SVC hierarchical edge cache optimization method and device based on unmanned aerial vehicle cooperation

Abstract

The invention discloses an SVC hierarchical edge cache optimization method and device based on unmanned aerial vehicle cooperation, and belongs to the technical field of wireless communication. A system applied to the method comprises a remote server, a ground base station, unmanned aerial vehicles and ground users. The unmanned aerial vehicle service stage keeps hovering, a ground user initiates a video image quality request to the associated unmanned aerial vehicle, and the associated unmanned aerial vehicle selects a video transmission link mode according to a cache hit condition; the optimization method comprises the following steps: establishing an optimization problem of minimizing the total video transmission delay of a system under the constraint conditions of the cache capacity and the coverage range of the unmanned aerial vehicle; based on the cache capacity and coverage constraint of the unmanned aerial vehicle, the optimization problem is decomposed into sub-problems of user unmanned aerial vehicle association and unmanned aerial vehicle cooperative association, unmanned aerial vehicle cache strategy and unmanned aerial vehicle service hovering position optimization; the optimal solution of each sub-problem is solved and obtained, and the global optimal solution is found through alternate iteration, so that the total video transmission delay of the system is minimized.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a method and apparatus for optimizing SVC (Scalable Video Coding) hierarchical edge buffering based on UAV collaboration. Background Technology

[0002] With the widespread application of high-bandwidth services such as high-definition video, virtual reality, and augmented reality, the demand for network resources is constantly increasing, and traditional network architectures are gradually facing the dilemma of being unable to meet these demands. Edge caching, as a technology that stores and distributes data at the network edge devices close to users, can effectively reduce data access latency, decrease core network load, and improve overall service quality. For applications with high real-time requirements, such as high-definition video streaming, augmented reality, and vehicle-to-everything (V2X) communication, edge caching can significantly improve user experience. Due to the high flexibility of drones, drone-assisted edge caching systems have gained attention. Considering the flexible deployment capabilities and high-probability line-of-sight (LAS) communication advantages of drones, drone-assisted SVC (Surface View Caching) video caching can be utilized.

[0003] However, there is relatively little research on drone-assisted SVC video caching. Existing system models typically use drones to assist base stations in transmitting video to users. Furthermore, drones have limited caching capacity, and each drone can only cache a small number of videos. Therefore, when the number of videos is large, the drone cache hit rate is low, and a large number of videos can only be transmitted from the base station to the ground user via the backhaul link to the core network, which greatly increases the load on the core network. Summary of the Invention

[0004] To address the shortcomings of existing technologies, the present invention aims to provide a method and apparatus for optimizing SVC hierarchical edge caching based on UAV collaboration, thereby solving the problems in existing technologies.

[0005] The objective of this invention can be achieved through the following technical solutions: A hierarchical edge caching optimization method based on UAV collaboration is proposed. The system to which this optimization method is applied includes: a remote server and a ground base station. A drone and A ground user; during the drone service phase, the drone remains hovering, and the ground user initiates a video quality request to the associated drone. The associated drone selects the video transmission link mode based on the cache hit status; the optimization method includes: Given the constraints of the drone's buffer capacity and coverage area, we establish an optimization problem to minimize the total latency of system video transmission. Based on the constraints of drone cache capacity and coverage, the optimization problem is decomposed into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; Solve the subproblems, obtain the optimal solutions for each subproblem, and find the global optimal solution through alternating iterations, so as to minimize the total video transmission delay of the system.

[0006] Furthermore, the optimization problem P1 for minimizing the total video transmission delay of the system is: in, It refers to the number of ground users. It is the number of drones. It refers to the number of videos. It is the number of encoding layers in the video. Is this a video request? The The probability of the layer, It is for users Transmit video The Layer delay, It is a user-drone association variable. These are the variables that correlate drones. It is a decision variable for drone caching. It is a base station caching decision variable. It is the three-dimensional position variable of the UAV; Representing binary optimization variables respectively , This indicates that all video quality can be obtained from the cloud video library via the base station. This means that each ground user is associated with a drone. This indicates a constraint on the drone's cache capacity. The collision avoidance distance between the two drones is... The maximum communication distance is , It is a drone The three-dimensional position, It is a drone The three-dimensional position, Indicates the flight altitude limit of the drone. This indicates the drone's flight altitude; the minimum altitude is... The highest height is .

[0007] Furthermore, the optimization sub-problem P2 of user drone association and drone cooperative association is as follows: The sub-problem P3 for optimizing the drone caching strategy is: The sub-problem P4 for optimizing the hovering position of the drone service is: .

[0008] Furthermore, the process of solving the subproblem includes: The greedy search method is used to solve the subproblems of user drone association and drone cooperative association optimization. The dynamic programming method is used to solve the subproblem of optimizing the drone caching strategy; We use the particle swarm optimization method to solve the hovering position optimization subproblem of UAV services.

[0009] Furthermore, the process of solving the sub-problems of user drone association and drone cooperative association optimization using the greedy search method includes: Introducing two constraints, problem P2 is rewritten as P5: in, This indicates the maximum associated payload for each drone. , Indicating collaborative drones For drones The signal-to-noise ratio of communication must meet a threshold condition. , It is the total transmission power for communication between drones. It is a drone The number of other drones that need to cooperate simultaneously yes arrive Channel gain, It is noise power; Greedy search is used to find the optimal solution for user drone association and drone collaboration association.

[0010] Furthermore, the process of solving the UAV caching strategy optimization subproblem using dynamic programming includes: Using the objective function transformation, problem P3 is simplified and re-expressed as P6: in, It is a drone caching layer The reduction in total system latency before and after is considered as the value of the item.

[0011] By modeling the caching strategy optimization problem of a single UAV as a 0-1 knapsack problem, the optimal caching strategy for the current UAV is solved using dynamic programming in a random order, combined with the current caching strategies of other UAVs. The optimal caching strategy for all drones was obtained through multiple iterations.

[0012] Furthermore, the specific steps for finding the global optimal solution through alternating iterations include: Step 1, Set the maximum number of iterations Convergence threshold Current iteration number The k-means clustering method is used to solve for the initial association and initial location of the user's drone, and records are... ; Step 2: Use a greedy search method to solve for the user-drone association variables and the collaboration association variables among drones. ; Step 3: Solve the drone caching strategy using dynamic programming. ; Step 4: Use the particle swarm optimization method to solve for the UAV's position. ; Step 5: Increment the iteration count by 1; Step 6: Check if the termination condition is met; if the termination condition is met, the iteration ends; otherwise, return to step 2 to continue the iteration. The termination condition includes: reaching the maximum number of iterations. or ,in, and They represent the first Second and third Total system latency for each iteration.

[0013] The SVC hierarchical edge caching optimization device based on drone collaboration performs the above method, including: Problem building module: Under the constraints of UAV buffer capacity and coverage range, establish an optimization problem to minimize the total latency of system video transmission; Problem decomposition module: Based on the constraints of drone cache capacity and coverage, the optimization problem is decomposed into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; Problem-solving module: Solve subproblems, obtain the optimal solutions to each subproblem, and find the global optimal solution through alternating iterations, thereby minimizing the total video transmission delay of the system.

[0014] A computer storage medium storing a readable program that, when executed, instructs a computing device to perform the aforementioned UAV-based SVC hierarchical edge caching optimization method.

[0015] An electronic device includes: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus; The memory is used to store at least one executable instruction that causes the processor to perform the operation corresponding to the above-described UAV-based SVC hierarchical edge caching optimization method.

[0016] The beneficial effects of this invention are: 1. This invention constructs a drone-collaborative SVC hierarchical edge caching model. The system consists of a ground base station, multiple drones, and multiple ground users. Initially, the drones randomly cache video in layers. Ground users are first clustered, with one drone serving users within a cluster. When a user initiates a video request, the drone makes an SVC layer distribution decision based on cache hit rate. If the associated drone's cache is not hit, a collaborating drone is sought to deliver the video to the ground user. If none of the drones have a cache, the video is directly obtained from the cloud via the base station. Adding a collaborative mechanism between drones to deliver video to ground users significantly increases the drone's cache hit rate, thereby reducing the number of times the base station accesses the cloud via the backhaul link to obtain video, thus reducing system transmission latency.

[0017] 2. This invention proposes a hierarchical edge caching optimization method for SVC based on UAV collaboration. Under the constraints of UAV cache capacity and coverage, an optimization problem is established to minimize the total latency of system video distribution. Since this problem is non-convex, it is decomposed into three sub-problems for solution. A greedy search method is used to obtain the optimized user-UAV association and the collaborative association between UAVs. A dynamic programming method is used to optimize the UAV caching strategy, and a particle swarm optimization method is used to solve for the 3D deployment of UAVs. Through a hierarchical decoupling solution strategy, a high-dimensional, non-convex NP-hard problem is successfully transformed into three independently solvable and efficient sub-problems, thus achieving a good balance between computational feasibility and solution quality. The computational complexity is reduced from exponential to polynomial, enabling the solution to have the potential for online adjustment in dynamic environments.

[0018] 3. After solving the three sub-problems separately, the optimal user-UAV association, collaborative association, UAV caching strategy, and UAV service hovering 3D position are obtained through alternating iterations. Simulation results show that the proposed method can significantly reduce the system's video transmission latency. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the system architecture corresponding to the SVC hierarchical edge caching optimization method of the present invention. Figure 2 This is a schematic diagram illustrating the variation of the system average latency with the number of users under different comparison algorithms for the SVC hierarchical edge caching optimization method of the present invention; Figure 3 This is a schematic diagram illustrating the variation of the system average latency with the number of users under different numbers of drones using the SVC hierarchical edge caching optimization method of the present invention; Figure 4 This is a schematic diagram illustrating the variation of the system average latency with the number of users under different UAV cache capacities using the SVC hierarchical edge caching optimization method of the present invention. Figure 5 This is a schematic diagram illustrating the variation of the system average latency with the number of users under different video popularity factors using the SVC hierarchical edge caching optimization method of the present invention. Figure 6 This is a schematic diagram illustrating the variation of the system average latency with the number of users under different video counts using the SVC hierarchical edge caching optimization method of the present invention. Detailed Implementation

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

[0022] Example 1 like Figure 1 As shown, the system applied to the SVC hierarchical edge caching optimization method based on UAV collaboration includes: a remote server, a ground base station, multiple UAVs with caching units, and multiple ground users; the number of ground users is... Using sets It indicates that the user The position is represented as The number of drones is Using sets It is stated that drones The position is represented as Ground base station The location is represented as The remote server operates at a fixed rate. Video is transmitted to users via a base station. The drone and the base station act as airborne and ground-based buffer nodes, respectively, providing services to ground users. Define a binary association variable. If ground users With over-the-air cache nodes If related, then ,otherwise Each ground user can associate with at most one drone cache node, therefore, Define a binary variable. Indicates drone With drones The associated variables, if the drone With drones If related, then ,otherwise .

[0023] The various model mechanisms involved in this invention are as follows: caching model: Assume the number of videos in the video library is Video collection This indicates that each video Encoded as Each video has a layer. Image quality. The first layer is the base layer. The subsequent layers are reinforcement layers. .video The The size of the layer is determined by It indicates, and its following And decrease. Based on the SVC principle, if and only if the user has already received the video. The layer to the first Only after the layer is reached can the user successfully decode the video. The Video quality varies. The base station can obtain all videos from the video library of a remote server via the backhaul link; therefore, the base station's caching decision variables are defined. The drone caches layered video content; the drone's caching decision variables are defined. If drone Cached video The Layer, then ,otherwise The drone's cache capacity is Therefore, the following condition is met: Assuming video popularity follows a Zipf distribution, video The request probability is expressed as: Due to each video have Image quality, user's view on video The preference for image quality follows a uniform distribution, and the probability of video image quality preference is defined as follows: Therefore, the video Image quality The request probability is expressed as: in, It is a video popularity factor.

[0024] Communication model: Based on the location distribution of ground users and the number of drones, the K-means clustering method is used to cluster the ground users, and the clusters are calculated. The center position is used as the initial horizontal position of the drone, and the initial altitude of all drones is set to the same value. Thus, drones were obtained. initial position In each cluster There is a drone inside. Communicating with users within the cluster, ground users and drones Establish access association to meet the requirements. During the service phase, the drones hovered above the cluster, allowing all ground users to view them. To its associated drones Initiate video quality Request, drone First, check if the video quality is cached. Required These layers, if they satisfy Then drone Direct communication with ground users via UAV-to-Ground (A2G) link transport layer If drone No cache ,satisfy And drones The video layer was cached, satisfying... ,and ,in It refers to the collision avoidance distance between drones. If this is the maximum communication distance between drones, then the drones and drones It is possible to establish connections for communication and collaboration, i.e., drones. First, the communication is transmitted to the drone via an air-to-air (A2A) link. transport layer ,make drones Received layer Then, the communication is transmitted to ground users via the A2G communication link. transport layer If drone satisfy Then the base station To ground users via base station to ground user communication link (G2G) transport layer The three communication links mentioned above will be discussed separately below.

[0025] (1) A2G communication link The bandwidth of the A2G communication link is uniformly set to The UAV communicates with ground users within its associated cluster using FDMA multiple access, and distributes the total bandwidth evenly among all users within the cluster. The number of users within Then the channel bandwidth for each user is The channels do not interfere with each other, and the total power transmitted by the drone to the user is [missing information]. .

[0026] A2G communication links consist of two types of links: line-of-sight (LoS) and non-line-of-sight (NLoS). (UAV) and ground users The link loss between them is expressed as: in, It is the line-of-sight link probability, expressed as: , It is an environmental regulator. It is the attenuation factor. It is a drone and ground users The angle of elevation between them is expressed as: . It is the line-of-sight path loss, expressed as: . It is the probability of a non-line-of-sight link, expressed as: . This is the non-line-of-sight path loss, expressed as: .

[0027] make and The channel gain between is Then drone For ground users The link transmission rate can be expressed as: Where N0 is the noise power.

[0028] Therefore, drones to ground users transport layer The time delay is expressed as: (2) A2A communication link drones Communication with other drones uses FDMA multiple access, with the total channel bandwidth uniformly set to... The total transmission power for communication between drones is uniformly set to... Drones The number of other drones that need to cooperate simultaneously is used This indicates that the A2A communication link is considered a Loss of Space (LoS) link for drones. To drones The link loss is expressed as: in, It is a drone and drones The distance. arrive Channel gain express.

[0029] drones For drones The link transmission rate can be expressed as: Therefore, drones To drones transport layer The time delay is expressed as: (3) G2G communication link The communication mode between the ground base station and the ground user adopts FDMA multiple access, and the total channel bandwidth is The ground base station allocates a transmission power of 100 kW to each user. The G2G communication link loss adopts a logarithmic distance path loss model, from the ground base station to the ground user. The link loss is expressed as: in, This is a reference distance. Reference distance Path loss at the location, It is the path loss index. It is a shadow decay term. Ground base stations and ground users The distance from the ground base station. Channel gain express.

[0030] Ground base station to ground users The link transmission rate can be expressed as: in, This represents the total number of ground users.

[0031] Because the remote server operates at a fixed rate to base station Transmitting data, therefore from the cloud to ground users transport layer The time delay is expressed as: Based on the above three communication links, to the user transport layer The time delay expression is as follows: Therefore, the total system delay can be expressed as: Example 2 Based on the system proposed in Example 1, this example proposes an SVC hierarchical edge caching optimization method based on UAV collaboration. During the service phase, the UAV hovers above the cluster, and all ground users... To its associated drones Initiate video quality Request, drone First, check if the video quality is cached. Required These layers, if they satisfy Then drone Direct communication with ground users via UAV-to-Ground (A2G) link transport layer If drone No cache ,satisfy And drones The video layer was cached, satisfying... ,and ,in It refers to the collision avoidance distance between drones. If this is the maximum communication distance between drones, then the drones and drones It is possible to establish connections for communication and collaboration, i.e., drones. First, the communication is transmitted to the drone via an air-to-air (A2A) link. transport layer ,make drones Received layer Then, the communication is transmitted to ground users via the A2G communication link. transport layer If drone satisfy Then the base station To ground users via base station to ground user communication link (G2G) transport layer .

[0032] When the drone communicates with ground users, the SVC hierarchical edge caching optimization method based on drone cooperation is implemented, including the following steps: S1, under the constraints of the UAV's buffer capacity and coverage range, establishes an optimization problem to minimize the total latency of system video transmission; The optimization problem of minimizing the total video transmission delay of the system is defined as P1: in, Representing binary optimization variables respectively , This indicates that all video quality can be obtained from the cloud video library via the base station. This means that each ground user is associated with a drone. This indicates a constraint on the drone's cache capacity. The collision avoidance distance between the two drones is... The maximum communication distance is , This indicates the flight altitude limit for the drone.

[0033] S2, based on the constraints of drone cache capacity and coverage, decompose the optimization problem established in S1 into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; The optimized drone caching strategy includes the following prerequisite: the actual cached content of the drone is less than the maximum cache capacity.

[0034] The sub-problem of optimizing user-drone association and drone collaborative association is defined as P2: The drone caching strategy optimization sub-problem is defined as P3: The drone service hovering position optimization sub-problem is defined as P4: S3 solves the subproblems, obtains the optimal solutions to each subproblem, and finds the global optimal solution through alternating iterations, thereby minimizing the total video transmission delay of the system.

[0035] The process of solving subproblems includes: The greedy search method is used to solve the subproblems of user drone association and drone cooperative association optimization. The dynamic programming method is used to solve the subproblem of optimizing the drone caching strategy; We use the particle swarm optimization method to solve the hovering position optimization subproblem of UAV services.

[0036] Specifically: 1. The process of solving the sub-problems of user drone association and drone cooperative association optimization using the greedy search method includes: Taking into account both the associated payload of the drone and the signal-to-noise ratio of the collaborative drones, two constraints are introduced, and problem P2 is rewritten as follows: in It is for users transport layer The time delay.

[0037] First, based on the distance between the user and the drone and the associated load of the drone, the associated access drone is determined for the ground user. If the access drone encounters a cache miss when making a video quality request to the ground user, the optimal cooperating drone is found for the access drone based on the cache hit status of other drones, communication distance, and SNR threshold.

[0038] 2. Using dynamic programming, solve the subproblem of optimizing the drone caching strategy to obtain the drone's caching strategy, including the following steps: Caching decisions for each drone The solution can be modeled as a classic 0-1 knapsack problem, where the buffer capacity of each drone is considered. Considered as backpack capacity, all video quality levels are layered. Treated as an item, layer Size Assuming the weight of the item, the drone caching layer The reduction in the total system latency before and after is considered the value of the item; the value of the item is expressed as: drones caching layer The greater the reduction in total system latency before and after, the better for items (layers). The greater the value, the smaller the total system latency. Therefore, the problem of minimizing the total system latency is equivalent to the problem of maximizing the value of all backpacks (drones).

[0039] Using the objective function transformation, problem P3 is simplified and re-expressed as P6: This embodiment uses dynamic programming with dependency constraints to solve the problem. The number of videos in the video library is... Video collection This indicates that each video Encoded as Therefore, the video library has 10 layers. Each layer uses Representation. Define state variables. Indicates before consideration There are 1 layer, and the remaining capacity of the drone is 1. At that time, drones The greatest value for the current layer. Check if the drone's current cache list contains a layer. The dependency layer.

[0040] If the drone's current cache list contains a layer The dependency layer, and the current remaining capacity of the drone. Greater than or equal to the current layer Size This means that the remaining capacity of the current drone can be used for caching. ,calculate And compare the current remaining capacity. Lower caching layer The value of the drone before and after. If ,make and update the current remaining capacity. Value of drones .like ,make , If the drone's current cache list does not contain the layer... The dependency layer, then , .

[0041] If the current drone has remaining capacity Smaller than the current layer Size This indicates that the current drone's remaining capacity cache is not being lowered. ,but , .

[0042] 3. Solve the UAV service hovering position optimization subproblem using the particle swarm optimization method, including the following steps: drones The three-dimensional position is represented as Total The drone has a particle dimension of 1. Let the particle swarm size be The first particle Initial positions obtained using the K-means clustering algorithm Other particles The position is generated using random perturbations near the initial position. .according to Three-dimensional position of a drone Calculate the expected total delay of the system, and consider the particle with the smallest expected total delay as the best historical position of the population. Each particle is based on its individual historical best position. and the group's historical global optimal position Update velocity and position; the three-dimensional velocity update formula is expressed as: in, , as well as They represent particles respectively Chinese drones In dimensions Upper The speed of each iteration; , as well as They represent particles respectively Chinese drones In dimensions Upper The position of the next iteration; It is the inertia factor, which controls the influence of historical speed on current speed. These are learning factors that control the influence of individual experience and group experience, respectively. yes Uniformly distributed random numbers within a range increase the randomness of the exploration.

[0043] The space constraints for particle exploration are satisfied as follows: 4. The process of finding the global optimal solution through alternating iteration is as follows: by alternately solving problems P2, P3 and P4, the optimal solutions for user and drone access association, drone cooperation association, drone caching strategy and drone service hovering position are obtained; The specific steps for finding the global optimum through alternating iterations include: Step 1, Initialization: Set the maximum number of iterations Convergence threshold Current iteration number The k-means clustering method is used to solve for the initial association and initial location of the user's drone, and records are... ; Step 2: Use a greedy search method to solve for the user-drone association variables and the collaboration association variables among drones. ; Step 3: Solve the drone caching strategy using dynamic programming. ; Step 4: Use the particle swarm optimization method to solve for the UAV's position. ; Step 5: Increment the iteration count by 1; Step 6: Check if the termination condition is met; if the termination condition is met, the iteration ends; otherwise, return to step 2 to continue the iteration. The termination condition includes: reaching the maximum number of iterations. or ,in, and They represent the first Second and third Total system latency for each iteration.

[0044] Based on a similar inventive concept, embodiments of the present invention also provide a computer storage medium storing a readable program that, when run by a processor, can execute the above-described SVC hierarchical edge caching optimization method based on UAV collaboration.

[0045] Based on a similar inventive concept, this invention provides an electronic device, including: a processor, a memory, a communication interface, and a communication bus, wherein the processor, the memory, and the communication interface communicate with each other through the communication bus; The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the above-described SVC hierarchical edge caching optimization method based on UAV cooperation.

[0046] Based on a similar inventive concept, embodiments of the present invention also provide a computer program product, including computer instructions, which instruct a computing device to perform the operations corresponding to the above-described SVC hierarchical edge caching optimization method based on UAV collaboration.

[0047] Example 3 In this embodiment, to verify the performance of the proposed UAV-based SVC hierarchical edge caching optimization method, it is compared with the following three algorithms: Compared to Algorithm 1: There is no collaboration between drones. When the drone associated with the ground user does not have a cache hit, the video is directly transmitted from the cloud through the base station. Other variable optimizations use the same method as in this paper. Comparison with Algorithm 2: The drone's caching strategy uses a fixed video popularity-based caching strategy, and other variable optimizations use the same method as in this paper; Algorithm 3: The drone's service hovering position is fixed and evenly distributed above the user area. Other variables are optimized using the same method as in this paper.

[0048] Figure 2 The method of this invention was compared with Comparison Algorithm 1, Comparison Algorithm 2, and Comparison Algorithm 3 under the same number of users to measure the average system latency. All simulations and experiments in this study were conducted on a fixed hardware platform equipped with an Intel® Core™ Ultra 7 155H processor with a base frequency of 3.80 GHz. The system was equipped with 32.0 GB of onboard memory and ran at a speed of 7467 MT / s. The storage device was a 954 GB solid-state drive. The system ran on Windows 11. All simulations and performance tests were performed in the PyCharm environment to ensure consistency of experimental conditions and reproducibility of results. The comparative experiments were conducted with uniform bandwidth power and relevant channel parameter settings, using four drones, each with a buffer capacity of... The system was set to 100 Mbit, with 30 videos and each video encoded in 3 layers. The base layer, enhancement layer 1, and enhancement layer 2 were set to 14 Mbit, 10 Mbit, and 6 Mbit, respectively. As shown in the graph, the average system latency of all four algorithms increases with the number of users. This is because, with fixed bandwidth and power resources, the bandwidth and power allocated to each user decreases as the number of users increases, leading to increased system transmission latency. Algorithm 2 has the highest average system latency because, under the video popularity-based caching strategy, the videos cached by the drones are concentrated in those with high popularity. Faced with random video requests from ground users, the cache hit rate for videos with lower popularity is low, and the video popularity-based caching strategy does not better adapt to users' random video requests. Algorithm 1 has the second highest average system latency because, without drone cooperation, when faced with more users' random video requests, the ground users' associated drones do not have cache hits, and the transmission latency is considerable as they can only obtain the data from the cloud via the base station. Algorithm 3 has a slightly higher average system latency compared to the algorithm of this invention. This is because the fixed drone positions, which are uniformly distributed across the user area, are not the optimal hovering positions for drone services. This increases the distance between the drone and the ground user, thereby reducing the transmission rate and increasing latency. Compared to the other three comparison algorithms, the algorithm of this invention allows drones to cooperate and optimizes drone caching strategies and drone positions, thus achieving the best system performance. Figure 3 The performance curves of the algorithm of this invention, showing the average system latency as a function of the number of users, are presented. The more drones there are, the lower the average system latency. This is because increasing the number of drones reduces the number of ground users associated with each drone, thereby increasing the transmission rate of the communication link between the drone and the ground users. Furthermore, the total system buffer capacity increases with the number of drones, improving the system buffer hit rate and thus reducing the average system latency. When the number of users is 20, the average system latency for drones with 3, 4, and 5 drones is not significantly different. This is because the average system latency for 20 users is inherently low, and 3 drones are sufficient to help the system transmit video to users.

[0049] Figure 4The performance curves of the algorithm of this invention, showing the average system latency as a function of the number of users under different drone buffer capacities, are presented. The larger the drone buffer capacity, the lower the average system latency. This is because increasing the drone buffer capacity allows the drone to cache more video, thereby improving the buffer hit rate, reducing the number of transmissions from the base station to ground users, and thus reducing the average system latency. When the number of users is 20, the average system latency for drone buffer capacities of 80 Mbit, 100 Mbit, and 120 Mbit is not significantly different. This is because the average system latency for 20 users is inherently low, and a drone buffer capacity of 80 Mbit is sufficient for the system to transmit video to users.

[0050] Figure 5 The performance curves of the algorithm of this invention as a function of the number of users under different video popularity factors are presented. The higher the video popularity factor, the lower the system average latency. This is because increasing the video popularity factor leads to a more concentrated video popularity, resulting in more concentrated requests from ground users. This improves the effectiveness of the UAV caching strategy optimization, thereby increasing the cache hit rate, reducing the number of transmissions from the base station to ground users, and thus reducing the system average latency. When the number of users is 20, the system average latency is already very small; therefore, the difference in system average latency under different video popularity factors is not significant.

[0051] Figure 6 The performance curves of the algorithm of this invention, showing the average system latency as a function of the number of users under different numbers of videos, are presented. The larger the number of videos, the greater the average system latency. This is because increasing the number of videos reduces the drone's cache hit rate and increases the number of base station transmissions, thus increasing the average system latency. When the number of users is 20, the average system latency is very small, and the change in average system latency under different numbers of videos is not significant.

[0052] Example 4 This embodiment proposes an SVC hierarchical edge caching optimization device based on UAV collaboration, specifically including: Problem building module: Under the constraints of UAV's buffer capacity and coverage range, an optimization problem is established to minimize the total latency of system video transmission; Problem decomposition module: Based on the constraints of drone cache capacity and coverage, the optimization problem is decomposed into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; Problem-solving module: Solve subproblems, obtain the optimal solutions to each subproblem, and find the global optimal solution through alternating iterations, thereby minimizing the total video transmission delay of the system.

[0053] The methods of the present invention can be implemented in hardware, firmware, or as software or computer code that can be stored in a recording medium (such as a CD-ROM, RAM, floppy disk, hard disk, or magneto-optical disk), or as computer code originally stored on a remote recording medium or a non-transitory machine-readable medium and subsequently stored on a local recording medium, downloaded via a network. Thus, the methods described herein can be processed by software stored on a recording medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware (such as an ASIC or FPGA). It is understood that the computer, processor, microprocessor controller, or programmable hardware includes storage components (e.g., RAM, ROM, flash memory, etc.) capable of storing or receiving software or computer code that, when accessed and executed by the computer, processor, or hardware, implements the methods described herein. Furthermore, when a general-purpose computer accesses the code used to implement the methods shown herein, the execution of the code transforms the general-purpose computer into a dedicated computer for performing the methods shown herein.

[0054] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. A hierarchical edge caching optimization method for SVC based on UAV collaboration, characterized in that, The system to which the optimization method is applied includes: a remote server and a ground base station. A drone and A ground user; during the drone service phase, the drone remains hovering, and the ground user initiates a video quality request to the associated drone. The associated drone selects the video transmission link mode based on the cache hit status; the optimization method includes: Given the constraints of the drone's buffer capacity and coverage area, we establish an optimization problem to minimize the total latency of system video transmission. Based on the constraints of drone cache capacity and coverage, the optimization problem is decomposed into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; Solve the subproblems, obtain the optimal solutions for each subproblem, and find the global optimal solution through alternating iterations, so as to minimize the total video transmission delay of the system.

2. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 1, characterized in that, The optimization problem P1 for minimizing the total video transmission delay of the system is: in, It refers to the number of ground users. It is the number of drones. It refers to the number of videos. It is the number of encoding layers in the video. Is this a video request? The The probability of the layer, It is for users Transmit video The Layer delay, It is a user-drone association variable. These are the variables that correlate drones. It is a decision variable for drone caching. It is a base station caching decision variable. It is the three-dimensional position variable of the UAV; Representing binary optimization variables respectively , This indicates that all video quality can be obtained from the cloud video library via the base station. This means that each ground user is associated with a drone. This indicates a constraint on the drone's cache capacity. The collision avoidance distance between the two drones is... The maximum communication distance is , It is a drone The three-dimensional position, It is a drone The three-dimensional position, Indicates the flight altitude limit of the drone. This indicates the drone's flight altitude; the minimum altitude is... The highest height is .

3. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 2, characterized in that, The optimization sub-problem P2 for user drone association and drone cooperative association is: The sub-problem P3 for optimizing the drone caching strategy is: The sub-problem P4 for optimizing the hovering position of the drone service is: 。 4. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 3, characterized in that, The process of solving the subproblem includes: The greedy search method is used to solve the subproblems of user drone association and drone cooperative association optimization. The dynamic programming method is used to solve the subproblem of optimizing the drone caching strategy; We use the particle swarm optimization method to solve the hovering position optimization subproblem of UAV services.

5. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 4, characterized in that, The process of solving the optimization subproblems of user drone association and drone cooperative association using the greedy search method includes: Introducing two constraints, problem P2 is rewritten as P5: in, This indicates the maximum associated payload for each drone. , Indicating collaborative drones For drones The signal-to-noise ratio of communication must meet a threshold condition. , It is the total transmission power for communication between drones. It is a drone The number of other drones that need to cooperate simultaneously yes arrive Channel gain, It is noise power; Greedy search is used to find the optimal solution for user drone association and drone collaboration association.

6. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 4, characterized in that, The process of solving the subproblem of drone caching strategy optimization using dynamic programming includes: Using the objective function transformation, problem P3 is simplified and re-expressed as P6: in, It is a drone caching layer The reduction in total system latency before and after is considered as the value of the item. By modeling the caching strategy optimization problem of a single UAV as a 0-1 knapsack problem, the optimal caching strategy for the current UAV is solved using dynamic programming in a random order, combined with the current caching strategies of other UAVs. The optimal caching strategy for all drones was obtained through multiple iterations.

7. The SVC hierarchical edge caching optimization method based on UAV collaboration according to claim 4, characterized in that, The specific steps for finding the global optimal solution through alternating iterations include: Step 1, Set the maximum number of iterations Convergence threshold Current iteration number The k-means clustering method was used to solve for the initial association and initial location of the user's drone, and the results were recorded. ; Step 2: Use a greedy search method to solve for the user-drone association variables and the collaboration association variables among drones. ; Step 3: Solve the drone caching strategy using dynamic programming. ; Step 4: Use the particle swarm optimization method to solve for the UAV's position. ; Step 5: Increment the iteration count by 1; Step 6: Check if the termination condition is met; if the termination condition is met, the iteration ends; otherwise, return to step 2 to continue the iteration. The termination condition includes: reaching the maximum number of iterations. or ,in, and They represent the first Second and third Total system latency for each iteration.

8. A drone-based collaborative SVC hierarchical edge caching optimization device, performing the method according to any one of claims 1-7, characterized in that, include: Problem building module: Under the constraints of UAV's buffer capacity and coverage range, establish an optimization problem to minimize the total latency of system video transmission; Problem decomposition module: Based on the constraints of drone cache capacity and coverage, the optimization problem is decomposed into: user drone association and drone collaboration association optimization sub-problem, drone cache strategy optimization sub-problem, and drone service hovering position optimization sub-problem; Problem-solving module: Solve subproblems, obtain the optimal solutions to each subproblem, and find the global optimal solution through alternating iterations, thereby minimizing the total video transmission delay of the system.

9. A computer storage medium storing a readable program, characterized in that, When the program runs, it can instruct the computing device to perform the SVC hierarchical edge caching optimization method based on UAV collaboration as described in any one of claims 1-7.

10. An electronic device, characterized in that, include: The processor, memory, communication interface, and communication bus are provided, wherein the processor, memory, and communication interface communicate with each other via the communication bus. The memory is used to store at least one executable instruction, which causes the processor to perform the operation corresponding to the SVC hierarchical edge caching optimization method based on UAV cooperation as described in any one of claims 1-7.

Images