A Dynamic Optimization Method for Updating Complex State Information of the Internet of Things Assisted by Unmanned Aerial Vehicles

Abstract

The present invention discloses a dynamic optimization method for updating complex state information of an Internet of Things (IoT) assisted by unmanned aerial vehicles (UAVs). In each time slot during the operation of the system, the UAV group collects the real-time positions of IoT devices. According to the real-time positions of the IoT devices, all UAVs obtain the optimization results by successively executing the method proposed by the present invention and perform dynamic deployment and resource allocation according to this result. In the method proposed by the present invention, considering the joint consideration of performance and computational complexity, the sub-problems of the transmission and calculation start times are split and the optimal solutions of the two split problems are obtained, ultimately realizing real-time and effective complex state information update. The optimization method proposed by the present invention establishes a dynamically autonomous UAV group in a distributed optimization manner, enabling the UAV group to perform efficient dynamic deployment and resource allocation according to the real-time positions of IoT devices, thereby realizing effective complex state information update in the UAV-assisted IoT under the nomadic movement scenario of IoT devices.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication status information update, and particularly to a dynamic optimization method for complex status information update assisted by an unmanned aerial vehicle (UAV) in the Internet of Things (IoT). Background Art

[0002] In recent years, status update applications such as intelligent driving and intelligent agriculture have attracted great attention in the academic and industrial fields. Since the freshness of status information is of great significance to the status update process, S.K. Kaul et al. defined the age of information (AoI) (S.K. Kaul, R.D. Yates, and M. Gruteser, “Real-time status: How often should one update?” in Proc. IEEE INFOCOM, 2012, pp. 2731–2735.) as a performance metric to evaluate the freshness of status information. However, the status information of some applications needs to be obtained through complex calculations of raw data. Therefore, the present invention defines the status update process in which status information needs to be obtained through complex calculations of raw data as complex status information update. However, IoT devices with limited energy can hardly afford the energy consumption generated during complex status information update when performing tasks. At the same time, the target areas of some applications are not single, and IoT devices will move nomadically to sense multiple target areas, that is, complete the sensing tasks of multiple target areas in the way of moving - sensing - moving. Facing the above two types of problems, the UAV equipped with a computing server can help the nomadically moving IoT devices complete complex status information update by virtue of its flexibility. Specifically, when the IoT device stays in the target area to perform complex status information update, the UAV hovers in the air to provide computing services for the IoT device. When the position of the IoT device changes, the UAV flies to the next hovering position and provides computing services for the IoT device. Therefore, the present invention aims to improve the performance of UAV-assisted IoT complex status information update in dynamic scenarios.

[0003] Implementing complex state information updates in UAV-assisted Internet of Things (IoT) requires considering three performance metrics. First, the average age of information can reflect the average freshness during the complex state information update process and can characterize the performance of real-time state update applications with moderate time sensitivity. Second, due to the long duration of some applications, the energy consumption of energy-constrained IoT devices is an important performance metric. Third, since UAVs are also energy-limited devices, their energy consumption is also an important performance metric. The above three metrics are affected by the resource allocation strategy. Specifically, the transmission start time of IoT devices and the calculation start time of UAVs during the complex state information update process directly affect the average age of information, and the hovering position of UAVs affects all three performance metrics simultaneously. Therefore, to optimize the three performance metrics, the above strategies need to be jointly optimized. However, solving the joint optimization problem of the above strategies requires overcoming the following challenges. First, the mobility of IoT device locations. Since the locations of IoT devices move continuously over time and their movement patterns are difficult to detect for UAVs, the optimization problem can only be carried out under the condition that the location information is not fully known. Second, the dynamics of complex state information updates. The transmission start time of IoT devices and the calculation start time of UAVs are coupled and highly variable, and there is a correlation between two consecutive complex state information updates, making the optimization problem have intractable Markov properties. Third, the large dimensionality of the UAV position space. The position space composed of the combinations of multi-UAV positions is relatively large, greatly increasing the complexity of solving the problem. Therefore, how to effectively jointly optimize multiple strategies in a dynamic scenario is a difficult problem. Summary of the Invention

[0004] The object of the present invention is to provide a dynamic optimization method for complex state information update in UAV-assisted IoT for the nomadic movement scenario of IoT devices, optimizing multiple strategies to achieve efficient complex state information update in UAV-assisted IoT in a dynamic scenario.

[0005] The technical solution for achieving the object of the present invention is: to provide a dynamic optimization method for complex state information update in UAV-assisted IoT. The method includes the following steps:

[0006] Step 1: The IoT device transmits real-time location information to the UAV, and all UAVs exchange their current positions and the location information of the IoT devices collected. Assume the total number of UAVs is M and all UAVs are numbered;

[0007] Step 2: Establish a dynamic optimization problem for UAV-assisted complex state information update;

[0008] Step 3: The first UAV generates an initial feasible solution {t 0 , c 0 , q 0}, where t 0 is the set of starting times for feasible IoT device transmissions, and c 0 is the set of starting times for feasible UAV calculations, and q 0 is the set of feasible UAV positions; meanwhile, let the value of variable m be 1;

[0009] Step 4: The m-th UAV obtains all actions from the action set {ascend, descend, move left, move right, move forward, move backward, stay still} that satisfy the constraints of the dynamic optimization problem in Step 2, and let the set consisting of all actions of the m-th UAV that satisfy the constraints of the dynamic optimization problem in Step 2 be and let the total number of elements in the set be Meanwhile, let the value of variable a be 1;

[0010] Step 5: The m-th UAV generates an initial feasible solution according to the a-th behavior in and determines the connection relationship between IoT devices and UAVs using the location-based matching game method, where is the set of feasible starting times for IoT device transmissions generated by the m-th UAV according to the a-th behavior, is the set of feasible starting times for UAV calculations generated by the m-th UAV according to the a-th behavior;

[0011] Step 6: Obtain the optimal solutions for the starting times of transmissions and calculations in the first to the (U - 1)-th complex state information updates, where U is the total number of complex state information updates;

[0012] Step 7: Obtain the starting time of transmission and the starting time of calculation

[0013] in the optimized U-th state update and obtain the objective function value under the optimization solution where the value range of variable u is all positive integers less than U, and the value range of variable is all positive integers not greater than M excluding the current value of variable m, is the starting time of transmission of the optimized m-th UAV at the u-th complex state information update, is the initial -th UAV's starting time of transmission at the u-th complex state information update, is the starting time of calculation of the optimized m-th UAV at the u-th complex state information update, is the initial The calculation start time of the No. u unmanned aerial vehicle (UAV) at the u-th complex state information update; Let the value of variable a increase by 1;

[0014] Step 9: Return to Step 5 until the value of variable a is greater than

[0015] Step 10: Denote The behavior number corresponding to the minimum value in as a * , and obtain the position of the m-th UAV after executing the a-th * behavior Replace with q m-1 the position of the m-th UAV in, and let the set of UAV positions after replacement be q m , where q m-1 is the initial UAV position at the m-th optimization;

[0016] Step 11: The m-th UAV flies from the current position to the optimized position and broadcasts q to all other UAVs m ; Let the value of variable m increase by 1;

[0017] Step 12: Return to Step 4 until the value of variable m is greater than the total number of UAVs M;

[0018] Step 13: All UAVs perform transmission start time and calculation start time settings according to , where is the set of transmission start times after the M-th optimization, is the set of calculation start times after the M-th optimization;

[0019] Step 14: Return to Step 1 until the auxiliary tasks of the UAVs end;

[0020] Compared with the prior art, the remarkable advantages of the present invention are as follows: (1) A dynamic optimization method is adopted to address the online optimization problem brought about by the mobility of Internet of Things (IoT) devices, enabling real-time and effective response to the changing positions of IoT devices; (2) An optimization architecture is established based on the analysis of the characteristics of the complex state information update process to solve the dynamic optimization problem in the complex state information update process with low complexity; (3) A greedy method is adopted to solve the problem of the huge dimensionality of the UAV position combination space, and a multi-UAV deployment strategy with good performance is obtained with low complexity; (4) The dynamic optimization method adopted by the present invention only requires multi-UAVs to autonomously execute the optimization program without an additional centralized controller, and the interaction information between UAVs and between UAVs and IoT devices is less, thus saving system overhead; BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1It is the dynamic model diagram of the drone-assisted Internet of Things device of the present invention;

[0022] Figure 2 It is the schematic diagram of the sum of the average age of information of the present invention changing with the number of drone antennas;

[0023] Figure 3 It is the schematic diagram of the sum of the energy consumption of the Internet of Things devices of the present invention changing with the number of drone antennas;

[0024] Figure 4 It is the schematic diagram of the sum of the energy consumption of the drones of the present invention changing with the number of drone antennas;

[0025] Figure 5 It is the schematic diagram of the sum of the weighted values of three performance indicators of the present invention changing with the number of drone antennas. Detailed implementation manners

[0026] Figure 1 It is the dynamic model diagram of the drone-assisted Internet of Things device of the present invention. In the drone-assisted Internet of Things, multiple Internet of Things devices move nomadically in time slots. Specifically, at the beginning of a time slot, each Internet of Things device moves towards its respective target point. When all Internet of Things devices reach the target point, the drone group performs real-time deployment according to the dynamic optimization method proposed by the present invention to assist the Internet of Things devices in updating complex state information. At the same time, the complex state information update tasks of all Internet of Things devices at the current target point need to be completed within this time slot so that the Internet of Things devices can perform a new round of complex state information update in the next time slot.

[0027] The embodiment of the present invention provides a dynamic optimization method for updating complex state information of a drone-assisted Internet of Things, and the method includes the following steps:

[0028] Step 1: The Internet of Things devices transmit real-time position information to the drones, and all drones exchange their current positions and the position information of the Internet of Things devices collected. Assume the total number of drones is M and all drones are numbered.

[0029] Step 2: Establish a dynamic optimization problem for updating complex state information assisted by drones. The optimization problem is as follows:

[0030]

[0031]

[0032]

[0033]

[0034]

[0035]

[0036]

[0037]

[0038]

[0039]

[0040] Among them, \(Z\) is the set of optimization variables, and are the weight factors of the average information age, the energy consumption of IoT devices, and the energy consumption of drones respectively, is the number of IoT devices served by the \(m\)-th drone, \(t\) u,m and \(t\) u+1,m are the transmission start times of the IoT devices served by the \(m\)-th drone at the \(u\)-th and \((u + 1)\)-th status updates respectively, and is the transmission delay of the \(k\)-th IoT device served by the \(m\)-th drone, \(\tau\) c is the coherence interval, \(D\) k,m is the task data volume of the \(k\)-th IoT device served by the \(m\)-th drone, \(B\) m is the communication bandwidth allocated to the \(m\)-th drone, \(p\) k,m is the transmit power of the \(k\)-th IoT device served by the \(m\)-th drone, is the channel coefficient between the \(m\)-th drone and the \(k\)-th IoT device it serves, \(g_0\) is the received power when the transceiver distance is 1 meter, \(d\) k,m is the distance between the \(m\)-th drone and the \(k\)-th IoT device it serves, \(L\) is the number of drone antennas, is the pilot transmit power of the \(i\)-th IoT device served by the \(m\)-th drone, \(g\) i,m is the channel coefficient between the \(m\)-th drone and the \(i\)-th IoT device it serves, is the Rice factor between the \(m\)-th drone and the \(i\)-th IoT device it serves, \(A_1\) and \(A_2\) are constants, \(\theta\) i,m is the elevation angle between the \(m\)-th drone and the \(i\)-th IoT device it serves, is the pilot transmit power of the \(k\)-th IoT device served by the \(m\)-th drone, \(\Omega\) m is a diagonal matrix and the element on the \(k\)-th main diagonal is \([\Omega\) m kk \(= K\)​k,m , K k,m is the Rice factor between the m-th unmanned aerial vehicle (UAV) and the k-th Internet of Things (IoT) device it serves, I m is the identity matrix of, is a matrix and the element in its l-th row and k-th column is θ k,m is the elevation angle between the m-th UAV and the k-th IoT device it serves, and are the set of complex state information update times and the set of UAVs respectively, c u,m and c u+1,m are the starting calculation times of the m-th UAV for the u-th and (u + 1)-th complex state information updates respectively, is the calculation delay of the k-th IoT device served by the m-th UAV, N ue is the upper limit of the number of IoT devices that a UAV can serve, C k,m is the number of CPU rotations required for the m-th UAV to complete the task of the k-th IoT device it serves, F is the computing power of the UAV, t 1,m is the starting transmission time of the IoT device served by the m-th UAV for the first state update, is the time taken for all IoT devices to move from the previous target point to the current target point, TF m is the flight time of the m-th UAV, c U,m is the starting calculation time of the m-th UAV at the U-th state update, T Z is the time slot length, and are the three-dimensional position coordinates of the m-th UAV after and before taking an action respectively, d1 and d2 are selectable flight distances, QS is the set of all allowed UAV hovering positions, and are the three-dimensional position coordinates of the x-th and y-th UAVs after taking an action respectively;

[0041] In Equation (1), the expression for the average age of information of the k-th IoT device served by the m-th UAV is as follows:

[0042]

[0043] where, c 1,m is the starting calculation time of the m-th UAV for the first complex state update, t u-1,m is the starting transmission time of the IoT device served by the m-th UAV for the (u - 1)-th complex state update, t U-1,mis the transmission start time of the Internet of Things device for the m - numbered drone service at the (U - 1) - th complex state update;

[0044] In Equation (1), the energy consumption of the k - th Internet of Things device for the m - numbered drone service to complete a task offloading is:

[0045]

[0046] In Equation (1), the energy consumption of the m - numbered drone in one time slot is:

[0047]

[0048] where c1, c2, c3, and c4 are constants related to the physical characteristics of the drone and the environment, P H is the hovering power of the drone, and are the position coordinates of the m - numbered drone before and after performing the action, respectively.

[0049] Step 3: The No. 1 drone generates an initial feasible solution {t 0 , c 0 , q 0} according to the constraint conditions of the dynamic optimization problem described in Step 2, where t 0 is the set of feasible transmission start times of the Internet of Things devices, c 0 is the set of feasible calculation start times of the drone, q 0 is the set of feasible drone positions; meanwhile, let the value of the variable m be 1.

[0050] Step 4: The m - numbered drone obtains all actions that satisfy the constraint conditions of the dynamic optimization problem in Step 2 from the action set {ascend, descend, move left, move right, move forward, move backward, stay still}, and let the set composed of all actions of the m - numbered drone that satisfy the constraint conditions of the dynamic optimization problem in Step 2 be and let the total number of elements in the set be Meanwhile, let the value of the variable a be 1.

[0051] Step 5: The m - numbered drone generates an initial feasible solution according to the a - th action in and determines the connection relationship between the Internet of Things devices and the drone by using the location - based matching game method, where, is the set of feasible transmission start times of the Internet of Things devices generated by the m - numbered drone according to the a - th action, is the set of feasible calculation start times of the drone generated by the m - numbered drone according to the a - th action; Step 5 specifically includes Steps 5.1 to 5.2;

[0052] Step 5.1: Generate an initial feasible solution: When the a-th behavior is to remain stationary, let where t m-1 is the set of initial transmission start times of IoT devices in the m-th optimization, and c m-1 is the set of initial calculation start times of the drone in the m-th optimization; when the a-th behavior is not to remain stationary, the m-th drone randomly generates an initial feasible solution according to the a-th behavior

[0053] Step 5.2: Establish connection relationships: The m-th drone establishes a distance-based preference relationship table for each IoT device and the drone based on the feasible drone positions generated by the a-th behavior; for IoT devices, any IoT device starts sending connection requests from the drone with a larger preference value in the preference relationship table. When the connection request is rejected by the drone, it continues to send connection requests to the next drone with a smaller preference value in the list until it is not rejected; for drones, any drone judges the IoT devices sending connection requests according to the preference list and the maximum number of IoT devices it can serve; when the number of IoT devices it serves is less than the maximum service quantity, the connection request is allowed; when the number of IoT devices it serves is already equal to the maximum service quantity and the preference value of the drone for the IoT device currently making a connection request is greater than the preference value of the already connected IoT devices, the connection request is received and the connection of the IoT device with a smaller preference value is rejected; when the number of IoT devices it serves is already equal to the maximum service quantity and the preference value of the drone for the IoT device currently making a connection request is less than the preference value of the already connected IoT devices, the connection request is rejected.

[0054] Step 6: Obtain the optimal solutions of the transmission start times and calculation start times in the first to the (U - 1)-th complex state information updates, where U is the total number of complex state information updates. Step 6 specifically includes Steps 6.1 to 6.4;

[0055] Step 6.1: Establish the following equivalent optimization problem:

[0056]

[0057]

[0058]

[0059]

[0060] where t 1,m is the transmission start time of the IoT device served by the m-th drone in the first complex state update, tu,m is the transmission start time of the Internet of Things device for the m - numbered drone service at the u - th complex state update, t u-1,m is the transmission start time of the Internet of Things device for the m - numbered drone service at the (u - 1)-th complex state update, t U-1,m is the transmission start time of the Internet of Things device for the m - numbered drone service at the (U - 1)-th complex state update is the transmission start time of the U - th complex state information update for all Internet of Things devices of the m - numbered drone service generated according to the a - th behavior is the calculation start time of the m - numbered drone at the U - th complex state information update generated according to the a - th behavior;

[0061] Step 6.2: Solve the optimal solution of the equivalent optimization problem using the Lagrange multiplier method where the integer - type variable u ranges from 1 ≤ u ≤ U;

[0062] Step 6.3: According to obtain where the variable u in ranges from 1 ≤ u ≤ U, are the optimal values of the equivalent variables corresponding to the 1 - st, u - th, and U - th complex state information updates of the m - numbered drone respectively, is the transmission start time of the m - numbered drone at the 1 - st complex state information update after optimization, is the transmission start time of the m - numbered drone at the u - th complex state information update after optimization, is the transmission start time of the m - numbered drone at the (u - 1)-th complex state information update after optimization, is the transmission start time of the m - numbered drone at the (U - 1)-th complex state information update after optimization.

[0063] Step 6.4: Let where the integer - type variable u ranges from 1 ≤ u ≤ U, is the calculation start time of the m - numbered drone at the u - th complex state information update after optimization.

[0064] Step 7: Obtain the transmission start time and the calculation start time in the optimized U - th state update. Step 7 specifically includes Steps 7.1 to 7.3;

[0065] Step 7.1: Let

[0066] Step 7.2: Let the lower and upper bounds of c U,m be and And define the function Wherein, is the starting moment of the (U - 1)-th complex state information update calculation for the optimized m-th unmanned aerial vehicle,

[0067] Step 7.3: When g(c lb ) = 0, then let When g(c ub ) = 0, then let When g(c lb ) < 0 and g(c ub ) > 0, then use the bisection method to obtain the solution c U,m that makes g(c * ) = 0, and let When g(c lb ) > 0, then let When g(c ub ) < 0, then let

[0068] Step 8: Obtain the optimal solution according to the objective function of the optimization problem in Step 2 The objective function value under Wherein, And The value range of the variable u is all positive integers less than U, and the variable has a value range of all positive integers not greater than M excluding the current value of the variable m, is the transmission starting moment of the optimized m-th unmanned aerial vehicle during the u-th complex state information update, is the initial -th unmanned aerial vehicle's transmission starting moment during the u-th complex state information update, is the calculation starting moment of the optimized m-th unmanned aerial vehicle during the u-th complex state information update, is the initial -th unmanned aerial vehicle's calculation starting moment during the u-th complex state information update; let the value of the variable a increase by 1.

[0069] Step 9: Return to Step 5 until the value of the variable a is greater than

[0070] Step 10: Denote The behavior number corresponding to the minimum value in as a * , and obtain the position of the m-th unmanned aerial vehicle after executing the a * -th behavior Replace with the position of the m-th unmanned aerial vehicle in q m-1 , and let the replaced unmanned aerial vehicle position set be q m , wherein, q m-1is the initial UAV position at the m-th optimization;

[0071] Step 11: The m-th UAV flies from the current position to the optimized position and broadcasts q to all other UAVs m ; Let the value of variable m increase by 1.

[0072] Step 12: Return to Step 4 until the value of variable m is greater than the total number of UAVs M.

[0073] Step 13: All UAVs perform transmission start time and calculation start time settings according to where, is the set of transmission start times after the M-th optimization, is the set of calculation start times after the M-th optimization.

[0074] Step 14: Return to Step 1 until the auxiliary tasks of the UAVs are completed.

[0075] Example:

[0076] Simulation parameter settings: The side length of the square area where the IoT devices are located is 500 meters, the number of IoT devices is K = 20, the transmission power of the IoT devices is all 0 dBm, the number of UAVs is M = 4, the number of antennas of each UAV is L = 32, the computing power of the UAVs is F = 2.4 GHz, the optional displacement distances of the UAVs are d1 = 250 meters and d2 = 25 meters, the flight time of any UAV is 8.3 meters per second, the system bandwidth is B = 5 MHz, the bandwidth allocated to any UAV is 1.25 MHz, the Rice factor related constants are A1 = 0 dB and A2 = 30 dB, the coherence interval is τ c = 200, the maximum movement time of the IoT devices is seconds, the slot length is T Z = 90 seconds, the number of times the complex state information of the IoT devices is updated within each slot is U = 10 times, the original data volume of the IoT devices is in the range of [0.1, 0.5] Mbits, and the noise power spectral density is N0 = -174 dBm / Hz. The weight of the UAV is about 2 KG, and its hovering power is about P H= 168.5 W, c1 = 357.2, c2 = 79.9, c3 = 0.02, c4 = 0.009 (Y. Zeng, J. Xu, and R. Zhang, “Energy minimization for wireless communication with rotary-wing UAV,” IEEE Transactions on Wireless Communications, vol. 18, no. 4, pp. 2329–2345, Apr. 2019).

[0077] Figure 2 This is a schematic diagram of the sum of the average age of information of the present invention varying with the number of UAV antennas. From Figure 2 it can be seen that the sum of the average age of information of the IoT devices hardly varies with the number of UAV antennas. This indicates that under the current system parameter settings, the change in the number of UAV antennas neither directly affects the sum of the average age of information of the IoT devices nor affects other related strategies. It should be noted that when the number of UAV antennas becomes smaller, the transmission time of the IoT devices increases, which may lead to a longer time for the complex state update process, thus increasing the average age of information. At the same time, the sum of the average age of information of the IoT devices under the method proposed in the present invention is less than that of the method that only optimizes the UAV position, and the performance gap between the method proposed in the present invention and the self-learning machine with a higher computational complexity is acceptable. This shows that the method proposed in the present invention can effectively improve the information freshness in the complex state information update process.

[0078] Figure 3 This is a schematic diagram of the sum of the energy consumption of the IoT devices of the present invention varying with the number of UAV antennas. From Figure 3 it can be seen that the sum of the energy consumption of the IoT devices decreases with the increase in the number of UAV antennas. This indicates that the spatial gain brought by the increase in the number of UAV antennas can reduce the transmission time of the IoT devices, thereby reducing the energy consumption of the IoT devices. For the sum of the energy consumption of the IoT devices, the method proposed in the present invention is inferior to the other two methods, which shows that under the method proposed in the present invention, the IoT devices need to sacrifice part of their energy to improve other performances.

[0079] Figure 4 This is a schematic diagram of the sum of the energy consumption of the UAVs of the present invention varying with the number of UAV antennas. In Figure 4 it, the sum of the energy consumption of the UAVs under the method proposed in the present invention is less than that of other optimization methods. This result shows that the method proposed in the present invention can effectively reduce the energy consumption of the UAVs. This also further indicates that there is a trade-off relationship among the average age of information, the energy consumption of the IoT devices, and the energy consumption of the UAVs, and the method proposed in the present invention sacrifices the energy of the IoT devices in exchange for the reduction of the average age of information and the energy consumption of the UAVs.

[0080] Figure 5 Schematic diagram of the weighted sum of three performance indicators of the present invention varying with the number of UAV antennas. From Figure 5 It can be seen that compared with the method of only optimizing the UAV position, the optimization method proposed in the present invention can effectively reduce the weighted sum of the three performance indicators, thereby improving the comprehensive performance of the system. Although the performance of the self-learning machine is better than the method proposed in the present invention, its computational complexity is relatively high and it is not suitable for dynamic optimization scenarios. At the same time, the weighted sum decreases with the increase in the number of UAV antennas, indicating that the increase in the number of UAV antennas can improve the comprehensive performance of the system.

[0081] The description of the above embodiments is relatively specific and detailed, but only expresses a feasible implementation manner of the present invention, and does not limit the scope of the invention patent. It should be noted that researchers and engineers in the field can add several deformations or improvements on the basis of this embodiment within the framework of the present invention, but these are all within the protection scope of the invention patent, and the protection scope of the invention patent shall be subject to the appended claims.

Claims

1. A dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle, characterized in that, It includes the following steps: Step 1: The Internet of Things devices transmit real-time location information to the drones. All drones exchange their current positions and the location information of the Internet of Things devices they have collected. Assume the total number of drones is M and all drones are numbered; Step 2: Establish a dynamic optimization problem for the drone-assisted complex state information update; Step 3: The No. 1 UAV generates an initial feasible solution {t 0 , c 0 , q 0} according to the constraint conditions of the dynamic optimization problem described in Step 2, where t 0 is a set of starting times for the transmission of feasible IoT devices, c 0 is a set of starting times for the calculation of feasible UAVs, and q 0 is a set of feasible UAV positions; meanwhile, let the value of the variable m be 1; Step 4: The m-th drone obtains all actions from the action set {ascend, descend, move left, move right, move forward, move backward, stay still} that satisfy the constraints of the dynamic optimization problem in Step 2, and let the set composed of all actions of the m-th drone that satisfy the constraints of the dynamic optimization problem in Step 2 be and let the set The total number of elements in is Meanwhile, let the value of variable a be 1; Step 5: The m-th drone generates an initial feasible solution according to the a-th behavior in and determines the connection relationship between the IoT devices and the drones by using the location-based matching game method, where is the set of feasible transmission start times of the IoT devices generated by the m-th drone according to the a-th behavior, is the set of feasible calculation start times of the drones generated by the m-th drone according to the a-th behavior; Step 6: Obtain the optimal solutions for the transmission start time and the calculation start time in the first to the (U - 1)-th complex state information updates, where U is the total number of complex state information updates; Step 7: Obtain the transmission start time and the calculation start time in the U-th optimized state update and Step 8: Obtain the optimal solution according to the objective function of the optimization problem in Step 2 of the objective function value wherein, and the value range of variable u is all positive integers less than U, and the variable has a value range of all positive integers not greater than M excluding the current value of variable m, is the transmission start time of the optimized m-th drone at the u-th complex state information update, is the initial drone's transmission start time at the u-th complex state information update, is the calculation start time of the optimized m-th drone at the u-th complex state information update, is the initial drone's calculation start time at the u-th complex state information update; let the value of variable a increase by 1; Step 9: Return to Step 5 until the value of variable a is greater than Step 10: Record The behavior number corresponding to the minimum value in * is a * , and obtain the position of the m-th drone after executing the a-th behavior Replace m-1 the position of the m-th drone in m , and let the set of replaced drone positions be q m -1 , where q -1 is the initial drone position at the m-th optimization; Step 11: The m-th unmanned aerial vehicle (UAV) flies from the current position to the optimized position and broadcasts q to all other UAVs m ; increment the value of variable m by 1; Step 12: Return to Step 4 until the value of variable m is greater than the total number of drones M; Step 13: All drones set the transmission start time and the calculation start time according to , where is the set of transmission start times after the Mth optimization, and is the set of calculation start times after the Mth optimization; Step 14: Return to Step 1 until the auxiliary tasks of the drones are completed.

2. A dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle according to claim 1, characterized in that The optimization problem described in Step 2 is as follows: Among them, \(Z\) is the set of optimization variables, and are the weight factors of the average age of information, the energy consumption of IoT devices, and the energy consumption of UAVs respectively, is the number of IoT devices served by the \(m\)-th UAV, \(t\) u,m and \(t\) u+1,m are the transmission start times of the IoT devices served by the \(m\)-th UAV at the \(u\)-th and \((u + 1)\)-th status updates respectively, and is the transmission delay of the \(k\)-th IoT device served by the \(m\)-th UAV, \(\tau\) c is the coherence interval, \(D\) k,m is the amount of task data of the \(k\)-th IoT device served by the \(m\)-th UAV, \(B\) m is the communication bandwidth allocated to the \(m\)-th UAV, \(p\) k,m is the transmit power of the \(k\)-th IoT device served by the \(m\)-th UAV, is the channel coefficient between the \(m\)-th UAV and the \(k\)-th IoT device it serves, \(g_0\) is the received power when the transceiver distance is 1 meter, \(d\) k,m is the distance between the \(m\)-th UAV and the \(k\)-th IoT device it serves, \(L\) is the number of UAV antennas, is the pilot transmit power of the \(i\)-th IoT device served by the \(m\)-th UAV, \(g\) i,m is the channel coefficient between the \(m\)-th UAV and the \(i\)-th IoT device it serves, is the Rice factor between the \(m\)-th UAV and the \(i\)-th IoT device it serves, \(A_1\) and \(A_2\) are constants, \(\theta\) i,m is the elevation angle between the \(m\)-th UAV and the \(i\)-th IoT device it serves, is the pilot transmit power of the \(k\)-th IoT device served by the \(m\)-th UAV, \(\Omega\) m is a diagonal matrix and the element on the \(k\)-th main diagonal is \([\Omega\) m kk \(= K\) k,m \(K\) k,m is the Rice factor between the \(m\)-th UAV and the \(k\)-th IoT device it serves, \(I\) m is the identity matrix, is a matrix and the element in its \(l\)-th row and \(k\)-th column is \(\theta\) k,m is the elevation angle between the \(m\)-th UAV and the \(k\)-th IoT device it serves, and​ are the set of complex status information update times and the set of UAVs, respectively, c u,m and c u+1,m are the starting calculation times of the m-th UAV for the u-th and (u + 1)-th complex status information updates, respectively, is the computing delay of the k-th IoT device served by the m-th UAV, N ue is the upper limit of the number of IoT devices that a UAV can serve, C k,m is the number of CPU rotations required for the m-th UAV to complete the task of the k-th IoT device it serves. F is the computing power of the UAV, t 1,m is the starting transmission time of the IoT device served by the m-th UAV for the first status update, is the time taken for all IoT devices to move from the previous target point to the current target point, TF m is the flight time of the m-th UAV, c U,m is the starting calculation time of the m-th UAV at the U-th status update, T Z is the time slot length, and are the three-dimensional position coordinates of the m-th UAV after and before taking an action, respectively, d1 and d2 are the selectable flight distances, and QS is the set of all allowed UAV hovering positions, and are the three-dimensional position coordinates of the x-th and y-th UAVs after taking an action, respectively; In Equation (1), the expression for the average age of information of the k-th Internet of Things device served by the m-th drone is as follows: Among them, c 1,m is the calculation start time when the m-th unmanned aerial vehicle (UAV) updates its complex state for the first time, and t u-1,m is the transmission start time when the Internet of Things (IoT) device served by the m-th UAV updates its complex state for the (u - 1)-th time, and t U-1,m is the transmission start time when the IoT device served by the m-th UAV updates its complex state for the (U - 1)-th time; In Equation (1), the energy consumption for the m-th drone to complete a task offloading for the k-th Internet of Things device is: In Equation (1), the energy consumption of the m-th drone in one time slot is: where c1, c2, c3, and c4 are constants related to the physical characteristics of the drone and the environment, and P H is the hovering power of the drone, and are the position coordinates of the m-th drone before and after performing the behavior, respectively.

3. A dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle according to claim 2, characterized in that Step 5 according to A m The ath action in produces the initial feasible solution The process is as follows: when the ath behavior is to remain stationary, let Among them, t m-1 is the initial IoT device transmission start time set in the mth optimization, c m-1 Calculate the starting time set for the initial UAV in the mth optimization; when the ath behavior is not to remain stationary, the mth UAV randomly generates an initial feasible solution according to the ath behavior 4. A dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle according to claim 3, characterized in that The location-based matching game method described in Step 5 is as follows: The m-th drone establishes a distance-based preference relation table for each Internet of Things device and drone based on the feasible drone positions generated by the a-th behavior; For the Internet of Things devices, any Internet of Things device starts sending connection requests from the drone with a larger preference value in the preference relation table. When the connection request is rejected by the drone, it continues to send connection requests to the next drone with a smaller preference value in the list until it is not rejected; For the drones, any drone judges the Internet of Things devices sending connection requests according to the preference list and the maximum number of Internet of Things devices it can serve; When the number of Internet of Things devices it serves is less than the maximum service quantity, it allows the connection request; When the number of Internet of Things devices it serves has reached the maximum service quantity and the preference value of the drone for the Internet of Things device making the connection request is greater than the preference value of the Internet of Things device with an established connection, it accepts the connection request and rejects the connection of the Internet of Things device with a smaller preference value; When the number of Internet of Things devices it serves has reached the maximum service quantity and the preference value of the drone for the Internet of Things device making the connection request is less than the preference value of the Internet of Things device with an established connection, it rejects the connection request.

5. A dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle according to claim 4, characterized in that, The process of obtaining the optimal solution described in Step 6 is as follows: Step 6.1: Establish the following equivalent optimization problem: Among them, t 1,m is the transmission start time when the Internet of Things device of the m-th drone service undergoes the first complex state update, t u,m is the transmission start time when the Internet of Things device of the m-th drone service undergoes the u-th complex state update, t u-1,m is the transmission start time when the Internet of Things device of the m-th drone service undergoes the (u - 1)-th complex state update, t U-1,m is the transmission start time when the Internet of Things device of the m-th drone service undergoes the (U - 1)-th complex state update, is the transmission start time for updating the U-th complex state information of all Internet of Things devices of the m-th drone service generated according to the a-th behavior, is the calculation start time for updating the U-th complex state information of the m-th drone generated according to the a-th behavior; Step 6.2: Solve the optimal solution of the equivalent optimization problem using the Lagrange multiplier method where the value range of the integer variable u is 1 ≤ u ≤ U; Step 6.3: According to obtain wherein, the value range of the variable u in is 1 ≤ u ≤ U, are respectively the optimal values of the equivalent variables corresponding to the m-th unmanned aerial vehicle (UAV) for the first, u-th, and U-th complex state information updates, is the starting moment of the first complex state information update transmission of the optimized m-th UAV, is the starting moment of the u-th complex state information update transmission of the optimized m-th UAV, is the starting moment of the (u - 1)-th complex state information update transmission of the optimized m-th UAV, is the starting moment of the (U - 1)-th complex state information update transmission of the optimized m-th UAV; Step 6.4: Let where the integer variable u ranges from 1 ≤ u ≤ U, is the starting time of the u-th complex state information update calculation for the optimized m-th UAV.

6. The dynamic optimization method for updating complex state information of an Internet of Things assisted by an unmanned aerial vehicle according to claim 5, characterized in that, Obtaining the transmission start time and the calculation start time in the optimal U-th state update described in step 7 is as follows: The process is as follows: Step 7.1: Let Step 7.2: Let the lower bound and upper bound of c U,m be and respectively, and define the function where is the starting moment of the (U - 1)-th complex state information update calculation for the optimized UAV No. m, Step 7.3: When g(c lb ) = 0, then let When g(c ub ) = 0, then let When g(c lb ) < 0 and g(c ub ) > 0, then use the bisection method to obtain the solution c U,m such that g(c * ) = 0 and let When g(clb) > 0, then let When g(c ub ) < 0, then let

Images