Unmanned aerial vehicle assisted V2V cache network optimization method for vehicle dynamic distribution

Abstract

The invention discloses a vehicle dynamic distribution-oriented unmanned aerial vehicle assisted V2V cache network optimization method, which comprises the following steps of: 1, designing a multi-unmanned aerial vehicle assisted hierarchical cache strategy, arranging files in a file library in a descending order from high to low according to popularity, and sequentially dividing the files into a core layer, a supplementary layer and a low-frequency layer, the cache mode of the core layer file is determined cache for each vehicle, the cache mode of the supplementary layer file is vehicle probability cache, and the low-frequency layer file is not cached by the vehicle; when a vehicle sends a file request, establishing a self cache and a V2V cache, and obtaining a request file three-level response from the cloud server through the unmanned aerial vehicle; 2, analyzing the average service time delay of the cache performance; step 3, jointly optimizing a cache strategy and unmanned aerial vehicle deployment; according to the invention, an efficient file distribution system with local priority, secondary cooperation and cloud supplementation is formed, the number of the unmanned aerial vehicles is optimized, and the cost and the performance are balanced.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a V2V caching network optimization method. Background Technology

[0002] In traditional terrestrial cellular networks, static base stations, limited by fixed deployment locations and resource allocation, struggle to effectively adapt to the highly dynamic and spatiotemporally uneven content requests from vehicles on urban roads, easily leading to regional network congestion and service interruptions during specific time periods or road sections. To improve content distribution efficiency, various new network architectures have emerged in recent years. Among them, drones, with their high mobility and rapid deployment capabilities, can act as aerial relays, providing vehicles with flexible line-of-sight links and edge services. Simultaneously, vehicles themselves possess communication and storage capabilities, enabling local storage and shared distribution of cached content within a certain range through vehicle-to-vehicle (V2V) communication, forming a V2V caching mechanism. Traditional network architectures relying on fixed base stations or single drones suffer from coverage blind spots and hotspot capacity bottlenecks, and treat vehicles merely as service request terminals, ignoring their potential as mobile relays and caching nodes. Existing solutions lack quantitative analysis of the dynamic distribution characteristics of vehicles, failing to effectively organize and utilize the edge caching transmission advantages of vehicles.

[0003] The following two existing technologies are as follows: Technical solution [1] "UAV-assisted content caching for human-centered consumer applications in the Internet of Vehicles", published in IEEE Consumer Electronics Bulletin, February 2024, Vol. 70, No. 1, pp. 927-938, Digital Object Identifier (doi): 10.1109 / TCE.2023.3349079.

[0004] Technical solution [2] "Optimal device-to-device cooperative caching system in wireless networks based on software-defined networks", included in the 30th Annual IEEE International Conference on Personal, Indoor and Mobile Radio Communications (PIMRC) held in Istanbul, Turkey in 2019, pages 1-7, digital object identifier (doi): 10.1109 / PIMRC.2019.8904149.

[0005] Considering the efficiency and reliability of content distribution, in order to improve the cache hit rate and reduce the dependence on the backhaul link, the technical solution [1] uses drones to assist in content caching, caching the content on drones to reduce the dependence on cloud servers. However, the fact that all content is cached by drones may lead to excessively long latency for users to apply for caching and reduced service quality.

[0006] Considering the hit rate of user-requested caches, the technical solution [2] constructs a D2D-assisted collaborative caching network, which expands the available cache resources by sharing content among neighboring devices, but cannot guarantee that all users can obtain the files requested for caching. Summary of the Invention

[0007] In view of this, the purpose of this invention is to provide an unmanned aerial vehicle-assisted V2V caching network optimization method for vehicle dynamic distribution, which improves the utilization rate of vehicle caching resources while ensuring that the latency of user requesting caching is small.

[0008] To achieve the above objectives, the present invention adopts the following technical solution: This invention provides an optimization method for a drone-assisted V2V caching network oriented towards dynamic vehicle distribution. It is applied to a file caching system between dynamic vehicles on a lane, drones deployed in the air above the lane, and a cloud server. The file caching system comprises a file library containing all files. Step 1: Design a multi-drone-assisted hierarchical caching strategy: Files in the file library are arranged in descending order of popularity from high to low, and are divided into a core layer, a supplementary layer, and a low-frequency layer. The core layer files are cached by a specific cache for each vehicle, the supplementary layer files are cached probabilistically by the vehicle, and the low-frequency layer files are not cached by any vehicle. Within a set distance, two vehicles can establish a V2V communication link to share their own cached files. When a vehicle sends a file request, it first checks whether it can respond through its own cache. If not, it checks whether it can respond through the V2V cache. If neither is possible, the drone will retrieve the requested file response from the cloud server, forming a three-level response. Step 2: Analyze cache performance and average service latency: Calculate the probability of each of the three response levels, establish latency models for each response mode, and calculate the average service latency of the system. Step 3: Jointly optimize caching strategy and drone deployment: Optimize with the goal of minimizing the average service latency of the system to obtain the core layer ratio, supplementary layer ratio, caching probability of supplementary layer, number of drones, and drone deployment location.

[0009] Preferably, the optimization constraints in step three are as follows: the core layer ratio and the supplementary layer ratio are in the range of 0-1, and the corresponding number of files is an integer; the number of drones is between 1 and the total number of vehicles in the road segment; the deployment location of drones is limited to the target road segment; and the sum of the core layer file size and the supplementary layer probability cache file size is equal to the total cache capacity of the vehicles.

[0010] Preferably, step three decomposes the optimization problem into a caching subproblem and a drone deployment subproblem. For the caching subproblem, the optimal value is determined by traversing the combination of hierarchical ratio parameters that satisfy the constraints through grid search, and the probability of supplementary layer file caching is optimized by function fitting method. For the drone deployment subproblem, the optimal number of drones is determined by a clustering algorithm based on service demand weighting and combined with elbow strategy, and the cluster center is used as the horizontal coordinate position of drone deployment.

[0011] Preferably, a single-lane road in one direction is selected, with a road segment length L and a number of vehicles J in the segment. The position of the vehicle is recorded as Vehicles arriving at the road segment entrance follow a Poisson distribution, and each vehicle is assumed to travel at a constant speed. Follow the interval The uniform distribution of drones on the surface, the number of drones deployed in the air , No. The location of the drone is recorded as The coordinates of the drone deployment location are x, y, y. = The drones are deployed at the same altitude, and the default vehicle and drone position coordinates are... Similarly, the system is a time-slot operating system, and the duration of each time slot is... , It is small enough that the positions of the drones and vehicles remain unchanged within a single time slot.

[0012] Preferably, the file library is F, containing N files. The documents are arranged in descending order of popularity, and the popularity distribution follows a certain pattern. Distribution, the nth file in the file library popularity for: ; in, This is used to control the concentration of popularity; all file sizes are 1, and files can be successfully transmitted within a single time slot.

[0013] Preferably, the set distance is a V2V communication distance threshold. , , .

[0014] Preferably, the cache space size of each vehicle is defined to be the same, and the cache space of each vehicle can store... One file, The core layer ratio is The ratio of supplementary layers is The caching probability of the supplementary layer is Let K be the number of drones, U be the x-coordinate of each drone, and let the set of file cache probabilities for the supplementary layer be denoted as K. The size of the low-frequency layer is Cache space constraints .

[0015] Preferably, the average service latency of the system is calculated by weighting and summing the probabilities of the three response modes with the corresponding delays.

[0016] Preferably, the probability that a file request for the target vehicle is obtained through its own cache is: ; The probability that a file request for the target vehicle will be responded to via a V2V cache is: ; in, Indicates that, with the target vehicle as the center, Cache files within a radius The number of vehicles, since vehicle arrival is a Poisson process and their speeds are independently distributed, the spatial distribution of vehicles constitutes a Poisson point process centered on the target vehicle. Number of vehicles within a radius Follows a Poisson distribution: ; but They also follow a Poisson distribution: ; When the target vehicle cannot obtain the requested file through the above two methods, the drone will provide the service through the cloud server. The probability that the target vehicle will be served by the drone is: ; Different channel models are established for signal transmission under different buffering methods. Ignoring hardware interference, the latency of the target vehicle's response through its own buffer is: ; The latency for the target vehicle to respond via the V2V cache is: ; in, This represents the channel capacity of the V2V buffered response. The V2V buffered response enables parallel transmission by multiple users through orthogonal frequency division multiple access (OFDMA) technology. for: ; in, This represents the signal-to-noise ratio of the V2V buffer response. The V2V channel is modeled as a Rayleigh channel. for: ; in, For the vehicle's transmission power, For Rayleigh's decline, for Communication distance, The power spectral density of additive white Gaussian noise, The latency for the target vehicle to respond via drone is: ; Among them, the channel capacity for UAV transmission to target vehicles for: ; in, The signal-to-noise ratio (SNR) represents the drone's response. The transmission between the drone and the target vehicle is line-of-sight propagation. for: ; in, For the drone's transmission power, This is the transmission loss factor of the channel. The distance between the drone and the target vehicle; The average service latency for the target vehicle file request is: .

[0017] Preferably, the caching probability for supplementary layer files An exponential function was used for fitting: ; Where, constant The calculation formula is: .

[0018] The beneficial effects of this invention are as follows: This invention adopts a hierarchical caching strategy based on file popularity. Popular files in the core layer are forcibly cached, medium-popular files in the supplementary layer are cached with probability, and unpopular files are not cached in the vehicle cache. This maximizes the hit probability of the vehicle's own cache and V2V collaborative cache, reducing reliance on drones and the cloud. This invention designs a three-level response for file caching, prioritizing low-latency local caching and short-range V2V communication to respond to requests, and only enabling drone supplementary responses when two levels of caching fail. This significantly reduces the average service latency of the system from the perspective of response path optimization, forming an efficient file distribution system that prioritizes local caching, followed by collaborative caching, and supplemented by cloud caching. It also optimizes the number of drones. This invention's optimization method improves the utilization rate of vehicle cache resources while ensuring low latency for user cache requests, achieving a balance between cost and performance. Attached Figure Description

[0019] The above and other objects, features and advantages of the present invention will become clearer from the following description of embodiments of the invention with reference to the accompanying drawings.

[0020] Figure 1 This is a flowchart of a specific embodiment of the present invention; Figure 2 This is a schematic diagram of a file caching system applied in a specific embodiment of the present invention; Figure 3 This is a schematic diagram of the decoupling iteration method used in a specific embodiment of the present invention; Figure 4 A schematic diagram illustrating the caching probability of each layer in a specific embodiment of the present invention; Figure 5 This is a schematic diagram of the vehicle response status distribution and UAV deployment location in a specific embodiment of the present invention; Figure 6 This is a schematic diagram of the optimal number of clusters for vehicle clustering in a specific embodiment of the present invention. Detailed Implementation

[0021] The present invention will be described below based on embodiments. Those skilled in the art should understand that the accompanying drawings provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0022] Unless the context explicitly requires it, the words "comprising," "including," and similar terms throughout the specification and claims should be interpreted as encompassing rather than being exclusive or exhaustive; that is, meaning "including but not limited to."

[0023] like Figure 1 The multi-UAV dynamic-assisted V2V caching network optimization method shown is applied to a file caching system (e.g., between dynamic vehicles on the lane, UAVs deployed in the air above the lane, and cloud servers) Figure 2 As shown in the figure, the road segment in this embodiment is a one-way single lane with a length L = 10 km. The number of vehicles in this road segment is J = 100. The position of the vehicle is recorded as Since it's a single-lane road, only the change in the vehicle's abscissa is considered. Vehicle arrivals at the road segment entrance follow a Poisson distribution, with an average arrival rate of... Vehicles per hour. Each vehicle travels at a constant speed. Follow the interval The number of drones deployed in the air is evenly distributed. To provide data services for ground vehicles, the first The location of the drone is recorded as The coordinates of the drone deployment location are U= The drone's longitudinal coordinates are aligned with the vehicle's for ease of service; the drones are deployed at the same altitude. Because the drone is positioned above the lane, the default vehicle position and drone position coordinates are... They are identical and equal to 0. The system is a time-slot operating system, and the duration of each time slot is... Assuming It is small enough that the positions of the drones and vehicles remain unchanged within a single time slot.

[0024] The file caching system involves a file library, denoted as F, which contains... One file, The files in the file library are arranged in descending order of popularity, and the popularity distribution of the files follows a certain pattern. of Distributed, nth file popularity Represented as: ; , A higher γ value indicates that vehicles are more likely to request popular files from the file library. When γ=0, all files have an equal probability of being requested. This is a uniform distribution, representing that users have no preference for any particular file and choose files completely randomly. γ=0.8 represents a value where popular files are concentrated but not extremely monopolized, which is more in line with actual user behavior. Setting all file sizes to 1 ensures successful file transfer within a single time slot.

[0025] In the above application scenario, the caching network optimization method in this specific embodiment includes the following steps: Step 1: Design a multi-drone-assisted hierarchical caching strategy; Step 2: Analyze cache performance and average service latency; Step 3: Jointly optimize caching strategies and drone deployment.

[0026] Specifically as follows: (1) Step 1 Utilizing the vehicle's cache space, a hierarchical caching strategy is designed, dividing the files in the file library into three layers: a core layer, a supplementary layer, and a low-frequency layer. The caching methods for these three layers differ, with the core layer having a size of [missing information]. , This indicates the proportion of core layer files in the file library; the caching method is deterministic caching, and the files are stored in the cache space of each vehicle; the supplementary layer size is... , This indicates the proportion of supplementary layer files in the file library, with probabilistic caching as the caching method, and the probability of a file being cached in the vehicle's cache space is [value missing]. The set of file caching probabilities for the supplementary layer is denoted as The size of the low-frequency layer is The file is not cached by the vehicle. The cache space size is defined to be the same for all vehicles, which can store... If there are multiple files, then there is a cache space size constraint. Furthermore, if the distance between the two vehicles is less than... Then, a V2V communication link can be established between the two vehicles, allowing them to share their own cache files.

[0027] (2) Step Two When a vehicle sends a file request, it first checks if it can respond through its own cache. If not, it checks if it can respond through a V2V cache. If neither is possible, the requested file will be retrieved from the cloud server via the drone.

[0028] The probability that a file request for the target vehicle is obtained through its own cache is: ; The probability that a file request for the target vehicle will be responded to via a V2V cache is: ; in, Indicates that, with the target vehicle as the center, Cache files within a radius The number of vehicles. Since vehicle arrival is a Poisson process and the velocities are independently distributed, the spatial distribution of vehicles constitutes a Poisson point process. Centered on the target vehicle, Number of vehicles within a radius Follows a Poisson distribution: ; but They also follow a Poisson distribution: ; When the target vehicle cannot obtain the requested file through the above two methods, the drone will provide the service through the cloud server. The probability that the target vehicle will be served by the drone is: ; Different channel models are established for signal transmission under different buffering methods. Ignoring hardware interference, the latency of the target vehicle's response through its own buffer is: ; The latency for the target vehicle to respond via the V2V cache is: ; in, This indicates the channel capacity of the V2V buffered response. V2V buffered responses enable multi-user parallel transmission using orthogonal frequency division multiple access (OFDMA) technology. for: ; in, This represents the signal-to-noise ratio (SNR) of the V2V cache response. The calculation formula is: ; In this embodiment, the vehicle's transmission power , Power spectral density of additive white Gaussian noise for Rayleigh fading , The average communication distance is .

[0029] The latency for the target vehicle to respond via drone is: ; Among them, the channel capacity for UAV transmission to target vehicles for: ; in, This represents the signal-to-noise ratio (SNR) of the drone's response. The transmission between the drone and the target vehicle is line-of-sight propagation, and the SNR... for: ; Among them, the vehicle user terminal transmit signal power The transmission loss factor of the channel , This refers to the distance between the drone and the target vehicle.

[0030] In summary, the average service latency for the target vehicle file request is: ; (3) Step Three Jointly optimize the layering parameters of each file layer ( , ), cache probability of supplementary layer ( ) and drone deployment ( , The following latency minimization problem is established: This represents seven constraints. The range of values ​​for the cache tiering ratio parameter is limited. Integer constraints for cache space hierarchies Due to cache space size constraints, Constraints on the value of cache probability To impose quantity constraints on the dynamic deployment of drones. It refers to the location constraints for the dynamic deployment of drones.

[0031] The optimization problem is decoupled into two sub-problems: the caching sub-problem and the drone deployment sub-problem. Optimization is then performed using a decoupled iterative method, such as... Figure 3 As shown.

[0032] Regarding the caching subproblem, constrained Integer limit for cache space tiering, cache tiering parameters , The possible combinations of values ​​are finite, so a grid search method is used to satisfy the constraints. - Traverse within the finite solution space to obtain , The optimal solution: Therefore, the core layer contains 11 files, the supplementary layer contains 46 files, and the low-frequency layer contains 43 files. The caching probability for the supplementary layer files is as follows. The optimization method employs a tailed distribution function fitting approach, comparing the average service delay obtained by fitting linear, power-law, logarithmic, and exponential functions. The exponential function was determined to be the fitter: ; Where, constant The calculation formula is: ; The final fitting yielded the following results and obtained Figure 4 The curve shown: ; Based on the optimization results of the caching subproblem, for the drone deployment subproblem, the number of drones is determined using a service demand-weighted K-means clustering method. (and drone deployment locations).

[0033] The K-means clustering method based on the elbow strategy is used to cluster vehicles in the road segment, and the number of clusters corresponds to the number of drones. The cluster center corresponds to the horizontal coordinate position of the UAV. Among them, as the number of clusters ( The increase of ) and the weighted sum of squared distances ( As the weighted sum of squared distances gradually decreases, the optimal number of clusters can be determined using the elbow strategy. The formula for calculating the weighted sum of squared distances is as follows: in, For the first Clusters, The cluster center is the location of the inflection point where the rate of decrease of the weighted sum of squared distances significantly slows down. The value is the optimal cluster number, such as Figure 6 As shown.

[0034] Get within the service section Vehicles need to be deployed A drone, After joint consideration, the optimal deployment locations for the drone swarm are finally obtained as follows: , After deployment, the auxiliary cache results are as follows: Figure 5 As shown.

[0035] Figure 5 This embodiment demonstrates the three-level response effect: Self-cached hit: 49 (49.0%): Approximately half of the requests can be responded to directly through the vehicle's own cache, indicating that the layered strategy of mandatory caching in the core layer and probabilistic caching in the supplementary layer is effective, allowing the local coverage of popular files to reach nearly 50%, achieving the low latency goal of prioritizing self-caching.

[0036] V2V cache hit: 20 (20.0%): 20% of requests share cache responses through vehicle-to-vehicle (V2V) communication, demonstrating the value of node collaboration in the Internet of Vehicles (IoV). Close-range shared cache between vehicles reduces reliance on drones and leverages the distributed nature of IoV, making it an important supplement to tiered caching strategies.

[0037] Drone hits: 31 (31.0%): Approximately 30% of requests required drones to retrieve files from the cloud in response. This percentage validates the necessity of drones as supplementary nodes and indicates that the first two levels of caching have handled nearly 70% of the requests, significantly reducing the pressure on direct cloud transmission.

[0038] Furthermore, only two drones are needed to achieve full coverage of the designated road section, avoiding the waste of resources caused by having too many drones.

[0039] It will be readily understood by those skilled in the art that, without conflict, the above-mentioned preferred solutions can be freely combined and superimposed.

[0040] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. For those skilled in the art, the present invention can be modified and varied in various ways. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principle of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for optimizing a drone-assisted V2V caching network for dynamic vehicle distribution, applied to a file caching system between dynamic vehicles on a lane, drones deployed in the air above the lane, and a cloud server, wherein the set of all files involved in the file caching system constitutes a file library, characterized in that: include: Step 1: Design a multi-drone-assisted hierarchical caching strategy: Files in the file library are arranged in descending order of popularity from high to low, and are divided into a core layer, a supplementary layer, and a low-frequency layer. The core layer files are cached by a specific cache for each vehicle, the supplementary layer files are cached probabilistically by the vehicle, and the low-frequency layer files are not cached by any vehicle. Within a set distance, two vehicles can establish a V2V communication link to share their own cached files. When a vehicle sends a file request, it first checks whether it can respond through its own cache. If not, it checks whether it can respond through the V2V cache. If neither is possible, the drone will retrieve the requested file response from the cloud server, forming a three-level response. Step 2: Analyze cache performance and average service latency: Calculate the probability of each of the three response levels, establish latency models for each response mode, and calculate the average service latency of the system. Step 3: Jointly optimize caching strategy and drone deployment: Optimize with the goal of minimizing the average service latency of the system to obtain the core layer ratio, supplementary layer ratio, caching probability of supplementary layer, number of drones, and drone deployment location.

2. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 1, characterized in that: The optimization constraints set in step three are as follows: the core layer ratio and the supplementary layer ratio are in the range of 0-1, and the corresponding number of files is an integer; the number of drones is between 1 and the total number of vehicles in the road segment; the deployment location of drones is limited to the target road segment; and the sum of the core layer file size and the supplementary layer probability cache file size is equal to the total cache capacity of the vehicles.

3. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 2, characterized in that: Step three decomposes the optimization problem into a caching subproblem and a drone deployment subproblem. For the caching subproblem, the optimal value is determined by traversing the combination of layer ratio parameters that satisfy the constraints through grid search, and the probability of supplementary layer file caching is optimized by function fitting method. For the drone deployment subproblem, the optimal number of drones is determined by a clustering algorithm based on service demand weighting and combined with elbow strategy, and the cluster center is used as the horizontal coordinate position of drone deployment.

4. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 1, 2, or 3, characterized in that: A single-lane road in one direction is selected, with a road length of L and a number of vehicles J in the road segment. The position of the vehicle is recorded as Vehicles arriving at the road segment entrance follow a Poisson distribution, and each vehicle is assumed to travel at a constant speed. Follow the interval The uniform distribution of drones in the air , No. The location of the drone is recorded as The coordinates of the drone deployment location are U= The drones are deployed at the same altitude, and the default vehicle and drone position coordinates are... Similarly, the system is a time-slot operating system, and the duration of each time slot is... , It is small enough that the positions of the drones and vehicles remain unchanged within a single time slot.

5. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 4, characterized in that: The file library is F, containing N files. The documents are arranged in descending order of popularity, and the popularity distribution follows a certain pattern. Distribution, the nth file in the file library popularity for: ； in, This is used to control the concentration of popularity; all files are 1 in size, and files can be successfully transmitted within a single time slot.

6. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 5, characterized in that: The set distance is the V2V communication distance threshold. , , .

7. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 5, characterized in that: All vehicles are defined to have the same cache space size, and each vehicle's cache space can store... One file, The core layer ratio is The ratio of supplementary layers is The caching probability of the supplementary layer is Let K be the number of drones, U be the x-coordinate of each drone, and let the set of file cache probabilities for the supplementary layer be denoted as K. The size of the low-frequency layer is Cache space constraints .

8. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 7, characterized in that: The average service latency of the system is calculated by weighting and summing the probabilities of the three response methods with the corresponding delays.

9. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 8, characterized in that: The probability that a file request for the target vehicle is obtained through its own cache is: ； The probability that a file request for the target vehicle will be responded to via a V2V cache is: ； in, Indicates that with the target vehicle as the center, Cache files within a radius The number of vehicles, centered on the target vehicle, Number of vehicles within a radius Follows a Poisson distribution: ； but They also follow a Poisson distribution: ； When the target vehicle cannot obtain the requested file through the above two methods, the drone will provide the service through the cloud server. The probability that the target vehicle will be served by the drone is: ； The latency for the target vehicle to respond using its own cache is: ； The latency for the target vehicle to respond via the V2V cache is: ； in, This represents the channel capacity of the V2V buffered response. The V2V buffered response enables parallel transmission by multiple users through orthogonal frequency division multiple access (OFDMA) technology. for: ； in, This represents the signal-to-noise ratio of the V2V buffer response. The V2V channel is modeled as a Rayleigh channel. for: ； in, For the vehicle's transmission power, For Rayleigh's decline, for Communication distance, The power spectral density of additive white Gaussian noise, The latency for the target vehicle to respond via drone is: ； Among them, the channel capacity for UAV transmission to target vehicles for: ； in, The signal-to-noise ratio (SNR) represents the drone's response. The transmission between the drone and the target vehicle is line-of-sight propagation. for: ； in, For the drone's transmission power, This is the transmission loss factor of the channel. The distance between the drone and the target vehicle; The average service latency for the target vehicle file request is: 。 10. The UAV-assisted V2V caching network optimization method for vehicle dynamic distribution according to claim 9, characterized in that: caching probability of supplementary layer files An exponential function was used for fitting: ； Where, constant The calculation formula is: 。

Images