing the risk of link retransmission and congestion, and ensuring the service fairness of low-priority tasks. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0022] Figure 1 This is the overall flowchart of the present invention.

[0023] Figure 2 This invention relates to a multi-UAV self-organizing network.

[0024] Figure 3 This is a flowchart of the multi-task preemptive path allocation strategy of the present invention.

[0025] Figure 4 This is a schematic diagram of the start and end times and path selection for the transmission of 35 tasks in an embodiment of the present invention.

[0026] Figure 5 This is a Gantt chart showing the execution of 35 tasks in an embodiment of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0028] Example 1 The invention will now be further described with reference to the accompanying drawings.

[0029] In this embodiment, 10 drones are randomly distributed within a 500m × 500m airspace. Each drone employs a random movement model: flying at a random speed of 10-15m / s in a random direction, pausing for 2 seconds after reaching a new location. Simultaneously, each drone's new location must be within the communication radius (250m) of at least one other drone. The continuous movement of the drones causes the network topology to change dynamically every second, and the link duration (lifetime) is determined by the relative motion between the drones.

[0030] The following is combined Figure 1 This invention introduces a link lifetime-aware distributed swarm communication control method for UAV swarms, comprising the following steps: Step S1. Construct a self-organizing network of drones, in which multiple data collection drones transmit data to the central base station / server through multi-hop relays; As shown in Figure 2, Each data-collecting drone executes a data collection task according to a preset trajectory and transmits the data to a designated central base station / server via a multi-hop wireless link. Each task is defined as the amount of data transmitted from the source node to the destination node. Data packets. When the distance between the source node and the destination node exceeds the maximum reliable transmission distance threshold. In such cases, multi-hop relay forwarding is required using other drone nodes in the network to ensure that the data successfully reaches the destination node.

[0031] Step S2. Set status fields for each drone node, including: node occupancy flag, current occupancy task number, and remaining node lock duration; The status fields in this embodiment are shown in the following table:

[0032] Step S3. Set task fields for each task, including: unique identifier, source node, destination node, data volume, dynamic priority, and task status; The task fields in this embodiment are shown in the following table:

[0033] Step S4. Based on the UAV self-organizing network, calculate the lifetime-weighted link cost between any two UAV nodes; Each drone clearly knows the movement trajectories of all other drones and the specific coordinates of the central base station / server. Therefore, at any given moment... Any two nodes and The distance can be determined by Calculation. Among them, Representing nodes respectively and At any moment The location coordinates. The transmission rate can be determined by... Calculation, where It's bandwidth. It is transmission power. It is the channel gain between the two drones. It is a distance of 1 m Channel gain at that location, This is the path loss coefficient. Then in Stable data packet transmission The delay can be caused by calculate.

[0034] Only the distance between nodes Only when the distance exceeds this threshold can a stable and effective communication link exist between them. When the distance exceeds this threshold, the link is considered to be interrupted. Due to the mobility of UAV nodes, the remaining link lifetime (LLT) and transmission rate between any two nodes change dynamically over time. drone nodes and nodes Remaining lifetime of the link The calculation is as follows:

[0035] The above formula means finding a maximum value. ,exist Within this time frame, at every moment Below, drone nodes and nodes The distances are all less than the threshold. .this It refers to the remaining lifetime of the downlink at time t. Because drone network topology is highly dynamic and links can be disconnected at any time, it is necessary to consider not only transmission latency but also whether the link duration is sufficient to support data transmission.

[0036] To address this, this application proposes a novel path selection metric—lifetime-weighted link cost. This metric comprehensively considers the impact of transmission latency and link duration on path availability, more accurately reflecting the actual transmission status of links in dynamic UAV network environments, and facilitating the effective selection of more stable and reliable transmission paths. At any given time... Neighbor nodes The lifetime-weighted link cost between them can be determined by .in, This is the risk aversion coefficient (the larger the coefficient, the more sensitive the user is to link failure). The link has failed. According to the formula... This allows us to obtain the total distance along the entire link from the source node to the target node. Subsequently, we will use Dijkstra's algorithm to find a path with the minimum total lifetime weighted link cost for each source node.

[0037] Step S5. A multi-task preemptive scheduling strategy is adopted, tasks are sorted according to task priority and generation time, lifespan-weighted Dijkstra algorithm is used, and a path is allocated for the task based on the lifespan-weighted link cost between any two nodes, the UAV node is locked and the status field and task field of the UAV node are updated, and the priority of unsuccessfully allocated tasks is dynamically adjusted based on dynamic adjustment rules. When multiple tasks request transmission simultaneously, node contention may occur, meaning a drone node can only serve one task at a time. To avoid this conflict, this invention introduces a status flag for each node: busy_flag, task_id, and TTL. Whenever a node is preempted, it broadcasts its own status information, allowing all drones in the network to understand the current resource usage in real time. After TTL seconds, the node will automatically release itself, the task_id will be cleared, the busy_flag will be reset to zero, and the node will broadcast its status information again.

[0038] Because different types of tasks have varying sensitivities to transmission latency, this embodiment introduces a priority mechanism in task scheduling to effectively meet task timeliness requirements and avoid individual tasks being unable to obtain node resources for extended periods (i.e., task "starvation"). The specific rules are as follows: Each task has an initial priority due to its different type; During the scheduling process, all tasks are preempted in the order of priority plus the timestamp generated. If a task fails to occupy a node in a scheduling round, its priority will be automatically increased by one in the next scheduling round to ensure that low-priority tasks also have a chance to complete and avoid starvation.

[0039] like Figure 3As shown, the multi-task preemptive path allocation strategy process in this embodiment is as follows: The scheduling loop continues to fire while the queue is not empty: Task sorting: Extract all tasks with status=WAIT, sort them by "high priority → low priority, then by creation time → late priority"; Node preemption and path optimization: For each sorted task, the lifetime-weighted Dijkstra algorithm searches the current set of idle nodes for the shortest lifetime-weighted link cost from the source node to the target node; if a feasible path is found, the involved nodes are locked (by setting busy_flag=1, writing task_id, and setting...). (Estimated transmission time), and set the task status to RUN; otherwise, the task remains in the queue waiting; Priority aging: Tasks that are not successfully assigned in this round will have their priority increased by ++, ensuring that low-priority tasks will eventually get a service opportunity.

[0040] The specific process of lifetime-weighted Dijkstra's algorithm is as follows: Input: Set of free nodes V, set of available edges E, LLT sum of each edge Source node s, destination node d; Preprocessing: Calculate the weight w(u,v) for each element in E; Core iteration: Maintain a min-heap Q to store the cumulative cost dist[] of undetermined nodes, and cyclically pop the smallest dist node and relax its adjacent edges; Termination and backtracking: The process ends when the popped node is d or the heap is empty; if dist[d]=∞, the current topology is considered unreachable, otherwise backtracking is performed according to prev[] to obtain the optimal path and return it to the scheduler.

[0041] When a task is in the RUN state, the source node continuously sends data to the central UAV along the locked path; once the task is completed, all nodes on the path are released (the triplet is reset and broadcast), the task status is set to DONE, it is removed from the queue, and the next round of scheduling is triggered.

[0042] In this embodiment, the Dijkstra algorithm's path finding process is as follows: Input information: Current available node set (Only nodes with busy_flag=0)

[0043] Current available edge set :satisfy And both endpoints are in .

[0044] Instantaneous rate of each available edge Remaining lifespan .

[0045] Source node Target node .

[0046] Output: from arrive shortest path Minimum lifetime weighted link cost If from arrive If it is unattainable .

[0047] The specific steps include: I. Pretreatment For each available communication link (u, v): First, calculate the time required for a data packet to be transmitted once through this link. .

[0048] Determine link validity: If the link has a remaining lifetime... Less than or equal to this single-hop transmission time This means that the link may fail before the data packet is fully transmitted. In this case, the link is considered unavailable, and its cost is reduced. Let it be positive infinity. If the remaining lifetime of the link is long enough... Then, the lifetime-weighted distance of the link is calculated according to the algorithm definition. .

[0049] II. Initialization Two key variables are maintained for each node: `dist[u]`: records the current known minimum cumulative cost from source node `s` to node `u`. Initially, `dist[s] = 0`, and `dist[u]` for all other nodes is set to positive infinity, indicating that no reachable path has been found or the cost is unknown. `prev[u]`: records the predecessor node of node `u` when the current minimum cumulative cost `dist[u]` is reached. This is used to eventually backtrack the complete path. Initially, `prev[u]` for all nodes is set to empty.

[0050] Create a priority queue (min-heap) Q and add all nodes to the queue. The queue is sorted according to the dist[u] value of each node, and the node with the smallest dist[u] is at the top of the heap and will be processed first.

[0051] III. Finding the shortest path (main loop) As long as priority queue Q is not empty, repeat the following process: Select the node with the lowest current cost: Take out and remove the node u (the top node of the heap) with the smallest current dist[u] value from queue Q.

[0052] Check if the goal has been achieved: If the extracted node u is the destination node d, it means that the optimal path from s to d has been found, and the loop can be terminated in advance.

[0053] Explore neighboring nodes: If u is not d, then check each of u's neighboring nodes v in turn. If the cost of link (u, v) is... If the value is positive infinity, then ignore this neighbor v and continue checking the next neighbor. For valid links, calculate the cost of a new candidate path: alt = dist[u] + This represents the potential total cost of reaching neighbor v through the current node u. If the calculated candidate cost alt is less than the current record value dist[v] of node v, update dist[v] = alt and prev[v] = u. Call the priority queue DecreaseKey(Q, v) to adjust the position of node v in the priority queue Q to maintain the property of the queue being ordered by dist.

[0054] IV. Tracing the Build Path Check the final cumulative cost dist[d] of the destination node d: If dist[d] is still positive infinity, it means that in the current network state, there is no feasible path from the source node s to the destination node d. If dist[d] is a finite value, it means that a feasible path exists. Construct this path by backtracking the predecessor node.

[0055] V. Output the final result like Figure 4 and Figure 5 As shown, if the path list `path` is empty, then in the current network state, there is no feasible path from `s` to `d` that satisfies the lifetime constraint. If the path list `path` is not empty, then `Path = [s, …, d]` is the optimal path from the source node `s` to the destination node `d`.

[0056] If flight causes premature link failure, any node on the path that detects insufficient actual LLT will immediately abort and broadcast a release signal. The scheduler will then place the task back into the WAIT queue and quickly reroute. The parameters can flexibly balance latency performance and link stability to adapt to the timeliness or reliability requirements of different scenarios.

[0057] Step S6. After the task is completed, release the node and trigger the next round of scheduling.

[0058] Example 2 This embodiment provides a swarm communication system for unmanned aerial vehicles (UAVs) based on link lifetime awareness, including: Self-organizing network construction module. It is used to build a self-organizing network of drones, in which multiple data acquisition drones perform multiple data acquisition tasks according to a preset trajectory and transmit the data to the central base station / server through multi-hop relay; Status field setting module. It is used to set status fields for each drone node, including: node occupancy flag, current occupancy task number, and remaining node lock duration; The task field setting module is used to set task fields for each task, including: unique identifier, source node, destination node, data volume, dynamic priority, and task status. Link cost calculation module. It is used to calculate the lifetime-weighted link cost between any two UAV nodes based on the UAV self-organizing network; The path calculation module is used to employ a multi-task preemptive scheduling strategy, sort tasks by task priority and generation time, use the lifetime-weighted Dijkstra algorithm, and allocate paths for tasks based on the lifetime-weighted link cost between any two nodes, lock UAV nodes, and update the status field and task field of the UAV nodes. Tasks that are not successfully allocated will have their priorities dynamically adjusted based on dynamic adjustment rules. The scheduling module is used to release nodes and trigger the next round of scheduling after a task is completed.

[0059] It should be understood that any parts not described in detail in this specification belong to the prior art.

[0060] It should be understood that the above description of the preferred embodiments is quite detailed, but this should not be construed as limiting the scope of protection of this invention. It is neither necessary nor possible to exhaustively describe all possible implementations. Those skilled in the art, guided by this invention, can make substitutions or modifications without departing from the scope of the claims, all of which fall within the scope of protection of this invention. The scope of protection of this invention should be determined by the appended claims.

Claims

1. A method for controlling swarm communication of unmanned aerial vehicles (UAVs) based on link lifetime awareness, characterized in that, Includes the following steps: Step S1. Construct a self-organizing network of drones, in which multiple data acquisition drones perform multiple data acquisition tasks according to a preset trajectory and transmit the data to the central base station / server through multi-hop relay; Step S2. Set status fields for each drone node, including: node occupancy flag, current occupancy task number, and remaining node lock duration; Step S3. Set task fields for each task, including: unique identifier, source node, destination node, data volume, dynamic priority, and task status; Step S4. Based on the UAV self-organizing network, calculate the lifetime-weighted link cost between any two UAV nodes; Step S5. A multi-task preemptive scheduling strategy is adopted, tasks are sorted according to task priority and generation time, lifespan-weighted Dijkstra algorithm is used, and a path is allocated for the task based on the lifespan-weighted link cost between any two nodes, the UAV node is locked and the status field and task field of the UAV node are updated, and the priority of unsuccessfully allocated tasks is dynamically adjusted based on dynamic adjustment rules. Step S6. After the task is completed, release the node and trigger the next round of scheduling.

2. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 1, characterized in that, The drone self-organizing network in step S1 includes: Multiple data acquisition drones execute multiple data acquisition tasks according to preset trajectories, and transmit the data to the central base station / server via multi-hop relay. Each task involves transmitting a certain amount of data from the source node to the destination node. Data packets that are transmitted when the distance between the source node and the destination node exceeds the maximum reliable transmission distance threshold. At the same time, it uses other drone nodes in the network for multi-hop relay forwarding.

3. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 2, characterized in that, The lifetime-weighted link cost between any two UAV nodes in step S4 is: in, Indicates at time Neighbor nodes The lifetime-weighted link cost between them Indicates at time Neighbor nodes Stable data packet transmission The required latency; Risk aversion coefficient; Indicates at time node The remaining lifetime of the link between them; This is a penalty item for link failure risk, reflecting the time required for transmission. with spare time The ratio; when this ratio increases, it means that the risk of the remaining link lifetime being insufficient to support the completion of the task increases, thus significantly increasing the link cost.

4. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 3, characterized in that, At any moment drone nodes and nodes Remaining lifetime of the link for: in, This indicates retrieving the maximum value from the set. Indicates from time Beginning, then The distance between the two nodes after seconds; This represents the maximum distance threshold at which drones can maintain reliable communication. This indicates the duration for which the link can remain effective; therefore... The physical meaning is that, from the current moment... From the beginning, until some point in the future The longest period of time a link can remain continuously effective until it is first terminated. .

5. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 4, characterized in that, The mechanism for locking the drone node in step S5 includes: When a node is occupied by a task, its occupation mark is set to occupied status, the task number is written, the lock countdown is set to the estimated transmission time of the task, and the status is broadcast to neighboring nodes. After the lock countdown reaches zero, the node is automatically reset to an idle state and the task number is cleared, and the status update is broadcast again; a new task is allowed to occupy the node only when it is in an idle state.

6. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 5, characterized in that, The multi-task preemptive scheduling strategy includes: Filter all tasks in the waiting state, sort them from highest to lowest priority value, and sort tasks with the same priority from earliest to latest creation time; sequentially search for the source-to-destination path with the lowest lifetime-weighted link cost in the set of idle nodes for each task; if the path is found successfully, mark the path node as occupied and associate it with the current task, and update the task to running status; for tasks that are not successfully assigned, increase their priority value by one level.

7. A method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 6, characterized in that, The dynamic adjustment rule for priority is as follows: The initial priority of a task is determined by its urgency. The priority value is increased by one level after each scheduling failure. High-priority tasks can preempt node resources already occupied by low-priority tasks.

8. The method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 7, characterized in that, The lifetime-weighted Dijkstra algorithm performs the following operations: Construct a network topology graph based on the set of idle nodes; Assign a lifetime-weighted link cost value to each edge in the network topology graph; The shortest path algorithm is used to iteratively calculate the shortest path from the source node to the destination node with the lifetime-weighted link cost. A complete transmission path is generated by tracing back the predecessor node.

9. A method for controlling UAV swarm swarm communication based on link lifetime awareness according to claim 8, characterized in that, It also includes a link failure handling mechanism: If insufficient remaining link lifetime is detected during data transmission, the task will be immediately terminated and a node release command will be broadcast. The status of the interrupted task will be reset to the waiting state and added back to the scheduling queue.

10. A swarm communication control system for unmanned aerial vehicles (UAVs) based on link lifetime awareness, characterized in that, include: Self-organizing network construction module. It is used to build a self-organizing network of drones, in which multiple data acquisition drones perform multiple data acquisition tasks according to a preset trajectory and transmit the data to the central base station / server through multi-hop relay; Status field setting module. It is used to set status fields for each drone node, including: node occupancy flag, current occupancy task number, and remaining node lock duration; The task field setting module is used to set task fields for each task, including: unique identifier, source node, destination node, data volume, dynamic priority, and task status. Link cost calculation module. It is used to calculate the lifetime-weighted link cost between any two UAV nodes based on the UAV self-organizing network; The path calculation module is used to employ a multi-task preemptive scheduling strategy, sort tasks by task priority and generation time, use the lifetime-weighted Dijkstra algorithm, and allocate paths for tasks based on the lifetime-weighted link cost between any two nodes, lock UAV nodes, and update the status field and task field of the UAV nodes. Tasks that are not successfully allocated will have their priorities dynamically adjusted based on dynamic adjustment rules. The scheduling module is used to release nodes and trigger the next round of scheduling after a task is completed. The link lifetime-aware UAV swarm swarm communication control system is used to execute the steps in the link lifetime-aware UAV swarm swarm communication control method according to any one of claims 1-9.

Images