Precise coverage optimization method for multi-UAV relay communication in a certain city

Abstract

The present invention provides a method for optimizing the precise coverage of multi-UAV relay communication in a determined city, comprising the following steps: modeling a selected urban area, constructing a mission scenario for an UAV, discretizing a map according to a preset resolution, and obtaining a mission unit; calculating a path loss for a designated urban area, and obtaining a path loss; setting a mission hotspot in the mission scenario, and initializing the UAV; constructing a UAV coverage utility function using the path loss, the UAV transmission power, and the mission hotspot in the mission area; setting a UAV action set, predicting, selecting, and judging the movement of the UAV according to the UAV action set, avoiding unreasonable UAV movement directions, and preventing the UAV from moving outside the mission area; iterating and accurately positioning the UAV coverage through a preset algorithm, and evaluating and predicting the coverage effect of a UAV group based on the UAV coverage utility function, and making a decision on the next action of the UAV.

Description

Technical Field

[0001] The present invention relates to the technical field of radio wave propagation calculation simulation technology, and in particular to a method for optimizing the precise coverage of multi-UAV relay communications in a specific city. Background Art

[0002] The communication modes of drone swarm communication networks are mainly divided into: pre-planning, air-ground communication, self-organizing network, and clustering network. In the pre-planning mode, the drone swarm is commanded by the ground control center before the mission is carried out, and the drone swarm operation track, coverage area, and deployment location are determined. The drone coverage problem is a basic problem of drone swarm communication. Many studies on drone coverage at home and abroad focus on the problem of maximizing coverage. Since drone swarms are often small in size and carry limited energy, many articles also focus on the problem of drone energy loss under drone coverage. In China, many literatures have proposed different algorithms for drone coverage problems. And these different algorithms can play their role in different environments and mission objectives. For example, a literature proposes a multi-drone coverage model based on energy-saving communication, and then uses a multi-drone energy-saving coverage deployment algorithm based on spatial adaptive performance (MUECD-SAP) to maximize coverage and control power, ensuring the optimal energy-saving coverage deployment. There are also literatures that have conducted simulation calculations for specific areas, designed drone coverage utility functions, and used log-linear learning deployment algorithms to achieve the goal of full network coverage for disaster areas. At the same time, they compared the number of drones, drone positions, altitudes, and maximum communication requirements of ground equipment with the optimal deployment algorithm, and obtained better results. Using binary log-linear learning algorithms for motion control to maintain optimal coverage is a common method. In addition, there are also articles that use clustering algorithms, particle swarm algorithms, and other algorithms to optimize drone coverage. These algorithms can maximize drone coverage, but the above methods often use statistical models rather than deterministic models to deal with path loss, so the effect is usually somewhat different from the real environment. Summary of the invention

[0003] In view of this, the purpose of the present invention is to provide a method for optimizing the precise coverage of multi-UAV relay communications in a certain city, so as to overcome or at least partially solve the above-mentioned problems existing in the prior art.

[0004] To achieve the above-mentioned purpose of the invention, the present invention provides a method for optimizing the precise coverage of multi-UAV relay communication in a certain city, comprising the following steps:

[0005] S101, modeling the selected urban area, constructing the mission scenario of the UAV, discretizing the map according to the preset resolution, and obtaining the mission unit;

[0006] S102, calculating the path loss for the designated urban area to obtain the path loss L;

[0007] S103, setting a mission hotspot in the mission scene and initializing the drone;

[0008] S104, using path loss L, UAV transmission power P s , mission area mission hotspot η(i) constructs drone coverage utility function U(a n , a -n );

[0009] S105, setting a drone action set, predicting, selecting and judging the drone movement according to the drone action set, avoiding unreasonable drone movement directions, and preventing the drone from moving outside the mission area;

[0010] S106. Iterate and accurately locate the drone coverage through a preset algorithm, evaluate and predict the coverage effect of the drone group based on the drone coverage utility function, and make a decision on the next action of the drone.

[0011] Furthermore, the initialization settings of the drone include initializing the drone height H, the drone transmission power P s , UAV transmission frequency F, UAV antenna parameters and the initial position of the UAV.

[0012] Furthermore, the step S101 specifically includes the following steps:

[0013] S201, selecting an urban area to be simulated, collecting a map of the selected urban area and establishing a mission scenario;

[0014] S202, set the resolution of the map when it is discretized, divide the mission scene into mission units i∈Ω of equal size, each mission unit i is the minimum unit for path loss calculation and also the minimum step length of the UAV movement, and perform equivalent processing on the scene information in each mission unit. The mission unit after equivalent processing carries the information of the surface and buildings in the mission unit, and Ω is the mission scene.

[0015] Further, the path loss is calculated for the designated urban area to obtain the path loss, which specifically includes the following steps:

[0016] S301, modeling the drone and setting the drone parameters;

[0017] S302, simulate the movement of the UAV in the mission area, starting from the origin, each time it moves a mission unit, simulate and calculate the path loss of the entire scene at the current point, obtain the path loss data in the scene, export the path loss file, and repeat this step until the UAV has simulated the path loss of all mission units in the scene, and obtain the path loss set of the UAV at all positions.

[0018] Furthermore, task hotspots are set in the task scene. Specifically, for the task scene area, the area that needs to be covered is selected, and a normalized task hotspot is set for the selected area according to the task goal:

[0019] η(i):Ω→R + (1)

[0020] Where Ω is the task scenario, R + represents the field of positive real numbers.

[0021] Furthermore, the drone coverage utility function is constructed, which specifically includes the following steps:

[0022] S401, setting a receiving power threshold Th of communication equipment in a task area;

[0023] S402, through the transmission power P s The difference between the path loss L and the received power P of each task unit in the task area is calculated r , the calculation formula is as follows:

[0024] P r =P s -L (2)

[0025] S403, considering the real environmental factors affecting the receiving power of each task unit in the calculation task area The calculation formula is:

[0026]

[0027] where σ 2 represents the variance of Gaussian white noise;

[0028] S404: Compare the received power of all task units at all locations with the received power threshold, and make a determination according to the following formula:

[0029]

[0030] S405. The global utility function is constructed as:

[0031]

[0032] The local utility function is constructed as:

[0033]

[0034] Where N is the set of drones, a n and a -n represent the action strategies taken by drone n and drones other than drone n, respectively, and P r (n) is the received power set of each unit of UAV n in the mission area, Ω n is the coverage of drone n, ξ is the set of neighboring drones of drone n, and N\ξ is the set of drones that do not overlap with the coverage of neighboring drones of drone n;

[0035] S406, conduct game analysis, and construct the potential energy function as follows:

[0036]

[0037] Therefore, formula (6) can be rewritten as:

[0038]

[0039] When the drone action strategy a n Change to When:

[0040]

[0041] Furthermore, the step S105 specifically includes the following steps:

[0042] S501, setting the movement rules of the drone at the mission boundary and corners to prevent the drone from moving outside the mission area, setting the movement rules of the drone inside the mission area, and using two sets of action sets with different step lengths to regulate the movement of the drone;

[0043] S502, the drone action set is set according to the scene unit, the action set with a smaller step length has a step length of one task unit, and the action set with a larger step length has a step length of two task units;

[0044] S503, setting a drone action set change threshold S according to the mission objective. When U≤S, the drone moves with a larger step motion set; when U>S, the drone moves with a smaller step motion set.

[0045] Furthermore, the preset algorithm specifically includes the following steps:

[0046] S601, setting the scene and drone information and initializing the drone position;

[0047] S602, randomly select a drone and record the action strategy of the current position as an , and select the action set corresponding to the selected drone n, and randomly predict the next action from the action set

[0048] S603, using formula (6) to calculate the coverage utility value at the current position and the coverage utility value at the predicted position, and using the probability selection function to select the current position an and the predicted position Make a selection, the probability selection function is as follows:

[0049]

[0050] Where β is the learning coefficient;

[0051] S604, using formula (5) to calculate the global utility of the drone in the new position, to measure the coverage effect of the drone;

[0052] S605. Repeat steps S602-S604 until the algorithm converges.

[0053] Compared with the prior art, the present invention has the following beneficial effects:

[0054] (1) The present invention models a specified urban environment and obtains the path loss data between the UAV and the ground in a specific urban environment, thereby improving the data and effect errors caused by the traditional method of using an empirical model to calculate the path loss;

[0055] (2) The drone action set set by the present invention expands the drone's range of motion and can achieve the drone's optimal coverage more quickly;

[0056] (3) The present invention can very intuitively demonstrate the coverage effect of drones in specific urban areas;

[0057] (4) The present invention can switch scenes according to task requirements and can adapt to changes in scenes. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only preferred embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0059] Figure 1 It is a schematic diagram of the overall process of a method for optimizing precise coverage of multi-UAV relay communications in a specific city provided by an embodiment of the present invention.

[0060] Figure 2It is a designated urban area task scenario diagram provided by an embodiment of the present invention.

[0061] Figure 3 This is an initial position map of the drone provided by an embodiment of the present invention.

[0062] Figure 4 This is a drone action set diagram provided by an embodiment of the present invention.

[0063] Figure 5 This is a global utility trend diagram of eight drones provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0064] The principles and features of the present invention are described below in conjunction with the accompanying drawings. The enumerated embodiments are only used to explain the present invention and are not used to limit the scope of the present invention.

[0065] Reference Figure 1-5 This embodiment provides a method for optimizing the precise coverage of multi-UAV relay communication in a specific city, and the method includes the following steps:

[0066] S101, modeling the selected urban area, constructing the mission scenario of the UAV, discretizing the map according to a preset resolution, and obtaining the mission unit.

[0067] S102: Calculate the path loss for the designated urban area to obtain the path loss L.

[0068] S103: Setting a mission hotspot in the mission scene and initializing the drone.

[0069] S104, using path loss L, UAV transmission power P s , mission area mission hotspot η(i) constructs drone coverage utility function U(a n , a -n ).

[0070] S105: Setting a drone action set, predicting, selecting and judging the drone movement according to the drone action set, avoiding unreasonable drone movement directions, and preventing the drone from moving outside the mission area.

[0071] S106. Iterate and accurately locate the drone coverage through a preset algorithm, evaluate and predict the coverage effect of the drone group based on the drone coverage utility function, and make a decision on the next action of the drone.

[0072] Exemplarily, initializing the drone includes initializing the drone height H, the drone transmission power P s , UAV transmission frequency F, UAV antenna parameters and the initial position of the UAV.

[0073] Step S101 specifically includes the following steps:

[0074] S201, selecting an urban area to be simulated, collecting a map of the selected urban area and establishing a mission scenario.

[0075] S202, set the resolution of the map when it is discretized, divide the mission scene into mission units i∈Ω of equal size, each mission unit i is the minimum unit for path loss calculation and also the minimum step length of the UAV movement, and perform equivalent processing on the scene information in each mission unit. The mission unit after equivalent processing carries the information of the surface and buildings in the mission unit, and Ω is the mission scene.

[0076] In step S102, path loss calculation is performed on the designated urban area to obtain the path loss, which specifically includes the following steps:

[0077] S301, modeling the UAV and setting the UAV parameters.

[0078] S302, simulate the movement of the UAV in the mission area, starting from the origin, each time it moves a mission unit, simulate and calculate the path loss of the entire scene at the current point, obtain the path loss data in the scene, export the path loss file, and repeat this step until the UAV has simulated the path loss of all mission units in the scene, and obtain the path loss set of the UAV at all positions.

[0079] In step S103, task hotspots are set in the task scene, specifically: for the task scene area, an area that needs to be covered emphatically is selected, and a normalized task hotspot is set for the selected area according to the task goal:

[0080] η(i):Ω→R + (1)

[0081] Where Ω is the task scenario, R + represents the field of positive real numbers.

[0082] like Figure 2 As shown, the urban area and the nearby sea area are designated as the drone scene, the mission hotspot is set to the urban area, and the drone is required to cover the urban area.

[0083] In step S104, a drone coverage utility function is constructed, which specifically includes the following steps:

[0084] S401. Set a receiving power threshold Th of communication equipment within a task area.

[0085] S402, through the transmission power P s The difference between the path loss L and the received power P of each task unit in the task area is calculatedr , the calculation formula is as follows:

[0086] P r =P s -L (2)

[0087] S403, considering the real environmental factors affecting the receiving power of each task unit in the calculation task area The calculation formula is:

[0088]

[0089] where σ 2 represents the variance of Gaussian white noise.

[0090] S404: Compare the received power of all task units at all locations with the received power threshold, and make a determination according to the following formula:

[0091]

[0092] S405. The global utility function is constructed as:

[0093]

[0094] The local utility function is constructed as:

[0095]

[0096] Where N is the set of drones, a n and a -n represent the action strategies taken by drone n and drones other than drone n, respectively, and P r (n) is the received power set of each unit of UAV n in the mission area, Ω n is the coverage of UAV n, ξ is the set of neighboring UAVs of UAV n, and N\ξ is the set of UAVs that do not overlap with the coverage of neighboring UAVs of UAV n.

[0097] S406, conduct game analysis, and construct the potential energy function as follows:

[0098]

[0099] Therefore, formula (6) can be rewritten as:

[0100]

[0101] When the drone action strategy a n Change to When:

[0102]

[0103]

[0104] Therefore, this potential game is an exact potential game and there exists one or more Nash equilibrium solutions.

[0105] Step S105 specifically includes the following steps:

[0106] S501, setting the movement rules of the drone at the mission boundary and corners to prevent the drone from moving outside the mission area, setting the movement rules of the drone inside the mission area, and using two sets of action sets with different step lengths to regulate the drone movement.

[0107] S502, the drone action set is set according to the scene unit, the action set with a smaller step length has a step length of one task unit, and the action set with a larger step length has a step length of two task units.

[0108] S503, setting a drone action set change threshold S according to the mission objective. When U≤S, the drone moves with a larger step motion set; when U>S, the drone moves with a smaller step motion set.

[0109] In step S106, the preset algorithm specifically includes the following steps:

[0110] S601: Set the scene and drone information and initialize the drone position.

[0111] S602, randomly select a drone and record the action strategy of the current position as a n , and select the action set corresponding to the selected drone n, and randomly predict the next action from the action set

[0112] S603, using formula (6) to calculate the coverage utility value at the current position and the coverage utility value at the predicted position, and using the probability selection function to select the current position an and the predicted position Make a selection, the probability selection function is as follows:

[0113]

[0114] Where β is the learning coefficient.

[0115] S604: Calculate the global utility of the UAV in the new position using formula (5) to measure the coverage effect of the UAV.

[0116] S605. Repeat steps S602-S604 until the algorithm converges.

[0117] The algorithms in Table 1 all use binary log-linear learning algorithms. The former method uses an empirical model to calculate the path loss, while this embodiment uses a ray tracing method to calculate the path loss between cities. In the simulation of a specified urban area, the two methods obtain different multi-UAV coverage positions, resulting in different coverage utilities of multiple UAVs. It can be seen from Table 1 that for different numbers of UAVs, the method used in this embodiment effectively improves the coverage effect of the city and achieves accurate coverage of relay communications for specific cities.

[0118] Table 1

[0119]

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for optimizing the precise coverage of multi-UAV relay communication in a certain city, characterized in that: The method comprises the following steps: S101, modeling the selected urban area, constructing the mission scenario of the UAV, discretizing the map according to the preset resolution, and obtaining the mission unit; S102, calculating the path loss for the designated urban area to obtain the path loss L; S103, setting a mission hotspot in the mission scene and initializing the drone; S104, using path loss L, UAV transmission power P s , mission area mission hotspot η(i) constructs drone coverage utility function U(a n ,a -n ), i represents the task unit; S105, setting a drone action set, predicting, selecting and judging the drone movement according to the drone action set, avoiding unreasonable drone movement directions, and preventing the drone from moving outside the mission area; S106, iterating and accurately positioning the drone coverage through a preset algorithm, evaluating and predicting the coverage effect of the drone group based on the drone coverage utility function, and making a decision on the next action of the drone; Constructing the drone coverage utility function includes the following steps: S401, setting a receiving power threshold Th of communication equipment in a task area; S402, through the transmission power P s The difference between the path loss L and the received power P of each task unit in the task area is calculated r , the calculation formula is as follows: P r =P s -L (2) S403, considering the real environmental factors affecting the receiving power of each task unit in the calculation task area The calculation formula is: where σ 2 represents the variance of Gaussian white noise; S404: Compare the received power of all task units at all locations with the received power threshold, and make a determination according to the following formula: S405. The global utility function is constructed as: The local utility function is constructed as: Where N is the set of drones, a n and a -n represent the action strategies taken by drone n and drones other than drone n, respectively, and P r (n) is the received power set of each unit of UAV n in the mission area, Ω n is the coverage of drone n, ξ is the set of neighboring drones of drone n, and N\ξ is the set of drones that do not overlap with the coverage of neighboring drones of drone n; S406, conduct game analysis, and construct the potential energy function as follows: Therefore, formula (6) can be rewritten as: When the drone action strategy a n Change to When: The preset algorithm specifically includes the following steps: S601, setting the scene and drone information and initializing the drone position; S602, randomly select a drone and record the action strategy of the current position as a n , and select the action set corresponding to the selected drone n, and randomly predict the next action from the action set S603, using formula (6) to calculate the coverage utility value at the current position and the coverage utility value at the predicted position, and using the probability selection function to select the coverage utility value at the current position a n And the predicted location Make a selection, the probability selection function is as follows: Where β is the learning coefficient; S604, using formula (5) to calculate the global utility of the drone in the new position, to measure the coverage effect of the drone; S605. Repeat steps S602-S604 until the algorithm converges.

2. According to claim 1, a method for optimizing the precise coverage of multi-UAV relay communication in a specific city is characterized in that: Initialize the drone settings including initializing the drone height H, the drone transmission power P s , UAV transmission frequency F, UAV antenna parameters and the initial position of the UAV.

3. According to claim 1, a method for optimizing the precise coverage of multi-UAV relay communication in a specific city is characterized in that: The step S101 specifically includes the following steps: S201, selecting an urban area to be simulated, collecting a map of the selected urban area and establishing a mission scenario; S202, set the resolution of the map when it is discretized, divide the mission scene into mission units i∈Ω of equal size, each mission unit i is the minimum unit for path loss calculation and also the minimum step length of the UAV movement, and perform equivalent processing on the scene information in each mission unit. The mission unit after equivalent processing carries the information of the surface and buildings in the mission unit, and Ω is the mission scene.

4. According to claim 1, a method for optimizing the precise coverage of multi-UAV relay communication in a specific city is characterized in that: Calculate the path loss for a specified urban area to obtain the path loss, which specifically includes the following steps: S301, modeling the drone and setting the drone parameters; S302, simulate the movement of the UAV in the mission area, starting from the origin, each time it moves a mission unit, simulate and calculate the path loss of the entire scene at the current point, obtain the path loss data in the scene, export the path loss file, and repeat this step until the UAV has simulated the path loss of all mission units in the scene, and obtain the path loss set of the UAV at all positions.

5. According to claim 1, a method for optimizing the precise coverage of multi-UAV relay communication in a specific city is characterized in that: Set task hotspots in the task scene. Specifically, select the area that needs to be covered in the task scene area, and set a normalized task hotspot for the selected area according to the task goal: η(i):Ω→R + (1) Where Ω is the task scenario, R + represents the field of positive real numbers.

6. The method for optimizing the precise coverage of multi-UAV relay communication in a specific city according to claim 1, characterized in that: The step S105 specifically includes the following steps: S501, setting the movement rules of the drone at the mission boundary and corners to prevent the drone from moving outside the mission area, setting the movement rules of the drone inside the mission area, and using two sets of action sets with different step lengths to regulate the movement of the drone; S502, the drone action set is set according to the scene unit, the action set with a smaller step length has a step length of one task unit, and the action set with a larger step length has a step length of two task units; S503, setting the drone action set change threshold S according to the mission objective, when U≤S, the drone adopts a larger step motion set for movement; when U>S, the drone adopts a smaller step motion set for movement.

Images