An island clustering and unmanned aerial vehicle trajectory planning method based on minimizing information age

Abstract

The application discloses an island clustering and unmanned aerial vehicle trajectory planning method based on minimizing information age, and the islands covered by an unmanned aerial vehicle assisted wireless power supply communication network are clustered according to the hovering point positions of the unmanned aerial vehicle on the islands, so that the clustering results of all the islands are obtained; the final clustering results and the hovering point access sequence in each cluster are determined according to the limited battery capacity constraint of the unmanned aerial vehicle in the cluster, the data unloading of the unmanned aerial vehicle after completing a task and flying back to a hybrid access point, and the average information age constraint of the hybrid access point; and finally, each cluster is sequentially flown according to the hovering point access sequence in the cluster according to the final clustering results. The technical scheme of the application reduces the average information age of the data received by the hybrid access point in the unmanned aerial vehicle assisted wireless power supply communication network.

Description

Technical Field

[0001] The present application belongs to the field of UAV-assisted wireless power supply communication technology, and in particular relates to an island clustering and UAV trajectory planning method based on minimizing information age. Background Art

[0002] In recent years, drones, due to their flexibility and mobility, have been deployed in wireless power supply communication networks to mitigate the near-far effect, sending radio frequency signals to sensor nodes for power and data reception. As practical applications demand increased freshness of acquired data, information age has attracted widespread attention and has become a new network performance metric. Information age primarily refers to the time between data generation and reception at the receiving end, characterizing the data's freshness and serving as a measure of the freshness of received data. By reducing the time interval between data generation and reception, the freshness of data received by the receiving end can be optimized. Based on drone-assisted wireless power supply communication networks in practical applications, it is particularly significant to rationally utilize the limited battery capacity of drones, generate data after sensor nodes capture energy, reduce the data generation cycle of sensor nodes, and lower the average information age of data received from sensor nodes.

[0003] In a drone-assisted wireless power supply communication network consisting of sensor nodes located on multiple islands (referred to as islands), the islands are divided into clusters. A drone flies to a single hovering point on an island, provides power to the corresponding sensor node, and receives data from the sensor node. After receiving data from all sensor nodes in the cluster, it returns to the hybrid access point to offload the received data, is recharged by the hybrid access point, and then flies to the next hovering point in the cluster. The order in which the drone visits the hovering point and the number of sensor nodes on the island corresponding to the hovering point affect the average information age of the data received by the hybrid access point in the drone-assisted wireless power supply communication network.

[0004] Due to the limited battery capacity of the drone, it must return to the hybrid access point to offload data received from the sensor nodes to the hybrid access point, where it is then recharged. Once fully charged, the drone continues to fly to the hovering point of the next cluster, traversing the hovering points within the cluster to complete the tasks of powering the sensor nodes and receiving data from them.

[0005] However, in drone-assisted wireless power supply communication networks, where the islands corresponding to hovering points have varying numbers of sensor nodes, and considering the optimization objective of average data information age, drones cannot simply fly along the shortest path between islands within a cluster. This is to avoid the following situation: the island corresponding to the first hovering point visited may contain a larger number of sensor nodes. Since the data generated by the sensor nodes on this island is received by the drone first, it will take the longest time for the hybrid access point to receive the data from sensor nodes on other islands in the cluster. Furthermore, the large number of sensor nodes on this island will increase the average information age of the data generated by all sensor nodes received by the drone.

[0006] For the drone-assisted wireless power supply communication network on the island containing different numbers of sensing nodes, combined with the limited battery capacity of the drone, how to reduce the average information age of the data received by the hybrid access points is a technical problem that needs to be solved in this field. Summary of the Invention

[0007] The purpose of this application is to provide an island clustering and UAV trajectory planning method based on minimizing information age to reduce the average information age of data received by hybrid access points.

[0008] In order to achieve the above objectives, the technical solutions of this application are as follows:

[0009] A method for island clustering and UAV trajectory planning based on minimizing information age is applied to UAV-assisted wireless power supply communication networks, including:

[0010] According to the position of the drone's hovering point on the island, all islands covered by the drone-assisted wireless power supply communication network are clustered to obtain the clustering results of all islands;

[0011] According to the limited battery capacity constraint of the UAVs in the cluster and the constraint of minimizing the average information age, the final clustering result and the order of visiting the hovering points in each cluster are determined;

[0012] According to the final clustering results, each cluster is flown in turn according to the order of visiting the hovering points within the cluster.

[0013] Furthermore, the final clustering result and the order of visiting hovering points in each cluster are determined according to the limited battery capacity constraint of the UAVs in the cluster and the constraint of minimizing the average information age, including the intra-cluster processing stage and the cluster merging stage;

[0014] In the intra-cluster processing stage, for any cluster, the following operations are performed:

[0015] The particle swarm optimization algorithm is used to obtain the shortest path of the hovering points in the cluster, and the average information age corresponding to the positive and reverse access according to the shortest path is calculated. The hovering point access sequence corresponding to the smaller average information age is selected as the initial access sequence.

[0016] Dynamically adjust the order of hover point visits based on the number of sensing nodes on the island corresponding to each hover point;

[0017] Calculate the energy consumed by the drone to complete the mission according to the hovering point visit sequence. If the energy exceeds the limited battery capacity of the drone, the islands in the current cluster are re-clustered and the intra-cluster processing phase is performed on any cluster after re-clustering. If the energy does not exceed the limited battery capacity of the drone, the intra-cluster processing phase is terminated.

[0018] After all clusters are processed in the intra-cluster processing stage, the cluster merging stage is performed on the obtained clustering results and the access order of the hovering points in the cluster, and the following operations are performed:

[0019] According to the cluster center position obtained by clustering, the islands in two clusters whose cluster centers are less than the distance threshold are merged into a new cluster;

[0020] The operations of the intra-cluster processing stage are performed on the new cluster, and the energy consumed by the UAV to complete the task according to the hovering point visit order is calculated. If it exceeds the limited battery capacity of the UAV, the merging operation of the new cluster is canceled. If it does not exceed the limited battery capacity of the UAV, the new cluster and the hovering point visit order within the new cluster are retained.

[0021] Furthermore, the dynamically adjusting the hover point access order according to the number of sensing nodes on the island corresponding to each hover point includes:

[0022] Step B1: Taking the first hovering point in the hovering point access sequence as the current hovering point, executing the hovering point exchange adjustment strategy to obtain a new hovering point access sequence.

[0023] Step B2: cyclically execute step B1 on the new hover point access sequence until the obtained hover point access sequence no longer changes.

[0024] Step B3: Taking the other hovering points in the hovering point access sequence as the current hovering points, execute the hovering point exchange adjustment strategy according to the methods of steps B1 and B2 until all the hovering points in the hovering point access sequence are traversed to obtain the final hovering point access sequence.

[0025] Furthermore, executing the hover point exchange adjustment strategy includes:

[0026] Compare the number of sensor nodes on the island corresponding to the current hovering point with the number of sensor nodes on the island corresponding to each hovering point after the current hovering point. If the number of sensor nodes on the island corresponding to the subsequent hovering point is smaller, swap the access order of the two hovering points; otherwise, stop comparing.

[0027] After each exchange of the access order of the two hovering points, the average information age corresponding to the new access order is calculated. If the average information age corresponding to the new access order is greater than or equal to the average information age corresponding to the original access order, the access order of the two hovering points is restored; otherwise, the exchanged access order is retained.

[0028] This application proposes an island clustering and drone trajectory planning method based on minimizing information age. Based on the limited battery capacity of a drone, the method determines island clustering results and flies between hovering points within the cluster, completing the tasks of power supply and data reception while hovering. By reducing the number of sensing nodes that experience a long time from data generation to data reception by hybrid access points, the average information age of data received from hybrid access points in a drone-assisted wireless power supply communication network is reduced. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 Flowchart of the island clustering and UAV trajectory planning method based on minimizing information age in this application;

[0030] Figure 2 Schematic diagram of the application scenario of this application. DETAILED DESCRIPTION

[0031] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0032] In order to reduce the average information age of data received by hybrid access points in UAV-assisted wireless power supply communication networks, such as Figure 1 As shown, this application proposes an island clustering and UAV trajectory planning method based on minimizing information age, including:

[0033] Step S1: Cluster all islands covered by the UAV-assisted wireless power supply communication network according to the hovering point position of the UAV on the island, and obtain clustering results for all islands.

[0034] The application scenarios of this application are as follows Figure 2 As shown in the figure, the UAV-assisted wireless power supply communication network includes: UAVs equipped with limited battery capacity, hybrid access points and sensing nodes randomly distributed on the islands. A certain number of islands are randomly distributed in the UAV-assisted wireless power supply communication network. Figure 2The figure is represented by circles. The drone's flight and hovering time are allocated as follows: First, a fully charged drone departs from the hybrid access point and selects a cluster according to the cluster index order from the island clustering results. Within each cluster, the drone uses a dynamically adjusted access order algorithm to determine the hovering point access order, and then visits each hovering point in that cluster in turn. The drone then flies to the corresponding hovering point on the island and, while hovering, provides energy to the sensor nodes on the island. The sensor nodes collect data from the environment using the captured energy and transmit this data to the drone. To ensure that the drone receives all data generated by all sensor nodes on the island corresponding to the hovering point within a single hover, the duration of the drone's hovering energy supply to the sensor nodes depends on the energy captured by the sensor node with the worst channel quality to transmit data to the drone, given that channel gain follows a large-scale fading model. When a sensor node transmits data to the drone, it uses a time division multiple access (TDMA) transmission protocol, transmitting data to the drone in ascending order of channel gain. After receiving data from all sensor nodes on the island corresponding to the hovering point, the drone will fly at a constant speed to the next hovering point, determined by the dynamically adjusted access order algorithm, to continue its tasks of powering and receiving data. After completing its tasks of powering and receiving data from all sensor nodes within the corresponding cluster, the drone will return to the hybrid access point, hovering and offloading data to it while the hybrid access point charges the drone. Once fully charged, the drone will select the next cluster based on the island's clustering results and complete its tasks in the aforementioned order.

[0035] In one example, Figure 2 In the paper, all islands covered by the UAV-assisted wireless power supply communication network are divided into cluster 1, cluster 2 and cluster 3. Each cluster includes several islands. Figure 2 The circles in the middle indicate that there are sensing nodes arranged in the islands, and the hovering point of the drone on the island is located at a height H just above the center of each island.

[0036] This application is aimed at a UAV-assisted wireless power supply communication network with different numbers of sensing nodes on the island, combined with the limited battery capacity of the UAV. In order to ensure that the UAV can receive data generated by all sensing nodes on the corresponding island m when hovering at the hovering point, there are N sensing nodes on the island m. m The drone uses a TDMA transmission protocol to transmit data to the sensing node. Due to the near-far effect and the large-scale fading model, the drone must ensure that the sensing node with the worst channel quality on island m can capture enough energy to transmit data. In other words, the sensing node farthest from the hovering point can successfully transmit data to the drone. Therefore, the duration of power supply to the sensing node during the drone's hovering depends on the energy required for the sensing node with the worst channel quality to transmit data to the drone.

[0037] This application focuses on the analysis of island clustering and trajectory planning for dynamically adjusting the order in which drones visit hovering points within a cluster. In this embodiment, a drone-assisted wireless power supply communication network consists of M non-overlapping islands. The drone flies at a constant altitude H and a constant speed. The hovering point is located at a height H above the center of each island, and each island has only one hovering point.

[0038] In a specific embodiment, the K-means clustering algorithm is used to cluster the hovering points according to their geographical locations, that is, islands corresponding to hovering points that are geographically close are grouped into one cluster.

[0039] The specific process of clustering is as follows:

[0040] Input: The position of the hovering point D = {x1, x2, ..., x M} and the number of clusters K;

[0041] Output: a set of K islands within a cluster;

[0042] Step F1: Initialize the positions of K cluster centers;

[0043] Step F2: Calculate the distances from all hovering points to the K cluster centers respectively, and assign the hovering points to the cluster corresponding to the cluster center closest to them;

[0044] Step F3: In each cluster, the average position of all hovering points in the cluster is used as the new cluster center position of the cluster;

[0045] Step F4: Repeat steps 2 and 3 until the cluster center position no longer changes or the preset number of iterations is reached, completing the clustering process of the islands corresponding to the hovering points.

[0046] Step S2: Determine the final clustering result and the order of visiting hovering points in each cluster based on the limited power constraints of the UAVs in the cluster and the constraint of minimizing the average information age.

[0047] This application proposes a drone-assisted wireless power supply communication network. When a drone accesses a cluster, its battery life must meet the energy constraints required to complete its mission within the cluster, while also optimizing the average age of data received by the hybrid access point. Within the cluster, the drone must fly to each island, power the sensing nodes and receive data on each island, and then return to the hybrid access point.

[0048] In a specific embodiment, the method determines the final clustering result and the hovering point access order within each cluster based on the limited battery capacity constraint of the drones within the cluster and the constraint of minimizing the average information age, including the intra-cluster processing stage and the cluster merging stage.

[0049] In the intra-cluster processing stage, for any cluster in the clustering results, the following operations are performed:

[0050] Step S2.1: Use the particle swarm optimization algorithm to obtain the shortest path to traverse the hovering points in the cluster, calculate the average information age corresponding to the positive and reverse access according to the shortest path, and select the hovering point access sequence corresponding to the smaller average information age as the initial access sequence.

[0051] This embodiment uses the particle swarm optimization algorithm (PSO) to find the shortest path between hovering points in a cluster. The steps are as follows:

[0052] Step A1: Initialize the order p of n particles visiting the hovering points in the cluster i .

[0053] Step A2: For each particle i, calculate its fitness value f i , that is, calculate according to the flight order p i Traverse the distance of the hovering points in the cluster, and use the hovering point access order corresponding to the minimum fitness value obtained in all iterations as the current optimal access order of particle i And update the minimum fitness value.

[0054] Step A3: Compare the minimum fitness values ​​of n particles and select the access order corresponding to the minimum fitness value as the optimal access order of the group particles

[0055] Step A4: Update the velocity and position of particle i according to the inertia weight ω:

[0056]

[0057] x i =x i +v i ;

[0058] c1 and c2 represent the cognitive coefficient and social coefficient respectively, and r1 and r2 are random numbers in the interval (0, 1).

[0059] The inertia weight ω satisfies the following formula:

[0060]

[0061] where ω max Indicates the maximum value of inertia weight, ω min Indicates the minimum value of the inertia weight.

[0062] Step A5: Until the number of iterations N is reached iter Otherwise, jump to step A2 and repeat the corresponding steps, and iter=iter+1.

[0063] The particle swarm optimization algorithm is a relatively mature technology in this field and will not be described in detail here.

[0064] Step S2.2: Dynamically adjust the order of visiting hovering points according to the number of sensing nodes on the island corresponding to each hovering point.

[0065] This embodiment dynamically adjusts the order of visiting hovering points based on the number of sensing nodes on the island corresponding to each hovering point. The adjustment process is as follows:

[0066] Step B1: Taking the first hovering point in the hovering point access sequence as the current hovering point, executing the hovering point exchange adjustment strategy to obtain a new hovering point access sequence.

[0067] Step B2: cyclically execute step B1 on the new hover point access sequence until the obtained hover point access sequence no longer changes.

[0068] Step B3: Taking the other hovering points in the hovering point access sequence as the current hovering points, execute the hovering point exchange adjustment strategy according to the methods of steps B1 and B2 until all the hovering points in the hovering point access sequence are traversed to obtain the final hovering point access sequence.

[0069] For the current hovering point, the hovering point exchange adjustment strategy is executed, including:

[0070] Compare the number of sensor nodes on the island corresponding to the current hovering point with the number of sensor nodes on the island corresponding to each hovering point after the current hovering point. If the number of sensor nodes on the island corresponding to the subsequent hovering point is smaller, swap the access order of the two hovering points; otherwise, stop comparing.

[0071] After each exchange of the access order of the two hovering points, the average information age corresponding to the new access order is calculated. If the average information age corresponding to the new access order is greater than or equal to the average information age corresponding to the original access order, the access order of the two hovering points is restored; otherwise, the exchanged access order is retained.

[0072] The following is a specific example to illustrate the above method of dynamically adjusting the hover point access order:

[0073] Assuming that the initial hovering point visit order is ABCDEFG, the corresponding number of sensing nodes on the island is 4-1-3-2-5-6-7;

[0074] First, take A as the current hovering point and compare it with B. Since A has 4 sensing nodes and B has 1 sensing node, they are swapped according to the rules. The order of the hovering points after the swap is BACDEFG.

[0075] Calculate the average information age corresponding to the access hover points corresponding to the access sequence BACDEFG and complete the energy supply and data reception. If the obtained average information age is greater than or equal to the average information age corresponding to the original access sequence ABCDEFG, then restore the access sequence to ABCDEFG. Otherwise, the new access sequence is BACDEFG.

[0076] Continue to use A as the current hover point and compare it with C using the same comparison method.

[0077] When comparing A with E, the comparison stops at this point, given that the hovering point of A has 4 sensor nodes, while the hovering point of E has 5 sensor nodes.

[0078] Assume that after the above comparison, the obtained hover point visit order is BCDAEFG, and the corresponding number of sensing nodes on the island is 1-3-2-4-5-6-7;

[0079] After that, if the hover point access order is BC, D, A, F, G, the same method is used to adjust the hover point B. The adjustment method is the same as above, and so on. Because hover point B has one sensor node and hover point C has three sensor nodes, it is found that the previous method cannot be used to adjust the hover point access order. The resulting hover point access order remains BC, D, A, F, G.

[0080] Next, take the second hovering point C in the above hovering point access sequence BCDAEFG as the current hovering point, and continue to adjust using the same method. Assume that the adjusted result is: BDCAEFG, and the corresponding number of sensing nodes on the island is 1-2-3-4-5-6-7, thus obtaining the final hovering point access sequence.

[0081] It should be noted that the above example is a result of obtaining a smaller average information age after each exchange. Depending on the specific calculation results, the final adjusted hover point access order may be different, which will not be repeated here.

[0082] Step S2.3: Calculate the energy consumed by the UAV to complete the task in the order of hovering point visits. If it exceeds the limited battery capacity of the UAV, re-cluster the islands in the current cluster and perform the intra-cluster processing phase on any cluster after re-clustering. If it does not exceed the limited battery capacity of the UAV, terminate the intra-cluster processing phase.

[0083] In this embodiment, after adjusting the order of visiting the hovering points within the cluster, if the energy consumed by the drone to sequentially visit and hover to complete the tasks of supplying power to the sensing nodes and receiving data from the sensing nodes exceeds its limited battery capacity, it means that for such a clustering result, the drone cannot complete the tasks of flying within the cluster and supplying power to the sensing nodes on the corresponding island and receiving data in a hovering state under the constraint of limited power, and the cluster needs to be re-divided.

[0084] For example, initially, the system is divided into five clusters: a, b, c, d, and e. The energy consumed by the drones in cluster b, flying and hovering according to the hovering point access sequence, providing power and receiving data, exceeds the limited battery capacity of the drones. Cluster b needs to be re-clustered. The clustering method is similar to steps F1-F4 and will not be repeated here. After clustering, nine clusters are obtained: a, b1-b5, c, d, and e.

[0085] After clustering, the operations in the intra-cluster processing stage are performed again, which will not be described in detail here.

[0086] After all clusters are processed in the intra-cluster processing stage, the clustering results and the order in which the drones visit the hovering points in the cluster are processed in the cluster merging stage. The following operations are performed:

[0087] According to the cluster center position obtained by clustering, the islands in two clusters whose cluster centers are less than the distance threshold are merged into a new cluster;

[0088] The operations of the intra-cluster processing stage are performed on the new cluster, and the energy consumed by the UAV to complete the task according to the hovering point visit order is calculated. If it exceeds the limited battery capacity of the UAV, the merging operation of the new cluster is canceled. If it does not exceed the limited battery capacity of the UAV, the new cluster and the hovering point visit order within the new cluster are retained.

[0089] As previously described, after each cluster has completed the intra-cluster processing phase, the clustering results for each cluster are obtained, including the islands included and the location of the cluster center. The order in which hovering points are visited within each cluster is also determined. Since the clustering results may result in two clusters being merged, hovering points within both clusters can be visited and the mission completed, provided the drone's battery capacity permits. Therefore, this application also performs a cluster merging phase.

[0090] During the cluster merging stage, based on the cluster center positions obtained from the clustering results, if the distance between two cluster centers is less than the distance threshold, the two clusters are merged and the intra-cluster processing stage operations are performed on the new cluster.

[0091] The energy consumed by the drone to complete the task according to the hovering point visit order is then calculated. If the energy exceeds the limited battery capacity of the drone, the merging operation of the new cluster is canceled. If the energy does not exceed the limited battery capacity of the drone, the new cluster and the hovering point visit order within the new cluster are retained.

[0092] Specifically, the system merges islands from two geographically close clusters into a single cluster, dynamically adjusts the visit order, and performs drone trajectory planning for the hovering points corresponding to the islands within the merged cluster. The system calculates the energy consumed by the drone to complete the mission by flying and hovering according to this visit order. If the energy consumed exceeds the limited battery capacity of the drone, the previous cluster merger is invalidated. If the energy consumed does not exceed the limited battery capacity of the drone, the previous cluster merger is valid, and the drone's hovering point visit order within the new cluster is updated.

[0093] The purpose of the merger is to make full use of the limited battery capacity of the UAV, reduce the number of times the UAV flies back to the hybrid access point to recharge, and reduce the cycle of data generation by the sensing node, thereby reducing the average information age of the data received by the hybrid access point in the UAV-assisted wireless power supply communication network.

[0094] Step S3: According to the final clustering result, each cluster is flown in turn according to the order of visiting the hovering points in the cluster.

[0095] After the above steps, the final clustering results and the hovering point access order within each cluster are obtained. The drone then traverses and visits the hovering points based on the clustering results and the hovering point access order within the cluster, completing the tasks of power supply and data reception while hovering. The drone can select hovering points within the cluster according to the subscript order of the clustering results and visit them sequentially. Within the corresponding cluster, the drone can fly according to the hovering point access order determined by the dynamically adjusted access order algorithm. Finally, it returns to the hybrid access point, unloads data to the hybrid access point, and is charged by the hybrid access point. The next cluster is then selected based on the island's clustering results, and the above steps are repeated. This application is not limited to the access order between clusters, and this will not be discussed in detail here.

[0096] The above-described embodiments merely represent several implementation methods of the present application. While the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art could make various modifications and improvements without departing from the spirit of the present application, all of which fall within the scope of protection of the present application. Therefore, the scope of protection of the present patent application shall be determined by the appended claims.

Claims

1. A method for island clustering and UAV trajectory planning based on minimizing information age, applied to UAV-assisted wireless power supply communication networks, characterized in that: The island clustering and UAV trajectory planning method based on minimizing information age includes: According to the position of the drone's hovering point on the island, all islands covered by the drone-assisted wireless power supply communication network are clustered to obtain the clustering results of all islands; According to the limited battery capacity constraint of the UAVs in the cluster and the constraint of minimizing the average information age, the final clustering result and the order of visiting the hovering points in each cluster are determined; According to the final clustering results, each cluster is flown in the order of visiting the hovering points within the cluster; The method determines the final clustering result and the hovering point access order within each cluster according to the limited battery capacity constraint of the UAVs within the cluster and the constraint of minimizing the average information age, including the intra-cluster processing stage and the cluster merging stage; In the intra-cluster processing stage, for any cluster, the following operations are performed: The particle swarm optimization algorithm is used to obtain the shortest path of the hovering points in the cluster, and the average information age corresponding to the positive and reverse access according to the shortest path is calculated. The hovering point access sequence corresponding to the smaller average information age is selected as the initial access sequence. Dynamically adjust the order of hover point visits based on the number of sensing nodes on the island corresponding to each hover point; Calculate the energy consumed by the drone to complete the mission according to the hovering point visit sequence. If the energy exceeds the limited battery capacity of the drone, the islands in the current cluster are re-clustered and the intra-cluster processing phase is performed on any cluster after re-clustering. If the energy does not exceed the limited battery capacity of the drone, the intra-cluster processing phase is terminated. After all clusters are processed in the intra-cluster processing stage, the cluster merging stage is performed on the obtained clustering results and the access order of the hovering points in the cluster, and the following operations are performed: According to the cluster center position obtained by clustering, the islands in two clusters whose cluster centers are less than the distance threshold are merged into a new cluster; The operations of the intra-cluster processing stage are performed on the new cluster, and the energy consumed by the UAV to complete the task according to the hovering point visit order is calculated. If it exceeds the limited battery capacity of the UAV, the merging operation of the new cluster is canceled. If it does not exceed the limited battery capacity of the UAV, the new cluster and the hovering point visit order within the new cluster are retained.

2. The island clustering and UAV trajectory planning method based on minimizing information age according to claim 1 is characterized in that: The dynamically adjusting the hover point access order according to the number of sensing nodes on the island corresponding to each hover point includes: Step B1: Taking the first hovering point in the hovering point access sequence as the current hovering point, executing the hovering point exchange adjustment strategy to obtain a new hovering point access sequence; Step B2: cyclically execute step B1 on the new hover point access sequence until the obtained hover point access sequence no longer changes; Step B3: Taking the other hovering points in the hovering point access sequence as the current hovering points, execute the hovering point exchange adjustment strategy according to the methods of steps B1 and B2 until all the hovering points in the hovering point access sequence are traversed to obtain the final hovering point access sequence.

3. The island clustering and UAV trajectory planning method based on minimizing information age according to claim 2 is characterized in that: The execution of the hover point exchange adjustment strategy includes: Compare the number of sensor nodes on the island corresponding to the current hovering point with the number of sensor nodes on the island corresponding to each hovering point after the current hovering point. If the number of sensor nodes on the island corresponding to the subsequent hovering point is smaller, swap the access order of the two hovering points; otherwise, stop comparing. After each exchange of the access order of the two hovering points, the average information age corresponding to the new access order is calculated. If the average information age corresponding to the new access order is greater than or equal to the average information age corresponding to the original access order, the access order of the two hovering points is restored; otherwise, the exchanged access order is retained.

Images