Multi-uav assisted wireless sensor network data collection method

Abstract

The present application relates to a kind of multi-unmanned aerial vehicle assisted wireless sensor network data acquisition method, belong to wireless sensor network technical field, comprising the following steps: S1: the maximum data acquisition time of UAV is obtained by dichotomy search method, and within the given time T, using time equal interval discrete technology is dispersed into a finite number of points with unmanned track;S2: according to initialization multi-UAV trajectory algorithm, solve UAV initial trajectory;S3: by the relaxation of binary variable, the original mixed integer non-convex problem is converted into continuous non-convex problem, and based on alternate optimization algorithm, the original problem is decomposed into two subproblems, by alternating iteration until convergence to obtain results, wherein, continuous convex approximation algorithm is used to solve subproblem, obtain UAV trajectory, uplink energy transmission scheduling and downlink data transmission scheduling;Until dichotomy search method result converges, the maximum time of UAV data acquisition is obtained.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless sensor networks, and relates to a multi-unmanned aerial vehicle (UAV) assisted wireless sensor network data collection method. BACKGROUND

[0002] With the fusion of the fifth generation mobile communication technology, the Internet of Things (IoT) and the Internet technology, thousands of IoT devices are applied in industry and social life, and the information exchange amount of a wireless sensor network (WSN) increases exponentially, which poses a great challenge to unmanned aerial vehicle (UAV) data collection. In addition, the ubiquitous nature and demand diversity of IoT devices also bring many technical problems to data collection in IoT. The low power consumption and long endurance requirement of IoT sensors are one of the key factors to ensure the stability and sustainability of the system. IoT sensors are usually used to detect specific changes in the environment, and then transmit the collected information to a data receiver for further processing. In a WSN, sensors upload data through multi-hop, which increases the information transmission delay and energy consumption. In addition, the sensors are constrained by battery energy storage, which limits the processing capacity and quality of service of the WSN. Although replacing the battery or charging can prolong the service life of the sensor, when the number of sensors is large, it is difficult to charge or replace the battery for the sensors. Therefore, improving the service life of the WSN has become an important research problem. SUMMARY

[0003] Therefore, the purpose of the application is to provide a multi-UAV assisted wireless sensor network data collection method by introducing UAV into the WSN. The method is equipped with a wireless power transmission (WPT) module for the UAV, and the UAV transmits energy to the sensor before collecting the data of the sensor. The sensor uploads data using the energy transmitted by the UAV and is used for subsequent sensing of new data. The UAV not only completes the data collection task, but also provides energy supply for the sensors in the WSN, further prolonging the service life of the WSN.

[0004] To achieve the above purpose, the application provides the following technical scheme:

[0005] A multi-UAV assisted wireless sensor network data collection method, comprising the following steps:

[0006] S1: obtaining the maximum data collection time of the UAV by using the dichotomy search method, and discretizing the unmanned track into a finite number of points by using the time equal interval discretization technology within a given time T;

[0007] S2: solving the initial UAV trajectory according to the initialized multi-UAV trajectory algorithm;

[0008] S3: converting the original mixed integer non-convex problem into a continuous non-convex problem by relaxing the binary variable, and decomposing the original problem into two sub-problems based on an alternating optimization algorithm, and obtaining the result by alternating iteration until convergence, wherein a continuous convex approximation algorithm is used to solve the sub-problems to obtain the UAV trajectory, uplink energy transmission scheduling and downlink data transmission scheduling; until the result of the dichotomy search method converges, the maximum time of UAV data collection is obtained.

[0009] Further, the step S1 specifically comprises: setting a loop to solve the maximum data collection time of the UAV in the outermost layer of the whole method; initializing T1=0 and T2, wherein T2 takes a large enough value, and the maximum data collection time T of the UAV is given as (T1+T2) / 2 at the beginning of each loop. The continuous UAV trajectory is discretized by using time equidistant discretization technology.

[0010] Further, the step S2 specifically comprises the following steps:

[0011] S21: the positions of the K sensors are given, and the order and time T of the UAV accessing the sensors are obtained by using the MTSP method mTSP , and an arbitrary time T is given;

[0012] S22: judging whether T≥T mTSP , if yes, it means that the UAV can fly over the head of each sensor, and the initial trajectory of the UAV is obtained by step S23, otherwise, step S24 is entered;

[0013] S23: the task completion time of the UAV is T, wherein the UAV u m spends time T mb to fly over the associated sensor, and the remaining time T-T mb is evenly distributed in G mb , during which the UAV hovers above the sensor, wherein the hovering time of the UAV at the hovering position is T-T

[0014]

[0015] T≥T mb is obtained

[0016] S24: the UAV cannot fly over the associated G mbabove the middle sensor, thus, a circular region with radius r is determined centering each sensor, by reasonably designing the UAV trajectory and radius r, the flight time of UAV is minimized; and according to the UAV flying over the circular region of the middle sensor in a given time range T, the UAV trajectory is designed mb above the middle sensor, thus, a circular region with radius r is determined centering each sensor, by reasonably designing the UAV trajectory and radius r, the flight time of UAV is minimized; and according to the UAV flying over the circular region of the middle sensor in a given time range T, the UAV trajectory is designed m The optimization problem of flight time is expressed as:

[0017]

[0018]

[0019] From the above formula, it can be seen that the objective function value does not change with r; therefore, by fixing r, the UAVu m hovering point is obtained and the optimal radius r of the given time range T is obtained by using the bisection search method, and the UAV initial trajectory of the case of T≤T mb is obtained

[0020] Further, step S3 specifically comprises the following steps:

[0021] S31: the UAV initial trajectory is obtained by using the above initialization UAV trajectory algorithm and setting the algorithm convergence threshold ε2;

[0022] S32: in the case of given UAV trajectory {q m [n]}, and introducing the relaxation variable {γ m,k [n]}, the problem is converted into a convex problem, and the energy transmission scheduling {α m,k [n]} of UAV and the data transmission scheduling {β m,k [n]} of sensor are obtained by solving the CVX toolbox;

[0023] S33: the UAV trajectory {q m,k [n]} is optimized under the given energy transmission scheduling {α m,k [n]} and data transmission scheduling {β m [n]}; by using the first-order Taylor expansion, the non-convex constraint is converted into a convex constraint, and the UAV trajectory {q m [n]} is obtained by using the continuous approximation method;

[0024] S34: it is judged whether the result converges, if yes, step S35 is entered, otherwise, step S32 is returned;

[0025] S35: the UAV trajectory, the energy transmission scheduling of UAV and the data transmission scheduling of sensor are obtained.

[0026] The present application has the beneficial effect that the present application can significantly reduce the maximum data acquisition time of the UAV, and prolong the service life of the WSN.

[0027] Other advantages, objects, and features of the present application will be better understood from the following specification taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to make the purposes, technical solutions and advantages of the present application clearer, the preferred detailed description of the present application will be combined with the drawings as follows, wherein:

[0029] Figure 1 A multi-UAV data acquisition system model diagram of the present application;

[0030] Figure 2 A multi-UAV trajectory optimization algorithm flowchart of the present application;

[0031] Figure 3 A multi-UAV WSN data acquisition algorithm flowchart of the present application;

[0032] Figure 4 A flight trajectory diagram of the UAV in the single-UAV data acquisition system of the present application;

[0033] Figure 5 A sensor wake-up time diagram in the single-UAV data acquisition system of the present application;

[0034] Figure 6 A flight trajectory diagram of the UAV in the multi-UAV data acquisition system of the present application;

[0035] Figure 7 A sensor wake-up time diagram in the multi-UAV data acquisition system of the present application;

[0036] Figure 8 A comparison diagram of the UAV number of the algorithm based on dichotomy and SCA and the GA, NIA, MACO and RSA algorithms affecting the maximum task completion time of the UAV;

[0037] Figure 9 A comparison diagram of the sensor number of the algorithm based on dichotomy and SCA and the GA, NIA, MACO and RSA algorithms affecting the maximum task completion time of the UAV. DETAILED DESCRIPTION

[0038] The present application can also be embodied in a different specific form, and the various details of the specification can be modified in various ways without departing from the spirit of the application. It should be noted that the drawings provided in the following examples are only intended to illustrate the basic concept of the present application, and the following examples and features in the examples can be combined with each other without conflict.

[0039] The accompanying drawings are only intended to illustrate, and are only schematic diagrams, not physical diagrams, and should not be understood as a limitation of the present application; in order to better illustrate the embodiments of the present application, some components in the drawings are omitted, enlarged or reduced, and do not represent the actual size of the product; for those skilled in the art, it is understandable that some well-known structures and their descriptions in the drawings can be omitted.

[0040] The same or similar reference numerals in the drawings of the embodiments of the present application correspond to the same or similar components; in the description of the present application, it should be understood that if the terms "upper", "lower", "left", "right", "front", "back" and the like indicate the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, therefore the terms describing the positional relationship in the drawings are only used for illustrative purposes, and should not be understood as a limitation of the present application, for those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0041] For the data collection problem in multi-UAV assisted WSN, the energy is transmitted to the sensor before the UAV collects the sensor data, then the sensor uploads the data to the UAV using the received energy, and ensures that the energy is sufficient before the next UAV visit. As shown in Figure 1 , in the WSN, M UAVs collect the data of K sensors on the ground. The set of UAVs is represented as M={u m ,1≤m≤M}, and the set of sensors is represented as K={SN k ,1≤k≤K}, where the kth sensor is represented as SN k , whose coordinates are w k =(x k ,y k ,0), and the data size to be collected from SN k is S k . Assuming that the flight height of all UAVs is H, according to the method of three-dimensional Euclidean coordinates, the position of UAV u m at time t is represented as um (t)=(x m (t),y m (t),H), where 0≤t≤T m After completing their data acquisition tasks, all UAVs return to the starting point, which in practice serves as a workstation for charging or unloading UAV data acquisition data.

[0042] Assuming UAVu m The time to complete the data acquisition task is T. m This method proposes a UAV data acquisition method based on TDMA, where the UAV acquires data in orthogonal frequency channels. Since the UAV trajectory is a continuous function of time, involving a large number of optimization variables, this method utilizes a time-interval discretization method to divide the time range [0, T] into segments. m Discretize into N equal time slots, i.e., T m =Nδ t , where δ t Let represent the time slot length, and let its value be small enough that the distance between the UAV and the sensor remains approximately constant within each time slot. The continuous-time UAV trajectory is discretized into N time slots, and the coordinates of the UAV in the nth time slot are {q}. m The expression `(n), 1≤n≤N}` is represented as:

[0043] q m [n]=(x[n],y[n],H),n∈{1,2,…,N}

[0044] UAVs have a maximum flight speed:

[0045]

[0046] In the above formula, V max This represents the UAV's maximum flight speed. After collecting data, the UAV returns to its starting point. In practical applications, after completing its mission, the UAV needs to return to its work platform to recharge and prepare for the next mission. Define q. I If the starting position of the UAV is given, then:

[0047]

[0048] In the system model, each sensor in the WSN data acquisition system goes through four stages: charging stage, data transmission stage, sleep stage, and data sensing stage. These four stages will be described in detail below.

[0049] Step 1: Charging stage, UAVu m Proximity sensor SN k Awaken it and supply it. The size of the energy, where UAVum To the sensor SN k The energy delivered is:

[0050]

[0051] In the above formula, is the time of UAV transmitting energy, represents the sensor SN k receiving the power of UAV u m , wherein d m,k [n] = (H 2 + || q m [n] - w k 2 ) 12 represents the distance between UAV u m and the sensor SN k in the n time slot; P U represents the constant transmission power of the UAV; β0represents the channel power gain when the distance is 1m.

[0052] In the time T m , the UAV adopts the TDMA transmission protocol. Assuming that the WSN is a sleep-wakeup mechanism, the UAV wakes up and charges the sensor when it approaches the sensor in the n time slot, which is defined as a binary variable:

[0053]

[0054] In the above formula, α m,k [n] = 1 indicates that the sensor SN k is woken up by the UAV u m at the n time, and the UAV transmits energy to the sensor, while determining the association relationship between the UAV and the sensor. On the contrary, the UAV does not transmit energy to the sensor SN k , that is, α m,k [n] = 0. This method adopts the TDMA transmission protocol, and the sensor can only receive the energy transmitted by one UAV at the same time. This constraint is represented as:

[0055]

[0056] Step 2: Data transmission phase, the sensor receives the energy provided by the UAV for subsequent uploading of data and sensing of new data. After the UAV transmits energy to the sensor, the sensor can upload data, which is defined as a binary variable:

[0057]

[0058] In the above formula, β m,k [n] = 1 indicates that the UAV u mSensor SN is collected at time n. k The data. Conversely, UAVu m Do not collect sensor SN k The data, namely β m,k [n] = 0. Since the UAV uses the TDMA protocol to collect data, the sensor can only upload data to one UAV at a time, as shown below:

[0059]

[0060] Therefore, sensor SN k The energy consumption for transmitting data is:

[0061]

[0062] In the above formula, P k and P c The transmit power and circuit power when the sensor sends data.

[0063] Step 3: Sleep stage. The sensor enters sleep mode, thereby reducing energy consumption. The energy consumed at this time is E. S Because of E S compared to and It is relatively small, not even on the same order of magnitude, and therefore can be ignored.

[0064] Step 4: Data sensing stage, sensor SN k The energy consumed by sensing data from the surrounding environment through the application is in, When the UAV visits again, the sensed data is uploaded to the UAV. Therefore, the energy consumed by the UAV to transmit this part of the energy is:

[0065]

[0066] In summary, the sensor's energy consumption can be divided into two parts: the first part is uploading data; the second part is sensing data. The total energy that the UAV needs to transmit to the sensor is:

[0067]

[0068] In the above formula, This indicates the charging time of the sensor by the UAV.

[0069] In four steps, step 2 can be carried out after step 1 starts, but the energy consumption of the sensor must be less than the energy transmitted by the UAV, step 3 can be interrupted by step 4, and step 4 suspends its sleep phase to sense new data. For step 2 and step 3, the UAV in this method uses a TDMA transmission protocol, and the UAV downlink energy transmission and uplink data collection are implemented in the same frequency band, but the time is orthogonal, that is, in a time slot, the UAV can only downlink energy transmission or uplink data collection, then:

[0070]

[0071] In this method, the data collection uses a rotor UAV, and the energy consumption includes propulsion energy consumption Communication energy consumption and energy consumption of WPT module transmission energy Since the UAV performs uplink data collection, the communication energy consumption of the UAV can be ignored. Therefore, the UAV energy consumption includes the energy consumption of UAV flight and WPT module charging. By discretizing the UAV flight time at equal intervals, the flight energy consumption model of UAVu m is:

[0072]

[0073] Assuming that the set of sensors served by UAVu m is G m , that is, |G m | is the number of sensors served by UAVu m . The energy consumed by the UAV to transmit energy to the sensor is

[0074]

[0075] Assuming that the maximum energy that can be stored by the battery of UAVu m is and the energy that can be stored by each UAV is equal. Each UAV needs to fly back to the starting point before the energy is exhausted:

[0076]

[0077] To ensure the continuous operation of the wireless sensor, the energy constraint of each sensor needs to be considered, so that the energy consumption of the sensor transmitting data is less than the energy transmitted by the UAV, and the sensor needs to transmit data after receiving energy. Therefore, for the energy received by the kth sensor and the energy consumption of transmitting data, the following constraints exist:

[0078]

[0079] Assuming that the set of sensors served by UAVu m is SNk The channel between them is a Loss-of-Stake (LoS) link channel, so UAVu at time n can be obtained. m With sensor SN k The channel gain can be modeled based on free-space path loss as follows:

[0080]

[0081] Therefore, at time n, the sensor SN k To UAVu m The data upload rate is expressed as:

[0082]

[0083] In the above formula, B is the channel bandwidth, σ 2 This represents the noise power at the UAV receiver. In WSN, each sensor needs to upload a file of size S. k Data volume:

[0084]

[0085] The optimization problem of this method is formulated as optimizing the UAV flight trajectory {q} m (n)}、UAV energy transmission scheduling{α m,k (n)}、Sensor data transmission scheduling {β m,k Let (n)} minimize the maximum task completion time for all UAVs. Assume the size S of the data uploaded by each sensor is... k and sensor total energy consumption constraints All are fixed; the specific optimization problem is described as follows:

[0086]

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093]

[0094]

[0095]

[0096]

[0097]

[0098]

[0099] In the above formula, N = T m / δ t Constraint C1 indicates that the sensor needs to upload S k size data; constraint C2 indicates that the energy transmitted by the UAV to the sensor is greater than the energy consumption of the sensor; constraint C3 indicates the energy consumption constraint of the sensor when transmitting data; constraint C4 indicates the total energy consumption constraint of the UAV; constraint C5 and constraint C6 respectively indicate the UAV downlink energy transmission scheduling and the sensor data transmission scheduling; constraint C7 indicates that the UAV downlink transmission energy and the uplink data collection cannot be performed at the same time; constraint C8 indicates that the sensor can only receive energy from one UAV at the same time; constraint C9 indicates that the sensor can only transmit data to one UAV at the same time; constraint C10 indicates that the energy received by the sensor at a certain time needs to be greater than the energy consumed; constraint C11 indicates the maximum flight speed constraint of the UAV; constraint C12 indicates that the UAV needs to return to the initial position after completing the task.

[0100] Problem P1 minimizes the maximum task completion time of all UAVs by optimizing the UAV trajectory {q m [n]}, UAV energy transmission scheduling {α m,k [n]} and sensor data transmission scheduling {β m,k [n]}. Wherein, the UAV flight trajectory {q m [n]} is a function of time slot n, and the UAV energy transmission scheduling {α m,k [n]} and the sensor data transmission scheduling {β m,k [n]} are binary variables. In addition, constraint C1 and constraint C2 are non-convex constraints. Therefore, problem P1 is a mixed integer non-convex problem, which is difficult to solve directly and needs to be solved by the following steps:

[0101] Step 1: Since the UAV energy transmission scheduling and the sensor data transmission scheduling in problem P1 are integer constraints, the binary variables in constraints C5 and C6 can be relaxed to continuous variables, and the continuous non-convex optimization problem P2 is obtained by substituting problem P1:

[0102]

[0103] s.t.C1-C4,C7-C12

[0104]

[0105]

[0106] Step 2: Introduce a slack variable η, given an arbitrary time length T, satisfying T m ≤T, and substituting the slack variable η into problem P2, we get problem P3:

[0107]

[0108] stC3,C4,C7-C14

[0109]

[0110]

[0111] C17:T m ≤T

[0112] Step 3: In problem P3, for any given UAV task completion time T, define the optimal value of problem P3 as η. * (T), where η * (T) is a function of time T. It can be seen that for any given UAV task completion time T, the condition is met if and only if η * (T)≥1, UAV not only transmits to the sensor It has the energy and can complete the data acquisition task. Therefore, problem P2 is equivalent to problem P4:

[0113]

[0114] stC18:η * (T)≥1

[0115] Step 4: Since problem P3 is a non-decreasing function, problem P4 can be solved effectively by using a binary search method to find the solution time T until constraint C18 is satisfied. Therefore, the difficulty in solving problem P1 lies in finding the optimal solution to problem P3 given time T. Since there always exists an optimal solution to problem P3 that satisfies constraint C18, problem P3 can be rewritten as problem P5:

[0116]

[0117] stC3,C4,C7-C14

[0118]

[0119]

[0120] In the above formula, N = T / N. Although time T is fixed in problem P5, constraint C15 still contains highly coupled variables {α}.m,k [n]} and {q m [n]} in constraint C16, constraint C16 is equivalent to constraint C17: m,k [n]} and {q m [n]} are highly coupled, and constraint C10 is a causal non-convex constraint. Therefore, the original problem is decomposed into two sub-problems for alternating optimization, i.e., fixing the UAV flight trajectory {q m [n]} to optimize the energy transmission schedule {a m,k [n]} and the data transmission schedule {b m,k [n]}, then fixing the energy transmission schedule {a m,k [n]} and the data transmission schedule {b m,k [n]} to optimize the flight trajectory {q m [n]}.

[0121] Step 5: An alternating optimization algorithm is proposed to decompose the problem P5 into two sub-problems for solving, which are step 6 and step 7, respectively.

[0122] Step 6: Given the UAV trajectory {q m [n]}, the energy transmission schedule {a m,k [n]} of the UAV and the data transmission schedule {b m,k [n]} of the sensor are optimized. By introducing the slack variable {g m,k [n]}, the problem P5 is equivalent to the problem P6:

[0123]

[0124] s.t.C3,C4,C7-C9,C11-C14

[0125]

[0126]

[0127]

[0128] From the above formula, in the optimal solution of the problem P6, the constraint C20 inequality needs to be strictly satisfied, which can be achieved by reducing the energy transmission schedule {a m,k [n]} and the data transmission schedule {b m,k [n]} to make the constraint C20 inequality hold. Under the condition of satisfying the constraint C20 inequality constraint, the problem P6 has an optimal solution. Therefore, the problem P6 given the UAV trajectory {q m [n]} is equivalent to the problem P5. At this time, it can be verified that all the constraints of the problem P6 are convex constraints, and the problem P6 is a convex optimization problem, which can be solved by the CVX toolbox to obtain the energy transmission schedule {a m,k[n]} and data transmission schedule {β m,k [n]}.

[0129] Step 7: Given the energy transmission schedule {a m,k [n]} and data transmission schedule {β m,k [n]} to optimize the UAV trajectory {q m [n]}. At this time, both constraint C15 and constraint C16 are non-convex constraints. In the lth iteration, given the local point The first-order Taylor expansion of constraint C15 and constraint C16 are respectively:

[0130]

[0131]

[0132] In the above formula, A k,r [n] and are respectively:

[0133]

[0134]

[0135] In the above formula, are respectively:

[0136]

[0137] For given any local point and the above lower bound expression, the original problem can be converted into problem P7:

[0138]

[0139] s.t.C3,C4,C7-C9,C11-C16

[0140]

[0141]

[0142] Step 8: Since problem P7 is a convex problem, it can be solved using standard convex optimization tools such as CVX. The UAV trajectory obtained by solving the second sub-problem is returned as a constant value to solve the first sub-problem P6, and the alternating iteration optimization is performed until it converges to a certain precision. The specific alternating optimization algorithm is shown in Algorithm 1.

[0143]

[0144] Step 9: For a given arbitrary time T, design the initial trajectory of UAV u for Algorithm 1 by solving the Min-Max Multiple Traveling Salesman Problem (MTSP) for UAV u m The flight time is T mb and UAV u m visits the set of sensors G mb . For a given arbitrary time T, whether T is greater than T mb There are two different cases, so two cases are proposed to solve the Min-Max MTSP problem to solve the initial trajectory:

[0145] (1) For the case of T≥T mb , the task completion time of UAV u is T, where UAV u m spends time T mb flying to the above of the associated sensors, and the remaining time T-T mb is evenly distributed among the sensors in G mb , during which time the UAV hovers above the sensors, where the hovering time of UAV u at the hovering location

[0146]

[0147] According to the UAV visiting order G mb and the hovering time distribution , the initial trajectory of UAV u for the case of T≥T mb can be obtained

[0148]

[0149] (2) For the case of T mb , the UAV cannot fly to the above of the sensors in G mb within the specified time T, so a circular region with a radius r is determined around each sensor, and by reasonably designing the UAV trajectory and the radius r, the flight time of the UAV is minimized. And according to the circular region that flies over the associated sensors in G mb within the given time range T, the initial trajectory of the UAV is designed. Given the radius r, the optimization problem of minimizing the flight time of UAV u m can be expressed as:

[0150]

[0151]

[0152] As shown in the above equation, the objective function value of problem P8 does not change with r. Therefore, UAVu can be obtained by solving problem P8 with r fixed. m hovering point The optimal radius r for a given time range T is obtained using a binary search method, and the initial trajectory of the UAV is calculated. The specific process of initializing the UAV trajectory algorithm is summarized in Algorithm 2.

[0153]

[0154]

[0155] Thus, the initialization of the UAV trajectory is complete. The overall algorithm is summarized as follows: First, time T is solved using binary search; second, Algorithm 2 is used to solve for the initial UAV trajectory at a given time T; finally, Algorithm 1 is used to alternately optimize the UAV flight trajectory {q}. m [n]}、Sensor data transmission scheduling {β m,k [n]} and UAV energy transfer scheduling {α m,k [n]} Solve problem P5 and iterate until the result converges. The overall algorithm process is summarized in Algorithm 3.

[0156]

[0157] The simulation considers multiple UAVs collecting data from K=20 sensors within a region. It assumes all sensors have equal transmit power, and each UAV is equipped with a WPT module with identical transmission power. Furthermore, it assumes all sensors need to upload the same amount of data, and the energy consumed in data transmission is fixed. Each UAV's battery is fully charged before the task begins, and the energy storage is consistent.

[0158] like Figure 4 The image shown is in Figure 4 The wake-up time of the sensors includes the time for UAV downlink transmission energy and the time for UAV uplink data acquisition. Figure 5 It can be seen that each sensor is in a sleep state most of the time, and can only be woken up when the UAV approaches it. This further verifies the TDMA data acquisition method, indicating that the UAV can only serve one sensor at a time. Furthermore, channel quality is a crucial factor in system performance. When the UAV is close to the sensor, the channel between the UAV and the sensor is good. Since both the UAV's power transmission and the sensor's data transmission power are fixed, the sensor's wake-up time is short, meaning less time is spent on downlink power transmission by the UAV and data transmission by the sensor. Conversely, when the UAV is far from the sensor, the channel quality is poor, leading to increased time for both UAV power transmission and sensor data transmission, resulting in higher energy consumption for both the sensor and the UAV.

[0159] Further study the multi-UAV assisted WSN data collection scenario, such as Figure 6 The scenario of WSN and Figure 4 is the same as the scenario of M = 3 data collection system, and each sensor needs to upload the same size of data. Simulation results show that the multi-UAV data collection system reduces the flight distance and time of UAV compared with the single UAV scenario. As shown in Figure 7 , at the same time, at most three sensors are woken up because the frequency bands used by each UAV are orthogonal. Compared with the single UAV scenario, the maximum UAV task completion time of the multi-UAV system in the proposed method is reduced by 57.4%, achieving the expected goal. Similar to Figure 5 , when the UAV is close to the sensor, a better communication channel can be obtained, reducing the wake-up time of the sensor. Conversely, the UAV is far away from the sensor, which will result in an increase in the wake-up time of the sensor.

[0160] To further demonstrate the performance gain of the proposed method, for the WSN data collection scenario with K = 20 sensors, the relationship between the maximum UAV task completion time and the number of dispatched UAVs is studied. Four different methods are compared with the proposed method, which are: Greedy Algorithm (GA), Nearest Insertion Algorithm (NIA), Modified Ant Colony Optimization Algorithm (MACO) and Random Service Algorithm (RSA). As shown in Figure 8 , the maximum task completion time of the UAV decreases with the increase of the number of UAVs M, verifying that multiple UAVs can complete the data collection task in a shorter time. It can be seen that the GA algorithm, the NIA algorithm and the MACO algorithm are better than the RSA algorithm, but they are slightly worse than the proposed method. In the MACO algorithm method, the UAV is in hover collection mode, and the flight speed of the UAV is fixed, resulting in a waste of a lot of time. In addition, as described in the second chapter, the UAV energy consumption model shows that the hover mode of the UAV is not the lowest energy consumption time, so the MACO algorithm consumes more energy of the UAV. In the proposed method, the UAV flight trajectory is a smooth curve, which can collect data in the process of flight, thereby reducing the time of UAV data collection. Therefore, with the increase of the number of UAVs, the proposed method can reduce the maximum task completion time of the UAV to a greater extent.

[0161] As shown in Figure 9As shown, further study the influence of the number of sensors in WSN on the maximum task completion time of UAVs in the case of a given number of UAVs M = 3, the maximum task completion time of UAVs increases with the increase of the number of sensors. Similarly, compared with GA and NIA, MACO and RSA algorithms, when the number of sensors is less than 20, the results obtained by GA and NIA and MACO algorithms are close to the proposed method. However, when the scale of WSN is larger, the algorithm is obviously superior to GA, NIA, RSA and MACO algorithms, and the proposed method can reduce the maximum task completion time of all UAVs to a greater extent.

[0162] Finally, it should be pointed out that the above embodiments are only used to illustrate the technical solutions of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solutions of the present application can be modified or replaced by equivalents without departing from the spirit and scope of the present application, and they should be covered in the scope of the claims of the present application.

Claims

1. A method for data acquisition in a wireless sensor network assisted by multiple unmanned aerial vehicles (UAVs), characterized in that: The method comprises the following steps: S1: initialize data collection time , set a large enough value , and set the algorithm convergence threshold ε ; S2: calculate the current maximum data acquisition time and initialize multi-UAV trajectory; S3: cyclically performing the following steps until the result converges: fixing the UAV trajectory, optimizing the UAV downlink energy transmission scheduling and uplink data collection scheduling; fixing the UAV downlink energy transmission scheduling and uplink data collection scheduling, and optimizing the UAV trajectory; S4: judge whether the number of collection is satisfied, if yes, let , update the trajectory of UAV, downlink energy transmission scheduling and uplink data collection scheduling, and let the current T be the maximum data collection time; If not, then let ; S5: cyclically execute steps S2-S4 until is satisfied, output the current UAV trajectory, uplink energy transfer schedule, downlink data transfer schedule, and maximum data collection time T .

2. The multi-UAV assisted wireless sensor network data collection method of claim 1, wherein: The step S2 calculates the current maximum data acquisition time , initializing the multi-unmanned aerial vehicle trajectory, specifically comprising the following steps: S21: Given K the position of each sensor, the order and time of the UAV accessing the sensors are solved by using the minimum maximum multi-traveling salesman problem (MTSP) ; S22: According to T whether greater than There are two different cases, respectively solve the minimum maximum MTSP problem, solve the initial trajectory: for In the case of UAV Time spent Fly above the associated sensors, remaining time exist The sensors are evenly distributed among them, For UAV Access the sensor set, during which time the UAV hovers above the sensors, where the UAV... In relation to the c Hovering position of each sensor when collecting data and transmitting energy hover time for: According to the UAV access order and hover time allocation , obtain UAV initial trajectory of the case ; For the case of , a circular region of radius r is determined around each sensor, and the UAV trajectory and radius r are designed to minimize the UAV flight time; and the initial trajectory of the UAV is designed to fly over the circular region of the associated T sensor within ; given the radius r , the optimization problem to minimize the UAV flight time is formulated as: wherein represents the maximum flight speed of the UAV; By fixing r Solving for UAV hovering point The maximum data acquisition time was obtained by using a binary search method. T optimal radius r Seeking Initial trajectory of the UAV in the situation . 3.The multi-UAV-assisted wireless sensor network data collection method of claim 1, wherein: Step S3 specifically comprises the following steps: S31: based on the initial trajectory of the UAV , introduce the relaxation variable , solve the energy transmission scheduling of the UAV by the CVX toolbox and the data transmission scheduling of the sensor ; S32: optimize the UAV trajectory given the energy transfer schedule and the data transfer schedule ; solve the UAV trajectory using a continuous approximation method ;​ S33: solving the UAV trajectory from step S32 As a constant value replaces the initial trajectory in step S31 , alternately iterated optimization until convergence to a certain accuracy; S34: obtaining the UAV trajectory, the energy transmission scheduling of the UAV, and the data transmission scheduling of the sensor.

Images