Energy scheduling and data collection method of intelligent reflector-assisted unmanned aerial vehicle

Abstract

The invention relates to the technical field of unmanned aerial vehicle wireless communication, and discloses an energy scheduling and data collection method of an intelligent reflecting surface assisted unmanned aerial vehicle, comprising the following steps: step 1, system deployment; step 2, dynamic clustering; step 3, channel modeling is carried out; step 4, joint optimization; step 5, scheduling energy; and step 6, data collection. According to the energy scheduling and data collection method of the intelligent reflector-assisted unmanned aerial vehicle, weighted clustering is carried out through an improved K-means algorithm of dynamic clustering, a cluster head node election mechanism and a re-clustering mechanism in combination with a sensor node geographic position and real-time residual energy, and a node with high residual energy and a central geographic position is preferentially selected as a cluster head; and when the cluster head energy is lower than the threshold value, re-clustering is triggered, so that the multi-hop transmission distance in the cluster is reduced, the node data acquisition and transmission energy consumption is reduced, the network energy management is improved, and the problem that the management is ignored in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of unmanned aerial vehicle wireless communication, in particular to an energy scheduling and data collection method for intelligent reflecting surface assisted unmanned aerial vehicle. BACKGROUND

[0002] In wireless rechargeable sensor networks (WRSN), mobile charging devices, especially unmanned aerial vehicles (UAVs), are widely used to provide energy replenishment and data collection services for energy-limited sensor nodes; existing solutions mainly optimize the UAV flight path, charging scheduling and data collection strategy to prolong the network lifetime; for example, some studies have proposed a charging request mechanism based on node energy state and a clustering data collection strategy, which reduces node energy consumption by reducing multi-hop communication, thereby improving the overall performance of the network.

[0003] As an emerging technology, intelligent reflecting surface (IRS) can enhance communication link quality by regulating the reflection characteristics of wireless signals; recent research has begun to explore the application of IRS in unmanned aerial vehicle communication, optimizing the IRS phase shift matrix to improve channel conditions and increase data transmission rate.

[0004] However, the existing technical solutions still have four core technical problems, which are difficult to meet the actual application requirements: Firstly, the existing technical solutions ignore the network energy management that determines the energy consumption of sensor nodes, for example, nodes usually consume a lot of energy to collect data and transmit them to the base station through multi-hop transmission; Secondly, the existing technical solutions ignore the limited battery capacity of sensor nodes and the energy consumption of data transmission, and sensor nodes often cannot be replenished with energy before they run out of energy; Furthermore, the existing technical solutions rarely consider the energy consumption and data delay of mobile chargers, and fail to fully consider the real-time and fairness of data collection; Finally, although IRS is an effective technology to enhance unmanned aerial vehicle communication, the existing technical solutions rarely apply IRS-UAV to WRSN, resulting in limited overall system performance; Therefore, an energy scheduling and data collection method for intelligent reflecting surface assisted unmanned aerial vehicle is proposed. SUMMARY

[0005] (I) Technical problems solved In view of the deficiencies of the prior art, the present application provides an energy scheduling and data collection method for intelligent reflecting surface assisted unmanned aerial vehicle, which has the advantages of dynamic clustering optimization, IRS assisted communication enhancement, UAV energy consumption and data delay joint optimization, etc., and solves the problems of poor energy management of sensor nodes, high data transmission delay, large energy consumption of mobile chargers and insufficient use of intelligent reflecting surface to enhance communication link quality in the prior art.

[0006] (II) Technical Solution To achieve the above-mentioned dynamic clustering optimization, IRS-assisted communication enhancement, and joint optimization of UAV energy consumption and data delay, the present application provides the following technical solution: An energy scheduling and data collection method for an intelligent reflecting surface assisted unmanned aerial vehicle, comprising the following steps: Step 1, system deployment: deploying 1 rotor UAV and 1 IRS in the WRSN, the IRS is carried on the UAV, the WRSN contains N sensor nodes; the initial position of the UAV coincides with the base station, and the UAV is responsible for supplementing energy and collecting data for the sensor nodes; Step 2, dynamic clustering: clustering and electing cluster head nodes using a clustering algorithm according to the geographical location and real-time residual energy of the sensor nodes; re-clustering is triggered when the residual energy of the cluster head node is lower than the threshold; Step 3, channel modeling: using a probabilistic line-of-sight channel model to characterize the communication links between the UAV and the sensor nodes, and between the UAV and the IRS, and calculating the LoS probability and channel gain; Step 4, joint optimization: constructing an objective function, and optimizing the IRS phase shift matrix, UAV flight path and speed, and sensor transmission power through an algorithm combining alternating optimization, semi-definite relaxation, and continuous convex approximation; the objective function is wherein is a weight coefficient, is a data delay penalty factor, is the charging energy consumption of the UAV, is the data collection energy consumption of the UAV, is the total energy consumption of the UAV; Step 5, energy scheduling: according to the optimization results, calculating the charging duration of the cluster head node according to wherein is the maximum capacity of the cluster head node, is the residual energy of the cluster head node at time t, is the charging rate of the cluster head node, is the energy consumption rate per unit time of the cluster head node; Step 6, data collection: using a time division multiple access protocol, the cluster head node aggregates the data within the cluster and transmits it to the UAV; the UAV receives the data through the IRS reflection link to ensure that the data transmission meets the signal-to-noise ratio threshold.

[0007] Preferably, the clustering algorithm in step 2 is an improved K-means algorithm, and the clustering process includes: Step 21. Initialize the number of cluster centers K, and randomly select K sensor nodes as initial cluster centers; Step 22. Calculate the weighted distance of each sensor node to each cluster center: wherein For geographical location weight, The remaining energy weight and , For sensor node coordinates, The coordinates of the cluster center, This represents the maximum battery capacity of the sensor node. The remaining energy of sensor node i; Step 23. Assign sensor nodes to the cluster with the smallest weighted distance, and update the cluster center coordinates and the set of nodes within the cluster; Step 24. Repeat steps 22 and 23 until the change in cluster center position is less than Output the clustering results; Step 25. When electing a cluster head node, the node with the highest remaining energy in the cluster and the central geographical location is selected first. The cluster head node is responsible for data aggregation and UAV communication within the cluster.

[0008] Preferably, the formula for calculating the LosS probability of the probabilistic line-of-sight channel model in step 3 is as follows: ,in Let a = 11.95 and b = 0.14 be the elevation angles of the UAV and the node, respectively, and these are environmental constants. The channel gain satisfies the following conditions under LoS state: NLoS status ,in As a reference distance of 1m, The path decay exponent. d is the additional attenuation factor for NLoS, and d is the communication link distance.

[0009] Preferably, the solution process of the hybrid algorithm in step 4 includes: Step 41. Initialize optimization variables: sensor communication scheduling matrix ( (Indicates whether sensor i communicates in time slot j), IRS phase shift matrix ( (The phase shift of the m-th reflecting unit), the UAV trajectory Q=[x(t),y(t),z(t)]; Step 42. Fix With Q, The relaxation is a continuous variable in the interval [0,1], and the communication scheduling subproblem is solved by linear programming. Step 43. Fix A and Q, and relax using the positive semidefinite relaxation technique. The unit modulus constraint transforms the non-convex problem into a semi-positive definite programming problem, and then the rank-1 solution is reconstructed by Gaussian randomization. Step 44. Fix A and The continuous convex approximation technique is used to perform a first-order Taylor expansion on the non-convex function in the UAV trajectory constraint, which is transformed into a convex problem to solve the horizontal and vertical trajectories. Step 45. Repeat steps 42 to 44 until the objective function converges. The convergence condition is that the difference between the objective functions in two iterations is less than 1. .

[0010] Preferably, the constraints on the objective function in step 4 include: ,in This represents the maximum energy capacity of the UAV. ,in The signal-to-interference-plus-noise ratio (SIR) of the sensor node and the UAV. For the sensor node's transmit power, For UAV-IRS channel gain, For IRS-sensor channel gain, Here, B is the IRS phase shift matrix, and B is the bandwidth. For additive white Gaussian noise power, The signal-to-noise ratio threshold; ,in Phase shift for IRS reflector unit; ,in For UAV flight speed, , These are the minimum and maximum flight speeds of the UAV, respectively.

[0011] Preferably, in step 5, the UAV is a cluster. Total service duration ,in Data transmission duration; total charging time of the UAV within one cycle. Where K is the number of clusters; total UAV flight time ,in , Let be the distance between cluster k and cluster k+1. The distance from the base station to cluster 1. Let K be the distance from cluster K to the base station.

[0012] Preferably, the formula for calculating the total energy consumption of the UAV in step 4 is as follows: ,in: These represent the blade profile power and inductive power during UAV hovering, respectively. The speed at the tip of the rotor blades. The average rotor induction speed during hovering. S represents the fuselage drag ratio, and S represents the rotor robustness. Where A is the air density and A is the rotor disk area; ; , UAV hovering power; , Power for UAV data collection.

[0013] Preferably, the data latency penalty factor in step 4 ,in The penalty coefficient is... UAV as the start of a cluster Charging time, For clusters Charging end time, For UAV from cluster The distance to u_k Here, cl represents the maximum latency of data packets for the cluster head node, and cl represents the storage capacity of the sensor node. s represents the data generation interval, and s represents the data packet size.

[0014] Preferably, in step 4, the sensor node transmit power is allocated using a water-filling algorithm, and the allocation formula is as follows: when hour, ; when hour, ; in Let be the channel gain of sensor node i. This represents the maximum transmit power of the sensor node.

[0015] Preferably, in step 4, the UAV trajectory also needs to satisfy the radius of curvature constraint. ,in Maximum acceleration of the UAV; UAV flight altitude constraint ,in These represent the minimum and maximum flight altitudes of the UAV, respectively.

[0016] (III) Beneficial Effects Compared with existing technologies, this invention provides a method for energy scheduling and data collection of intelligent reflective surface-assisted unmanned aerial vehicles (UAVs), which has the following beneficial effects: 1. This intelligent reflective surface-assisted UAV energy scheduling and data collection method uses an improved K-means algorithm for dynamic clustering, a cluster head node election mechanism, and a re-clustering mechanism. It combines the geographic location of sensor nodes and real-time remaining energy for weighted clustering, prioritizing nodes with high remaining energy and central geographic location as cluster heads. When the energy of a cluster head falls below a threshold, re-clustering is triggered, reducing multi-hop transmission distance within a cluster, lowering energy consumption for node data collection and transmission, improving network energy management, and solving the problem of existing technologies neglecting this management aspect.

[0017] 2. The energy scheduling and data collection method of this intelligent reflective surface-assisted UAV relies on the calculation of cluster head node charging time, planning of total UAV service time, and calculation mechanism of total charging time and flight time in the energy scheduling step. Based on the joint optimization results, it accurately matches the charging needs of cluster head nodes, and reasonably allocates the charging and data transmission time of UAV for each cluster, ensuring that sensor nodes are replenished in time before energy is exhausted, thus solving the problem of untimely node energy replenishment in the existing technology.

[0018] 3. The energy scheduling and data collection method of this intelligent reflective surface-assisted UAV utilizes a hybrid algorithm and objective function constructed through joint optimization steps, as well as a time-division multiple access protocol. The hybrid algorithm collaboratively optimizes the IRS phase shift matrix, UAV flight path and speed, and sensor transmission power. The objective function balances data latency penalty and UAV energy consumption. Combined with the time-division multiple access protocol to allocate transmission time slots, the method reduces data latency and controls the total energy consumption of the UAV, ensuring the real-time performance and fairness of data collection.

[0019] 4. This intelligent reflector-assisted UAV energy scheduling and data collection method is based on the integrated design of the intelligent reflector and UAV in system deployment, the probabilistic line-of-sight channel model for channel modeling, and the jointly optimized IRS phase shift matrix. The intelligent reflector is mounted on the UAV, and the communication link characteristics are accurately characterized by the probabilistic line-of-sight model. The phase shift of the reflector unit is optimized to enhance signal transmission, giving full play to the synergistic advantages of the intelligent reflector and the UAV. This fills the gap in the existing technology for this application in wireless rechargeable sensor networks and improves the overall system performance. Attached Figure Description

[0020] Figure 1 This is a flowchart of the energy scheduling and data collection method of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments and accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0022] Please see Figure 1 A method for energy scheduling and data collection of an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) includes the following steps: Step 1, System Deployment: Deploy one rotorcraft UAV and one IRS in the WRSN. The IRS is mounted on the UAV. The WRSN contains N sensor nodes. The initial position of the UAV coincides with the base station and is responsible for replenishing energy to the sensor nodes and collecting data. Step 2, Dynamic Clustering: Based on the geographic location and real-time remaining energy of the sensor nodes, a clustering algorithm is used to create clusters and elect cluster head nodes; when the remaining energy of the cluster head node is lower than the threshold, re-clustering is triggered. Step 3, Channel Modeling: The probabilistic line-of-sight channel model is used to characterize the communication links between UAV and sensor nodes, and between UAV and IRS, and the link Loss probability and channel gain are calculated; Step 4, Joint Optimization: Construct the objective function and optimize the IRS phase shift matrix, UAV flight path and velocity, and sensor transmit power using an algorithm that combines alternating optimization, semidefinite relaxation, and continuous convex approximation. The objective function is: ,in These are the weighting coefficients. As a data latency penalty factor, Energy consumption for charging UAVs Energy consumption for UAV data collection Total energy consumption of the UAV; Step 5, Energy Scheduling: Based on the optimization results, according to... Calculate the charging time of the cluster head node, where The maximum capacity of the cluster head node. The remaining energy of the cluster head node at time t. The charging rate for the cluster head node, The energy consumption rate per unit time of the cluster head node; Step 6, Data Collection: Using the Time Division Multiple Access (TDMA) protocol, the cluster head node aggregates the data within the cluster and transmits it to the UAV; the UAV receives the data through the IRS reflection link to ensure that the data transmission meets the signal-to-noise ratio threshold.

[0023] Example 1: This embodiment provides a basic scheme for energy scheduling and data collection of an intelligent reflective surface-assisted unmanned aerial vehicle (UAV). The complete execution steps, operational details, and technical features of this scheme are described in detail below: Step 1. System Deployment Steps: Deploy one rotorcraft UAV and one intelligent reflector IRS in the Wireless Rechargeable Sensor Network (WRSN). The IRS is mounted under the UAV fuselage via a mechanical fixing structure, ensuring that the IRS reflector unit faces the ground sensor node. The WRSN contains N=50 sensor nodes evenly distributed, with each node having a maximum battery capacity of 5000mAh. Initial remaining energy is randomly distributed between 1000mAh and 4500mAh. The UAV's initial position coincides with the base station, whose coordinates are set to 0,0,0. The UAV's maximum energy capacity is 10000mAh, and its maximum flight speed is... Minimum flight speed Flight altitude constraints to Between; the output power of the charging module on the UAV Data collection module power hovering power The IRS contains M=200 passive reflective elements, and the phase shift of each reflective element is adjustable in the range of [0, 2π].

[0024] Step 2. Dynamic Clustering: The improved K-means algorithm is used for clustering. The specific operation is as follows: Step 21: Initialize the number of cluster centers K=5. This value is determined based on the distribution density of sensor nodes to ensure that each cluster contains 8-12 nodes. Randomly select 5 nodes from N sensor nodes as initial cluster centers. When selecting, ensure that the initial cluster centers are evenly distributed within the WRSN coverage area and avoid concentration in a certain local area. Step 22: Calculate the weighted distance from each sensor node to each cluster center. The formula for calculating the weighted distance is: In the formula, This is the geographic location weight, with a value of 0.6; The remaining energy weight is set to 0.4, and ω1+ω2=1. This weight setting can balance the spatial clustering and energy balance of the clusters; xᵢ and yᵢ are the horizontal coordinates of the sensor nodes. , The horizontal coordinates of the cluster center; This represents the maximum battery capacity of the sensor node. The real-time remaining energy of sensor node i; Step 23: Assign each sensor node to the cluster with the smallest weighted distance, then update the cluster center coordinates. The new cluster center coordinates are the mean of the coordinates of all nodes within the cluster. Number of nodes within a cluster The number of nodes within the cluster is counted, and the set of nodes within the cluster is updated simultaneously. Step 24: Repeat steps 22 and 23. After each iteration, calculate the change in cluster center position, i.e., the Euclidean distance between the new cluster center and the old cluster center. Stop the iteration when the change in position of all cluster centers is less than 10⁻³ and output the clustering results. Step 25: Elect a cluster head node. Traverse the nodes in each cluster, select the top 3 nodes with the highest remaining energy, and calculate the average distance from these 3 nodes to all other nodes in the cluster. Select the node with the smallest average distance, i.e., the one with the central geographical location, as the cluster head node. The cluster head node is responsible for summarizing the data of each node in the cluster and performing data aggregation, and then communicating with the UAV. When the remaining energy of the cluster head node is lower than the threshold, re-clustering is triggered. The threshold is set to 20% of E_max, i.e., 1000mAh. Steps 21-25 are repeated when re-clustering.

[0025] Step 3. Channel Modeling Steps: The communication link is characterized using a probabilistic line-of-sight channel model. The specific operations are as follows: Calculate the Loss of Sight (LoS) probabilities between the UAV and sensor nodes, and between the UAV and the IRS. The formula for calculating the LoS probability is: In the formula, a=11.95 and b=0.14 are environmental constants; The elevation angle between the UAV and the node. Using the height z and horizontal distance of the UAV Calculation, the calculation method is as follows ; The formula for calculating channel gain in the Loss of Space (LoS) state is as follows: The formula for calculating channel gain in NLoS state is: In the formula, The channel gain at a reference distance of 1m is set to -30dB. This represents the path decay exponent in the Loss of Space (LoS) state. The path decay exponent in NLoS state; The additional attenuation factor in NLoS state; d is the spatial distance of the communication link, calculated as follows: , Let z be the horizontal distance between the UAV and the node, and z be the flight altitude of the UAV.

[0026] Step 4. Joint optimization step: Construct the objective function Minimize λ, where , The weighting coefficients balance data latency penalties and energy utilization. The objective function is solved using a hybrid algorithm, as follows: Step 41: Initialize the optimization variables. The sensor communication scheduling matrix A is an N×T matrix, and T is the number of time slots in the task cycle, which is set to 100. The initial value is randomly set to 0 or 1 to ensure that at most one sensor node communicates in each time slot; IRS phase shift matrix , The phase shift of the m-th reflecting unit is initially set randomly within the range of [0, 2π); the UAV trajectory Q = [x(t), y(t), z(t)] is initially set to start from the base station, pass above the horizontal projection points of each cluster head node in sequence, and the flight altitude z(t) = 100m; Step 42: Fix Θ and Q, and... The relaxation is a continuous variable in the interval [0,1]. A linear programming model is constructed with the objective function being to minimize λ. The constraints include the number of communication nodes in each time slot ≤ 1 and sensor transmission power constraints. The simplex method is used to solve the communication scheduling subproblem to obtain the optimal communication node allocation scheme for each time slot. Step 43: Fix A and Q, and use the semidefinite relaxation technique to relax the unit modulus constraint of Θ, that is... relaxation The non-convex problem is transformed into a semi-positive definite programming problem. The optimal solution of Θ is obtained by solving it using the interior point method. Then, the rank-1 solution is reconstructed using the Gaussian randomization method to ensure that Θ satisfies the unit modulus constraint. Step 44: Fix A and Θ, use the continuous convex approximation technique to perform a first-order Taylor expansion on the non-convex function in the UAV trajectory constraint, transform the non-convex constraint into a convex constraint, construct a convex optimization problem, and solve the horizontal trajectory x(t), y(t) and the vertical trajectory z(t); Step 45: Repeat steps 42-44, calculating the objective function value after each iteration. The objective function value is calculated when the difference between two iterations is less than 10⁻. 4 The iteration stops when the optimal IRS phase shift matrix Θ, UAV flight path Q, and sensor transmit power are output. .

[0027] Step 5. Energy scheduling step: Calculate the charging time of the cluster head node based on the joint optimization results. The formula for calculating the charging time is: In the formula, The maximum energy capacity of the cluster head node is set to 5000 mAh. Let be the remaining energy of the cluster head node at time t; The charging rate of the cluster head node is set to 10W. The energy consumption rate per unit time of the cluster head node is 2W; Simultaneously calculate UAV as a cluster Total service duration , Data transmission duration; total charging time of the UAV within one cycle. Total UAV flight time , The total flight distance of the UAV is calculated as follows: , Let k be the horizontal distance between cluster k and cluster k+1. The horizontal distance from the base station to cluster 1. Let K be the horizontal distance from cluster K to the base station.

[0028] Step 6. Data collection steps: Using the time division multiple access protocol, allocate data transmission time slots for each cluster head node according to the results of the communication scheduling matrix A. The cluster head node transmits the aggregated intra-cluster data to the UAV within the corresponding time slot. The UAV receives data via the IRS reflection link, and the signal-to-interference-plus-noise ratio is monitored in real time during the reception process. ,make sure , The signal-to-noise ratio threshold is set to 10dB.

[0029] In this embodiment: During system deployment, the integrated design of IRS and UAV achieves a combination of mobility and enhanced communication, expanding the communication coverage of UAV and sensor nodes. At the same time, the design that the initial position of UAV coincides with the base station facilitates energy replenishment after the mission, ensuring the continuous operation capability of UAV.

[0030] The dynamic clustering process improves the K-means algorithm by combining geographical location and remaining energy for weighted clustering, resulting in stronger spatial aggregation of nodes within the cluster, reducing the data transmission path length within the cluster, and lowering node energy consumption. The election criteria for cluster head nodes and the re-clustering mechanism ensure that cluster head nodes always have sufficient energy, avoiding cluster function failure, extending network lifetime, and solving the problem of poor energy management of sensor nodes.

[0031] The channel modeling stage employs a probabilistic line-of-sight model to accurately characterize the channel characteristics under different link states, providing reliable channel input for subsequent joint optimization, avoiding optimization biases caused by traditional idealized channel models, and improving the accuracy of communication link quality assessment.

[0032] The joint optimization stage decomposes the complex non-convex problem into multiple convex subproblems through a hybrid algorithm, efficiently solving for the optimal parameter configuration. This reduces data latency penalties while controlling the total energy consumption of the UAV, addressing the issues of high energy consumption and high data transmission latency in mobile chargers. The optimization of the IRS phase shift fully leverages the advantages of IRS-enhanced communication, improving the data transmission rate.

[0033] The energy scheduling process ensures timely energy replenishment for cluster head nodes by precisely calculating charging and service durations, preventing data loss or network interruptions caused by energy depletion. It also provides data for UAV energy planning. The data collection stage employs a time-division multiple access protocol to avoid communication interference, and the IRS reflection link enhances signal strength to ensure that data transmission meets the signal-to-noise ratio requirements, thereby improving data transmission reliability and efficiency and reducing delays caused by retransmissions.

[0034] Example 2: This embodiment further optimizes the dynamic clustering method, refining operational details to improve clustering performance. The specific operations are as follows: 1. Adaptive Determination of Cluster Core Quantity: The number of cluster cores K is adaptively adjusted based on the distribution density of sensor nodes. First, the density index of node distribution is calculated. The density index is calculated as follows: Density Index = Total Number of Nodes / Network Coverage Area. When the density index is ≥ 0.0015 nodes / m², K = 6; when the density index is between 0.001 and 0.0015 nodes / m², K = 5; when the density index is < 0.001 nodes / m², K = 4. This ensures that the number of nodes in each cluster is balanced, avoiding uneven energy consumption caused by clusters that are too large or too small.

[0035] 2. Optimized selection of initial cluster centers: A distance constraint is introduced when randomly selecting initial cluster centers to ensure that the distance between any two initial cluster centers is not less than 1 / K of the diagonal length of the network coverage area; the method for calculating the diagonal length of the network coverage area is as follows: This constraint can prevent the initial cluster centers from being too concentrated, which would lead to uneven clustering. In practice, the first cluster center is randomly selected first, and then a node from the remaining nodes that is not less than a set threshold distance from the first cluster center is selected as the second cluster center. This process is repeated until K initial cluster centers are selected.

[0036] 3. Dynamic adjustment of weighted distance weights: The geographical location weight ω1 and the remaining energy weight ω2 are dynamically adjusted according to the real-time network status. When the standard deviation of the remaining energy of the sensor nodes in the network is ≥1000mAh, it indicates that the energy difference between the nodes is large. At this time, the value of ω2 is increased to 0.5 and the value of ω1 is 0.5. When the standard deviation of the geographical location of the sensor nodes is ≥80m, it indicates that the nodes are scattered. At this time, the value of ω1 is increased to 0.7 and the value of ω2 is 0.3. When both are at a medium level, ω1=0.6 and ω2=0.4, so that the clustering strategy adapts to the real-time network status.

[0037] 4. Refined Calculation of Cluster Center Update: Node energy weights are introduced during cluster center update. The calculation method for the new cluster center coordinates is as follows: In the formula, The remaining energy of sensor node i; , Let i be the horizontal coordinate of sensor node i. This calculation method brings the cluster center closer to the energy-rich node, shortens the communication distance between nodes within the cluster and the cluster head node, and reduces data transmission energy consumption.

[0038] 5. Cluster head node backup mechanism: When electing a cluster head node, in addition to determining the primary cluster head node, the node with the second highest remaining energy and the second best geographical location is selected as the backup cluster head node; when the remaining energy of the primary cluster head node is lower than 30% of the threshold, i.e., 1500mAh, the backup cluster head node switching is initiated, and the primary cluster head node transmits the data aggregation results within the cluster to the backup cluster head node to avoid the energy consumption and delay caused by re-clustering.

[0039] In this embodiment: The adaptive determination of the number of cluster centers ensures that the clustering results match the node distribution density, avoiding excessive energy consumption for intra-cluster communication due to excessively large clusters or excessively long flight paths for UAVs due to excessively small clusters, thus further optimizing network energy allocation.

[0040] The distance constraint of the initial cluster center ensures the spatial balance of clustering, reduces the number of subsequent iterations, improves clustering efficiency, and avoids communication interference caused by excessive concentration of nodes in local areas.

[0041] The dynamic adjustment of the weighted distance weights enables the clustering strategy to respond to different network states. When there are large energy differences, it focuses on energy balance, and when the distribution is scattered, it focuses on spatial clustering, thus improving the flexibility and adaptability of clustering.

[0042] The refined calculation of cluster center updates brings the cluster center closer to energy-rich nodes, shortens intra-cluster communication distance, reduces node energy consumption, and improves the energy reliability of cluster head nodes.

[0043] The backup mechanism of the cluster head node avoids frequent re-clustering, reduces energy consumption and data transmission latency during the clustering process, ensures network service continuity, and further solves the problems of poor energy management and high data transmission latency of sensor nodes.

[0044] Example 3: This embodiment further refines the implementation steps for channel modeling and joint optimization algorithms to improve modeling accuracy and optimization efficiency. The specific operations are as follows: 1. Real-time calculation and updating of elevation angle: The UAV obtains its own position via GPS module during flight. and sensor node location Real-time calculation of horizontal distance , The calculation method is as follows And based on the horizontal distance and UAV flight altitude Calculate elevation angle , Updated every 0.1 seconds The value is then updated in real time to determine the Loss probability. This ensures the timeliness of channel state tracking.

[0045] 2. Calculation of the statistical average of channel gain: Considering the randomness of channel conditions, the statistical average of channel gain is calculated using the sliding window method. The sliding window size is set to 10 update cycles, i.e., 1 second. The calculation method for the statistical average of channel gain is as follows: In the formula, The channel gain within each update cycle; this calculation method can reduce the impact of channel fluctuations on subsequent optimizations and improve the robustness of the optimization results.

[0046] 3. Parallel computation optimization of hybrid optimization algorithm: In the joint optimization process, the communication scheduling subproblem, the IRS phase shift optimization subproblem and the UAV trajectory optimization subproblem adopt parallel computation method and are solved simultaneously on different computing cores. During the iteration process, intermediate results are exchanged through shared memory. This optimization can shorten the time of each iteration from 0.5 seconds to 0.2 seconds, improve optimization efficiency and meet the real-time scheduling requirements of UAV.

[0047] 4. Time slot optimization allocation for the communication scheduling subproblem: When solving the communication scheduling subproblem, a data urgency weight and a data delay penalty factor are introduced. Larger cluster head nodes are allocated more time slot resources; this ensures that urgent data is transmitted first, further reducing data latency and improving the real-time performance of data collection.

[0048] 5. Implementation of Curvature Radius Constraint for UAV Trajectory: The curvature radius constraint is explicitly defined in the UAV trajectory optimization stage. The calculation formula for the curvature radius constraint is as follows: In the formula, The flight speed of the UAV; The maximum acceleration of the UAV is set to 5 m / s². During implementation, a circular arc transition segment satisfying the radius of curvature constraint is inserted between the paths of the two cluster head nodes to avoid increased energy consumption and decreased stability caused by the UAV's sharp turning, thus ensuring a smooth trajectory.

[0049] In this embodiment: Real-time updates of elevation angle enable the Loss probability and channel gain to dynamically track the UAV's flight status, improving the timeliness and accuracy of channel modeling, providing real-time channel input for joint optimization, and avoiding optimization biases caused by static channel models.

[0050] The statistical averaging of channel gain reduces the impact of random channel fluctuations, making the optimization results more robust, avoiding over-adjustment caused by instantaneous channel fluctuations, and improving the stability of system operation.

[0051] Parallel computation optimization using hybrid optimization algorithms improves iteration efficiency, meets the real-time scheduling requirements of UAVs, ensures that optimization results are applied to actual flight and communication scheduling in a timely manner, and reduces performance loss caused by optimization delays.

[0052] The time slot optimization allocation of the communication scheduling subproblem prioritizes the transmission of urgent data, further reducing data latency, improving the real-time performance of data collection, and solving the problem of high data transmission latency.

[0053] The curvature radius constraint of the UAV trajectory ensures a smooth flight path, reduces the extra energy consumption caused by sharp turns, improves UAV flight stability, extends UAV lifespan, and solves the problem of high energy consumption of mobile chargers.

[0054] Example 4: This embodiment details the implementation of system constraints, UAV total energy consumption calculation, and sensor transmit power allocation to ensure stable system operation and optimized resource allocation. The specific operations are as follows: 1. Real-time verification and adjustment of constraints: UAV Total Energy Constraint: UAV Total Energy Constraint Must meet ≤ , The maximum energy capacity of the UAV is 10000mAh, or 37Wh. Real-time computing ,when achieve When the 90% threshold, or 33.3Wh, is reached, the UAV flight speed is adjusted from... =25m / s decreased to 20m / s; when achieve When the charge reaches 95% (35.15Wh), the charging time for each cluster is shortened by 80% of the original duration. Ensure that total energy consumption does not exceed the maximum value; Signal-to-noise ratio constraint: signal-to-interference-plus-noise ratio Must meet , The signal-to-noise ratio threshold is 10dB; real-time monitoring ,when When the signal strength is less than 10 dB, adjust the IRS phase shift matrix Θ to increase the reflected signal strength, or increase the transmit power of the sensor node. and not exceeding until ≥10dB; IRS phase shift constraint: phase shift of the IRS reflector Must satisfy 0≤ <2π, the phase shift of each reflective unit is monitored in real time by the IRS controller. When the phase shift exceeds the range, the bias voltage is automatically adjusted to pull the phase shift back to the [0,2π) range; UAV speed constraint: UAV flight speed Must meet ≤ ≤ The flight control system uses speed sensors to provide real-time speed feedback; when the speed is below a certain threshold... When the speed is 5m / s, increase the motor output power; when the speed is higher than... Reduce motor output power when the speed is 25m / s.

[0055] 2. Detailed calculation of total UAV energy consumption: flight The formula for calculating flight energy consumption is: ,in The flight power of a UAV is calculated as follows: In the formula, =10W is the blade profile power when the UAV is hovering; =8W is the sensing power when the UAV is hovering; =100m / s is the rotor blade tip velocity; V0=5m / s is the average rotor induction velocity during hovering; =0.05 is the fuselage drag ratio; =1.2kg / m³ is the air density; S=0.1m² is the rotor robustness; A=0.5m² is the rotor disk area; This represents the total flight time of the UAV. Charging energy consumption The formula for calculating charging energy consumption is: , Total charging time for UAV; Hovering energy consumption The formula for calculating hovering energy consumption is: , =8W is the hovering power of the UAV. UAV as a cluster Total service duration; Data collection energy consumption The formula for calculating energy consumption for data collection is: , =5W is the data collection power of the UAV. UAV as a cluster The time it takes to collect data.

[0056] 3. Algorithm for allocating sensor transmit power using a water-filling algorithm: First, calculate the channel gain for each sensor node. Right now Then judge and The size relationship, among which The power is additive white Gaussian noise and is -80 dBm, i.e., 10⁻ 8 W, =10dB, which is 10. =0.5W is the maximum transmit power of the sensor node; when The formula for calculating the transmit power of a sensor node is as follows: when At that time, the sensor node transmit power Ensure that the signal strength meets the signal-to-noise ratio requirements; The transmit power is recalculated every 0.5 seconds after allocation. This ensures that power allocation is always adapted to the current channel conditions.

[0057] In this embodiment: Real-time verification and adjustment of constraints ensure system stability, UAV total energy consumption constraints prevent UAVs from interrupting their missions due to energy depletion, signal-to-noise ratio constraints ensure data transmission reliability, and IRS phase shift constraints and UAV speed constraints ensure the safe operation of hardware devices, thus solving the problem of limited system performance.

[0058] The refined calculation of total UAV energy consumption accurately quantifies the composition of various energy consumption components, providing precise energy consumption input for joint optimization, making the optimization objectives more realistic. At the same time, through energy consumption composition analysis, the largest proportion of flight energy consumption can be reduced in a targeted manner, further improving energy utilization.

[0059] The water-filling algorithm for sensor transmit power allocation enables power configuration on demand. Nodes with good channel conditions use lower power, while nodes with poor channel conditions use maximum power. This minimizes sensor node energy consumption and extends node lifespan while meeting signal-to-noise ratio requirements, solves the problem of poor energy management of sensor nodes, avoids power waste, and improves overall energy utilization.

[0060] In summary, this intelligent reflective surface-assisted UAV energy scheduling and data collection method utilizes an improved K-means algorithm for dynamic clustering, a cluster head node election mechanism, and a re-clustering mechanism. It combines the geographic location of sensor nodes with real-time remaining energy for weighted clustering, prioritizing nodes with high remaining energy and centrally located positions as cluster heads. Re-clustering is triggered when the cluster head energy falls below a threshold, reducing intra-cluster multi-hop transmission distance, lowering energy consumption for node data collection and transmission, improving network energy management, and addressing the issue of existing technologies neglecting this management aspect. Furthermore, this intelligent reflective surface-assisted UAV energy scheduling and data collection method relies on the cluster head node charging time calculation, UAV service total duration planning, and total charging time and flight time accounting mechanism in the energy scheduling steps. Based on the joint optimization results, it accurately matches the cluster head node charging needs and rationally allocates the charging and data transmission time of the UAV for each cluster, ensuring that the sensor nodes are replenished in time before their energy is exhausted, thus solving the problem of untimely node energy replenishment in the existing technology.

[0061] Furthermore, this intelligent reflective surface-assisted UAV energy scheduling and data collection method utilizes a hybrid algorithm with joint optimization steps, a time-division multiple access protocol (TDMA) to construct an objective function, and collaboratively optimizes the IRS phase shift matrix, UAV flight path and speed, and sensor transmission power through the hybrid algorithm. The objective function balances data latency penalties and UAV energy consumption, and the TDMA allocates transmission time slots to reduce data latency and control the total UAV energy consumption, ensuring real-time and fair data acquisition.

[0062] Furthermore, this intelligent reflector-assisted UAV energy scheduling and data collection method, based on the integrated design of the intelligent reflector and UAV in system deployment, the probabilistic line-of-sight channel model for channel modeling, and the jointly optimized IRS phase shift matrix, mounts the intelligent reflector on the UAV. By accurately characterizing the communication link characteristics through the probabilistic line-of-sight model, it optimizes the phase shift of the reflector unit to enhance signal transmission, fully leverages the synergistic advantages of the intelligent reflector and the UAV, fills the gap in the application of this technology in wireless rechargeable sensor networks, improves the overall system performance, and solves the problems of poor energy management of sensor nodes, high data transmission latency, high energy consumption of mobile chargers, and failure to fully utilize the intelligent reflector to enhance the quality of communication links in existing technical solutions.

[0063] The relevant modules involved in this system are all hardware system modules or functional modules that combine computer software programs or protocols with hardware in the prior art. The computer software programs or protocols involved in these functional modules are technologies known to those skilled in the art and are not improvements to this system. The improvement of this system lies in the interaction or connection between the modules, that is, in improving the overall structure of the system to solve the corresponding technical problems that this system aims to address.

[0064] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for energy scheduling and data collection of an intelligent reflective surface-assisted unmanned aerial vehicle (UAV), characterized in that, Includes the following steps: Step 1, System Deployment: Deploy one rotorcraft UAV and one IRS in the WRSN. The IRS is mounted on the UAV. The WRSN contains N sensor nodes. The initial position of the UAV coincides with the base station and is responsible for replenishing energy to the sensor nodes and collecting data. Step 2, Dynamic Clustering: Based on the geographic location and real-time remaining energy of the sensor nodes, a clustering algorithm is used to create clusters and elect cluster head nodes; when the remaining energy of the cluster head node is lower than the threshold, re-clustering is triggered. Step 3, Channel Modeling: The probabilistic line-of-sight channel model is used to characterize the communication links between UAV and sensor nodes, and between UAV and IRS, and the link Loss probability and channel gain are calculated; Step 4, Joint Optimization: Construct the objective function and optimize the IRS phase shift matrix, UAV flight path and velocity, and sensor transmit power using an algorithm that combines alternating optimization, semidefinite relaxation, and continuous convex approximation. The objective function is: ,in These are the weighting coefficients. As a data latency penalty factor, Energy consumption for charging UAVs Energy consumption for UAV data collection Total energy consumption of the UAV; Step 5, Energy Scheduling: Based on the optimization results, according to... Calculate the charging time of the cluster head node, where The maximum capacity of the cluster head node. The remaining energy of the cluster head node at time t. The charging rate for the cluster head node, The energy consumption rate per unit time of the cluster head node; Step 6, Data Collection: Using the Time Division Multiple Access (TDMA) protocol, the cluster head node aggregates the data within the cluster and transmits it to the UAV; the UAV receives the data through the IRS reflection link to ensure that the data transmission meets the signal-to-noise ratio threshold.

2. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The clustering algorithm in step 2 is an improved K-means algorithm, and the clustering process includes: Step 21. Initialize the number of cluster centers K by randomly selecting K sensor nodes as initial cluster centers; Step 22. Calculate the weighted distance from each sensor node to each cluster center: in For geographical location weight, The remaining energy weight and , For sensor node coordinates, The coordinates of the cluster center are... This represents the maximum battery capacity of the sensor node. The remaining energy of sensor node i; Step 23. Assign sensor nodes to the cluster with the smallest weighted distance, and update the cluster center coordinates and the set of nodes within the cluster; Step 24. Repeat steps 22 and 23 until the change in cluster center position is less than Output the clustering results; Step 25. When electing a cluster head node, the node with the highest remaining energy in the cluster and the central geographical location is selected first. The cluster head node is responsible for data aggregation and UAV communication within the cluster.

3. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The formula for calculating the LoS probability of the probabilistic line-of-sight channel model in step 3 is as follows: ,in Let a = 11.95 and b = 0.14 be the elevation angles of the UAV and the node, respectively, and these are environmental constants. The channel gain satisfies the following conditions under LoS state: NLoS status ,in As a reference distance of 1m, The path decay exponent. d is the additional attenuation factor for NLoS, and d is the communication link distance.

4. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The solution process of the hybrid algorithm in step 4 includes: Step 41. Initialize optimization variables: sensor communication scheduling matrix ( (Indicates whether sensor i communicates in time slot j), IRS phase shift matrix ( (The phase shift of the m-th reflecting unit), the UAV trajectory Q=[x(t),y(t),z(t)]; Step 42. Fix With Q, The relaxation is a continuous variable in the interval [0,1], and the communication scheduling subproblem is solved by linear programming. Step 43. Fix A and Q, and relax using the positive semidefinite relaxation technique. The unit modulus constraint transforms the non-convex problem into a semi-positive definite programming problem, and then the rank-1 solution is reconstructed by Gaussian randomization. Step 44. Fix A and The continuous convex approximation technique is used to perform a first-order Taylor expansion on the non-convex function in the UAV trajectory constraint, which is transformed into a convex problem to solve the horizontal and vertical trajectories. Step 45. Repeat steps 42 to 44 until the objective function converges. The convergence condition is that the difference between the objective functions in two iterations is less than 1. .

5. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The constraints on the objective function in step 4 include: ,in This represents the maximum energy capacity of the UAV. ,in The signal-to-interference-plus-noise ratio (SIR) of the sensor node and the UAV. For the sensor node's transmit power, For UAV-IRS channel gain, For IRS-sensor channel gain, Here, B is the IRS phase shift matrix, and B is the bandwidth. For additive white Gaussian noise power, The signal-to-noise ratio threshold; ,in Phase shift for IRS reflector unit; ,in For UAV flight speed, , These are the minimum and maximum flight speeds of the UAV, respectively.

6. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, In step 5, UAV is a cluster. Total service duration ,in Data transmission duration; total charging time of the UAV within one cycle. Where K is the number of clusters; total UAV flight time ,in , Let be the distance between cluster k and cluster k+1. The distance from the base station to cluster 1. Let K be the distance from cluster K to the base station.

7. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, The formula for calculating the total energy consumption of UAV in step 4 is as follows: ,in: These represent the blade profile power and inductive power during UAV hovering, respectively. The speed at the tip of the rotor blades. The average rotor induction speed during hovering. S represents the fuselage drag ratio, and S represents the rotor robustness. Where A is the air density and A is the rotor disk area; ； , UAV hovering power; , Power for UAV data collection.

8. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, Data delay penalty factor in step 4 ,in The penalty coefficient is... UAV as the start of a cluster Charging time, For clusters Charging end time, For UAV from cluster The distance to u_k Here, cl represents the maximum latency of data packets for the cluster head node, and cl represents the storage capacity of the sensor node. s represents the data generation interval, and s represents the data packet size.

9. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, In step 4, the sensor node transmission power is allocated using a water-filling algorithm, and the allocation formula is as follows: when hour, ; when hour, ; in Let be the channel gain of sensor node i. This represents the maximum transmit power of the sensor node.

10. The energy scheduling and data collection method for an intelligent reflective surface-assisted unmanned aerial vehicle (UAV) according to claim 1, characterized in that, In step 4, the UAV trajectory also needs to satisfy the radius of curvature constraint. ,in Maximum acceleration of the UAV; UAV flight altitude constraint ,in These represent the minimum and maximum flight altitudes of the UAV, respectively.

Images