A UAV-assisted vehicle network collaborative caching method, device and storage medium

Abstract

The present invention discloses a UAV-assisted vehicle network collaborative caching method, device and storage medium. The method can make full use of the limited cache resources in the vehicle self-organizing network and effectively reduce the delay of vehicle request content. It includes: firstly clustering the mobile vehicles, and finding the cluster center closer to the active users according to the activity of the vehicle users, that is, the optimal two-dimensional position of the UAV. Then, the flight trajectory and hovering time of the UAV are optimized according to the activity of the vehicle cluster, so that the UAV can better serve the ground vehicle users. Finally, the global content popularity is obtained according to the cluster activity and the popularity of the content within the cluster. Since the cache capacity of the UAV and the base station is limited, a target optimization problem of minimizing the delay is constructed, and then the particle swarm algorithm is used to solve the problem to obtain the optimal cache placement solution.

Description

Technical Field

[0001] The present invention relates to the field of communication technology, and in particular to a drone-assisted vehicle-mounted network collaborative caching method, device and storage medium. Background Art

[0002] With the rapid development of 5G and various applications, the wireless data traffic in vehicle networks has increased dramatically, mainly due to the repeated downloading of some popular content. Introducing edge caching technology in vehicle networks and caching the content that users are interested in on edge servers can effectively reduce the traffic load of vehicle networks and improve the driving experience of vehicle users. However, the caching capacity of vehicle networks is limited, and how to reasonably cache and effectively transmit content has become a challenge.

[0003] The traditional caching method in the vehicle network does not use drones to assist in caching the content in the vehicle network, which results in a high content request latency for the entire network and a low content request success rate. Summary of the invention

[0004] In view of the defects in the existing technology, the collaborative caching between drones and base stations should be considered to distribute the content in drones and base stations in a coordinated manner, so as to improve the diversity of cached content in the network and reduce redundancy. At the same time, the hovering position, hovering time, flight trajectory and content popularity of drones should be optimized to reduce the latency of vehicle users in obtaining content and improve the success rate of content requests.

[0005] Purpose of the invention: In order to overcome the shortcomings of the prior art, the purpose of the present invention is to provide a drone-assisted vehicle network collaborative caching method, device and storage medium, whereby the vehicle can obtain requested content from the drone and the base station. By applying this method, the content request latency and cost can be effectively reduced, while the successful content delivery rate in the network can be improved.

[0006] Technical solution: To solve the above technical problems, the technical solution adopted by the present invention is:

[0007] In a first aspect, the present invention provides a drone-assisted vehicle network collaborative caching method, comprising:

[0008] Step S1: Cluster the vehicles to obtain N C vehicle clusters;

[0009] Step S2: For each vehicle cluster, the hovering position of the UAV in the vehicle cluster is optimized according to the vehicle positions in the cluster and the activity of the vehicle users;

[0010] Step S3, optimizing the flight trajectory and hovering time of the UAV according to the activity of the vehicle cluster and the tolerance time of the vehicle users;

[0011] Step S4, optimizing the global content popularity according to the activity of the vehicle cluster and the popularity of the content within the cluster;

[0012] Step S5: Based on the optimized hovering positions of the drones in each vehicle cluster, the flight trajectories and hovering times of the drones, and the global content popularity, a particle swarm algorithm is used to solve the pre-constructed target optimization problem to obtain a cache placement solution;

[0013] Step S6: According to the obtained cache placement solution, the drone and the base station cache the vehicle request content.

[0014] In some embodiments, step S1: clustering the vehicles to obtain N C Vehicle clusters, including:

[0015] In the urban road scene, N V There are N mobile vehicles, and the mobile vehicles are divided into N C vehicle clusters, the jth (j∈{1,2,3,...,n}) vehicle in the i-th vehicle cluster is denoted by V i,j .

[0016] In some embodiments, step S2, for each vehicle cluster, optimizing the hovering position of the drone in the vehicle cluster according to the vehicle positions in the cluster and the activity of the vehicle users, includes:

[0017]

[0018] Among them, l i is the hovering position of the UAV in the i-th vehicle cluster, For vehicle V i,j activity, For vehicle V i,j The position of V i,j represents the jth vehicle in the i-th vehicle cluster, and n represents the total number of vehicles in the i-th vehicle cluster.

[0019] In some embodiments, step S3, optimizing the flight trajectory and hovering time of the drone according to the activity of the vehicle cluster and the tolerance time of the vehicle user, includes:

[0020] The activity of each vehicle cluster is w i , sort the activity of all vehicle clusters to get the activity of the vehicle cluster

[0021] The flight trajectory of the drone includes: based on the activity of the vehicle clusters, the drone first flies to the most active vehicle cluster, hovers at the hovering position of the vehicle cluster to provide content services to the vehicle, and then flies to the second most active vehicle cluster, and so on, until the tolerance time t of the vehicle user is exceeded;

[0022] The hovering time of the drone in each vehicle cluster is allocated according to the activity of the vehicle cluster to better serve the vehicle users. The hovering time of the drone in each vehicle cluster meets the following requirements:

[0023]

[0024]

[0025] in, are the hovering time of the UAV in each vehicle cluster according to the activity of the vehicle cluster; t UAV The total hovering time of the drone.

[0026] In some embodiments, step S4, optimizing the global content popularity according to the activity of the vehicle cluster and the popularity of the content within the cluster, includes:

[0027]

[0028] Among them, P is the global content popularity, R k is the activity of the vehicle cluster, O k The content popularity of each cluster.

[0029] In some embodiments, in step S5, the target optimization problem is represented as P1;

[0030]

[0031] T=α·T1+β·T2

[0032]

[0033]

[0034] st:a:,:={0,1}

[0035]

[0036]

[0037]

[0038] x min ≤x UAV ≤x max

[0039] y min ≤y UAV ≤y max

[0040] h>0

[0041] Among them, A Q,2 represents the cache placement matrix, T is the total content request latency of vehicle users; α represents the number of contents directly obtained from the drone, β represents the number of contents obtained from the base station, T1 represents the content request latency directly obtained from the drone, T2 represents the content request latency obtained from the base station, and the requested content is q, where q∈{1,2,3,...,Q}, Q is the total number of contents, and the content size is s q , is the transmission rate between the UAV and the vehicle user, ζ BS,UAV is the transmission rate between the drone and the base station, st represents the constraint condition, a :,: ={0,1} means that the elements in the cache placement matrix can only be 0 or 1, a q,: =1 means that the drone or base station has content q, a cached. q,: = 0 means that there is no content q cached in the drone or base station. The cache capacity of the base station and the drone is S respectively. BS and S UAV , the drone position is [x UAV ,y UAV ,h],x UAV ,y UAV 、x min 、x max ,y min ,y max They represent the two-dimensional horizontal and vertical coordinates of the UAV and the corresponding minimum and maximum values, respectively, and h represents the height of the UAV;

[0042] The first two constraints indicate that content cannot be cached repeatedly in the drone and base station. The third and fourth constraints indicate that the size of the cached content in the drone and base station cannot exceed their own cache capacity. The last three constraints limit the location of the drone.

[0043] In some embodiments, the method for constructing the target optimization problem includes:

[0044] The communication link between the drone and the ground is modeled by a probabilistic loss model. The path loss of the wireless backhaul link is determined by the line-of-sight LoS and non-line-of-sight NLoS link states. The path losses of LoS and NLoS are PL and , respectively. LoS , PL NLoS :

[0045] PL LoS =20log(4πfd / c)+ηLoS

[0046] PL NLoS =20log(4πfd / c)+η NLoS

[0047] Where d is the distance between the drone and the requesting vehicle or base station, f is the carrier frequency, c is the speed of light, and η LoS and η NLoS are the attenuation factors of LoS and NLoS links respectively;

[0048] The probability of LoS and NLoS links is denoted as P Los , P NLoS :

[0049]

[0050] P NLoS =1-P Los

[0051] Where θ is the elevation angle between the drone and the vehicle or base station, X and Y are constants that depend on the environment; the average path loss PL between the drone and the vehicle or base station is:

[0052] PL=PL LoS ×P Los +PL NLoS ×P NLoS

[0053] Transmission rate between drone and vehicle users for:

[0054]

[0055] in, is the bandwidth between the drone and the vehicle, is the received power of the vehicle, σ 2 is the noise power;

[0056] The transmission rate between the drone and the base station is BS,UAV for:

[0057]

[0058] in, is the bandwidth between the base station and the drone, P UAV is the received power of the UAV;

[0059] After a mobile vehicle sends a content request, different content acquisition methods will produce different request delays. The total content request delay of the vehicle user is T, the requested content is q, where q∈{1,2,3,...,Q}, and the content size is sq ;

[0060] When a mobile vehicle obtains content directly from a drone, the content request latency T1 obtained directly from the drone is:

[0061]

[0062] When the content requested by the vehicle user is not cached in the drone, the drone needs to send a content request to the base station. The base station checks the local content library and sends the requested content to the drone, which then forwards it to the requesting vehicle. The content request latency T2 obtained from the base station is:

[0063]

[0064] The total content request delay T of vehicle users is:

[0065] T=α·T1+β·T2

[0066] Among them, α represents the number of contents directly obtained from the drone, and β represents the number of contents obtained from the base station.

[0067] In some embodiments, a particle swarm algorithm is used to solve a pre-constructed target optimization problem, including:

[0068] (1) Initialize the cache placement scheme and calculate the content request latency for each scheme;

[0069] (2) updating the particle velocity and its corresponding cache placement scheme according to the individual extreme value of the particle and the group extreme value of the particle group;

[0070] (3) Iterate step (2) until a preset number of times is reached, and find the cache placement solution with the minimum latency as the solved cache placement solution.

[0071] In a second aspect, the present invention provides a drone-assisted vehicle-mounted network collaborative caching device, including a processor and a storage medium;

[0072] The storage medium is used to store instructions;

[0073] The processor is configured to operate according to the instructions to execute the method according to the first aspect.

[0074] In a third aspect, the present invention provides a device, comprising:

[0075] Memory;

[0076] processor;

[0077] as well as

[0078] Computer programs;

[0079] The computer program is stored in the memory and is configured to be executed by the processor to implement the method described in the first aspect above.

[0080] In a fourth aspect, the present invention provides a storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the method described in the first aspect.

[0081] Beneficial effects:

[0082] Compared with the prior art, the present invention has the following significant advantages:

[0083] 1. The present invention takes minimizing the content request delay as the optimization goal and uses the particle swarm algorithm to optimize the cache placement scheme to obtain the placement scheme with the minimum delay.

[0084] 2. The present invention provides a method for optimizing the hovering position of a drone, which can enable the drone to hover closer to active users, thereby reducing content request latency.

[0085] 3. The present invention provides a method for optimizing the trajectory and hovering time of a UAV, so that the UAV can provide better content services for ground vehicles according to the different activities of vehicle clusters, thereby improving the success rate of vehicle users' content requests. BRIEF DESCRIPTION OF THE DRAWINGS

[0086] Figure 1 A schematic diagram of a method flow according to an embodiment of the present invention;

[0087] Figure 2 is a schematic diagram of a system model according to an embodiment of the present invention;

[0088] Figure 3 This is a flight trajectory diagram of a drone in an embodiment of the present invention. DETAILED DESCRIPTION

[0089] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

[0090] In the description of the present invention, "several" means more than one, "many" means more than two, "greater than", "less than", "exceed", etc. are understood to exclude the number itself, and "above", "below", "within", etc. are understood to include the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0091] In the description of the present invention, the description with reference to the terms "one embodiment", "some embodiments", "illustrative embodiments", "examples", "specific examples", or "some examples" means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0092] Example 1

[0093] In a first aspect, this embodiment provides a drone-assisted vehicle network collaborative caching method, which is used in a vehicle self-organizing network and includes:

[0094] Step S1: Cluster the vehicles to obtain N C vehicle clusters;

[0095] In the urban road scene, N V There are N mobile vehicles, and the mobile vehicles are divided into N C vehicle clusters, the jth (j∈{1,2,3,...,n}) vehicle in the i-th vehicle cluster is denoted by V i,j ;

[0096] Step S2: For each vehicle cluster, the hovering position of the UAV in the vehicle cluster is optimized according to the vehicle positions in the cluster and the activity of the vehicle users;

[0097]

[0098] Among them, l i is the hovering position of the UAV in the i-th vehicle cluster, For vehicle V i,j activity, For vehicle V i,j The position of V i,j represents the jth vehicle in the i-th vehicle cluster, and n represents the total number of vehicles in the i-th vehicle cluster.

[0099] Step S3, optimizing the flight trajectory and hovering time of the UAV according to the activity of the vehicle cluster and the tolerance time of the vehicle users;

[0100] The activity of each vehicle cluster is w i , sort the activity of all vehicle clusters to get the activity of the vehicle cluster

[0101] The flight trajectory of the drone includes: based on the activity of the vehicle clusters, the drone first flies to the most active vehicle cluster, hovers at the hovering position of the vehicle cluster to provide content services to the vehicle, and then flies to the second most active vehicle cluster, and so on, until the tolerance time t of the vehicle user is exceeded;

[0102] The hovering time of the drone in each vehicle cluster is allocated according to the activity of the vehicle cluster to better serve the vehicle users. The hovering time of the drone in each vehicle cluster meets the following requirements:

[0103]

[0104]

[0105] in, are the hovering time of the UAV in each vehicle cluster according to the activity of the vehicle cluster; t UAV The total hovering time of the drone.

[0106] Step S4, optimizing the global content popularity according to the activity of the vehicle cluster and the popularity of the content within the cluster;

[0107]

[0108] Among them, P is the global content popularity, R k is the activity of the vehicle cluster, O k The content popularity of each cluster.

[0109] Step S5: Based on the optimized hovering positions of the drones in each vehicle cluster, the flight trajectories and hovering times of the drones, and the global content popularity, a particle swarm algorithm is used to solve the pre-constructed target optimization problem to obtain a cache placement solution;

[0110] Step S6: According to the obtained cache placement solution, the drone and the base station cache the vehicle request content.

[0111] In some embodiments, the system model diagram of the present application is as follows Figure 2 As shown, in the urban road scene, N V Mobile vehicles are divided into i(i∈{1,2,3,...,N C}) vehicle clusters, the jth (j∈{1,2,3,...,n}) vehicle in the i-th cluster is denoted by V i,j The drone is connected to the base station through a wireless link, and the base station is also connected to the remote content service provider through a backhaul link. The drone provides content services for each vehicle cluster by constantly moving. The cache capacity of the base station and the drone is S respectively. BS and S UAV, the mobile vehicle request content is q(q∈{1,2,3,...,Q}), and the drone position is [x UAV ,y UAV ,h].

[0112] When a mobile vehicle needs to obtain content, it first sends a content request to the drone. If the drone has the content cached, it will directly return the content to the requesting vehicle. Otherwise, the drone will forward the request to the base station. If the base station has the content cached, it will send it to the drone, and the drone will forward the content to the requesting vehicle. If neither the drone nor the base station has the content cached, the user needs to obtain the content from a remote content service provider.

[0113] The communication link between the drone and the ground can be modeled by a probabilistic loss model. According to the 3GPP specification, the path loss of the wireless backhaul link is determined by the line-of-sight (LoS) and non-line-of-sight (NLoS) link states. The path losses of LoS and NLoS are:

[0114] PL LoS =20log(4πfd / c)+η LoS

[0115] PL NLoS =20log(4πfd / c)+η NLoS

[0116] Where d is the distance between the drone and the requesting vehicle or base station, f is the carrier frequency, c is the speed of light, and η LoS and η NLoS are the attenuation factors of LoS and NLoS links respectively.

[0117] The probability of LoS and NLoS links depends on the density and height of buildings, the surrounding environment, and the elevation angle between the drone and the vehicle user or base station. Therefore, the probability of LoS and NLoS links is expressed as:

[0118]

[0119] P NLoS =1-P Los

[0120] Among them, θ (radians) is the elevation angle between the drone and the vehicle or base station, and X and Y are constants that depend on the environment. Therefore, the average path loss for communication between the drone and the vehicle or base station is:

[0121] PL=PL LoS ×P Los +PL NLoS ×P NLoS

[0122] From this, we can conclude that the transmission rate between the drone and the vehicle user is:

[0123]

[0124] in, is the bandwidth between the drone and the vehicle, is the received power of the vehicle, σ 2 is the noise power. The transmission rate between the drone and the base station is

[0125]

[0126] in, is the bandwidth between the base station and the drone, P UAV is the received power of the UAV.

[0127] After a mobile vehicle sends a content request, different content acquisition methods will produce different request delays. The total content request delay of the vehicle user is T. The present invention only considers the downlink transmission delay. Assuming that the request content is q and the content size is s q .

[0128] When the mobile vehicle obtains content directly from the drone, the content request latency is:

[0129]

[0130] When the content requested by the vehicle user is not cached in the drone, the drone needs to send a content request to the base station. The base station checks the local content library and sends the requested content to the drone, which then forwards it to the requesting vehicle. The content request latency at this time is:

[0131]

[0132] The total content request delay T of vehicle users is:

[0133] T=α·T1+β·T2

[0134] Among them, α represents the number of contents directly obtained from the drone, and β represents the number of contents obtained from the base station.

[0135] The request success rate is the ratio of the number of user content requests successfully responded by the drone to the total number of content requests within the vehicle user tolerance time, that is, the ratio of the number of cache hits of the drone and the base station to the total number of cached contents within the user tolerance time. It directly affects the QoE of the vehicle user. Define the cache placement matrix as A Q,2 , a q,: =1 means that the drone or base station has content q, a cached. q,:=0 means no cache. The same content cannot be cached repeatedly in the drone and the base station. The probability matrix of the requested content is P = [p1, p2, p3, ..., p Q ], the request success rate within the user tolerance time can be expressed as:

[0136] H=P×A Q,2

[0137] The request success rate can reflect the rationality of the drone flight trajectory and the drone hovering time. A better drone flight trajectory and a more reasonable drone hovering time allocation can make the vehicle user's request success rate higher.

[0138] Content request latency is one of the important indicators for evaluating cache strategies. It can directly reflect the quality of the cache placement solution. Considering the limited cache capacity, the target optimization problem is defined as:

[0139]

[0140] T=α·T1+β·T2

[0141]

[0142]

[0143] st:a:,:={0,1}

[0144]

[0145]

[0146]

[0147] x min ≤x UAV ≤x max

[0148] y min ≤y UAV ≤y max

[0149] h>0

[0150] Among them, A Q,2 represents the cache placement matrix, T is the total content request latency of vehicle users; α represents the number of contents directly obtained from the drone, β represents the number of contents obtained from the base station, T1 represents the content request latency directly obtained from the drone, T2 represents the content request latency obtained from the base station, and the requested content is q, where q∈{1,2,3,...,Q}, Q is the total number of contents, and the content size is s q , is the transmission rate between the UAV and the vehicle user, ζ BS,UAV is the transmission rate between the drone and the base station, st represents the constraint condition, a :,: ={0,1} means that the elements in the cache placement matrix can only be 0 or 1, a q,: =1 means that the drone or base station has content q, a cached. q,: = 0 means that there is no content q cached in the drone or base station. The cache capacity of the base station and the drone is S respectively. BS and S UAV , the drone position is [x UAV ,y UAV ,h],x UAV ,y UAV 、x min 、x max ,y min ,y max They represent the two-dimensional horizontal and vertical coordinates of the UAV and the corresponding minimum and maximum values, respectively, and h represents the height of the UAV;

[0151] Among them, the first two constraints indicate that the content cannot be cached repeatedly in the drone and base station, the third and fourth constraints indicate that the size of the content cached in the drone and base station cannot exceed their own cache capacity, and the last three constraints limit the location of the drone.

[0152] The hovering position of the drone in the vehicle cluster directly affects the total content request latency of the users in the entire cluster. In order to optimize the hovering position of the drone in each cluster, the present invention comprehensively considers the position of the vehicle and the activity of the vehicle users, with the goal of hovering the drone closer to the active users, thereby further reducing the content request latency and improving the request success rate. First, i points are randomly generated as the initial cluster centers, and the position of each cluster center is l i =[x i ,y i ], the vehicle’s position is Calculate the distance d from each vehicle to each cluster center:

[0153]

[0154] Each vehicle is assigned to the cluster that is closest to it. Then, the weighted geographic location of each vehicle is recalculated using the activity of the vehicle user as a weighting factor. Recompute the cluster center of each vehicle cluster by averaging the weighted geographic locations of the vehicles:

[0155]

[0156] in For vehicle V i,jThe activity of vehicle users follows the M-Zipf distribution. i,j is the αth active user in the cluster, then vehicle V i,j Activity level w α It can be expressed as:

[0157]

[0158] Among them, β is the platform factor of M-Zipf distribution, and δ is the skewness parameter of M-Zipf distribution. The new position l after optimization i =[x i ',y i '] is the two-dimensional coordinate of the drone.

[0159] The flight trajectory and hovering time of drones are important factors affecting the success rate of vehicle user requests. Traditional drones mostly fly along fixed trajectories to provide content services to ground vehicles, and hover for the same time in each area. This situation ignores the different activity levels of vehicle users in each area, resulting in a low success rate of user requests. Therefore, the present invention comprehensively considers the activity level of vehicle clusters and the tolerance time of vehicle users, and optimizes the flight trajectory and hovering time of drones to improve the success rate of user requests. The activity level of each vehicle cluster is w i (i∈{1,2,3,...,N C}), sorting the activity of vehicle clusters can get R k = {R1, R2, R3, ..., R NC}, the tolerance time of each request vehicle is t, and the flight speed of the drone is v UAV , the total hovering time of the drone is t UAV .

[0160] The drone first flies to the most active vehicle cluster, hovers at the hovering position of this vehicle cluster to provide content services to the vehicle, then flies to the second most active vehicle cluster, and so on, until the user tolerance time is exceeded. Since more vehicle users in more active vehicle clusters send content requests, the drone needs longer time to respond to the content requested by the mobile vehicle. If the hovering time of each cluster is evenly distributed, the user requests in the active vehicle cluster cannot be met. Therefore, the present invention allocates the hovering time of the drone in each vehicle cluster in a gradient according to the activity of the cluster to better serve the vehicle users. The hovering time of the drone in each vehicle cluster meets:

[0161]

[0162]

[0163] Assume there are four vehicle clusters and the flight trajectory of the drone is as follows Figure 3As shown in the figure, the activity of the clusters is ranked as w1>w4>w2>w3. According to the previous formula, the optimal hovering position of the drone in each cluster is calculated as l1=[x1,y1], l2=[x2,y2], l3=[x3,y3], l4=[x4,y4]. At this time, the flight trajectory of the drone is from cluster 1 to cluster 4, then to cluster 2, and finally to cluster 3. According to the activity of the cluster, the hovering time of the drone in each cluster is allocated as t1>t4>t2>t3, t1+t2+t3+t4=t UAV , at this time, the service time of the drone in vehicle cluster 1 is:

[0164]

[0165] The service time of the drone in vehicle cluster 4 is:

[0166]

[0167] When the user tolerates time When the drone reaches vehicle cluster 4, the user has stopped waiting for content response, so the service time is 0.

[0168] The service time of the drone in vehicle cluster 2 is:

[0169]

[0170] When the user tolerates time When the drone completes the content service for vehicle cluster 1 and vehicle cluster 4, it cannot continue to provide services for users of vehicle cluster 2.

[0171] The service time of the drone in vehicle cluster 3 is:

[0172]

[0173] When the user tolerates time , the UAV cannot reach vehicle cluster 3 within the user’s tolerance time, so the service time is 0.

[0174] It can be seen that the tolerance time t of vehicle users and the hovering time t of drones in each cluster i The flight trajectory of the drone directly affects the service time of the drone in each cluster. The longer the service time of the drone, the more time the user has to obtain the requested content, and the higher the success rate of the user's content request.

[0175] The content popularity directly calculated based on historical requests may be inaccurate. The present invention defines a global content popularity that comprehensively considers the activity of the vehicle cluster and the content popularity within the cluster. The content of interest to the more active vehicle cluster is more likely to be requested by most users. The content popularity within each vehicle cluster isq This is consistent with the Zipf distribution:

[0176]

[0177] Where s is the parameter of the Zipf distribution, e q It is the ranking of the number of requests for content q. The larger the parameter s is, the higher the probability of popular content appearing. The request frequency of the corresponding content will increase, and the probability of these contents being cached will also increase.

[0178] The activity ranking of vehicle clusters is From the above formula, we can get the content popularity of each cluster as O k ={o k,1 ,o k,2 ,o k,3 ,...,o k,Q}, so the global content popularity P = [P1, P2, P3, ..., P Q ] can be expressed as:

[0179]

[0180] Problem P1 is an NP-hard problem. The present invention uses a particle swarm algorithm to solve the optimal value of P1 to obtain the optimal cache placement strategy. Particle swarm algorithm is a search method based on group collaboration that simulates the foraging behavior of bird flocks. Its basic idea is to find the optimal solution through collaboration and information sharing between individuals in the group. In the process of searching for food, the flock of birds will share their position in the form of information with their companions around them to determine whether they are in the position closest to the food. By constantly comparing their own position with the position closest to the food in the group, all birds will gradually fly towards the food source. Eventually, the entire flock will gather near the food.

[0181] Assume that there are m = [1, 2, 3, ..., M] particles, and the position of each particle in the Gth iteration is The corresponding fitness value is The position of each particle represents a cache placement scheme, and the fitness value of the particle represents the latency of the vehicle user to obtain the requested content under the cache placement scheme. As of the Gth iteration, the optimal position of each particle is The corresponding optimal fitness value is The optimal position of the entire group is g best , the corresponding optimal fitness value of the group is F best , then the velocity and position of particle m in the G+1th iteration are:

[0182]

[0183]

[0184] Where w is the inertia weight of the speed. A larger w is conducive to global search, while a smaller w is conducive to local search. c1 is the individual learning factor. The larger c1 is, the more particles tend to go to the position with the highest individual fitness value. c2 is the group learning factor. The larger c2 is, the more particles tend to go to the optimal position of the group. a and b are random numbers in the range of [0,1].

[0185] When selecting the optimal cache placement plan, each particle first searches for the optimal solution in the search space separately, takes the maximum fitness value as the individual extreme value and records its corresponding cache placement plan, then compares it with other particles, finds the individual extreme value of the optimal particle as the group extreme value of this round of iteration and records the corresponding cache placement plan. In the next round of iteration, each particle updates its own cache placement plan based on the individual extreme value and the group extreme value and repeats the above search process until the cache placement plan with the smallest content request delay is found.

[0186] Finally, if Figure 1 As shown, the process of the invention method based on the above optimization problem is as follows:

[0187] (1) Initialize vehicle position;

[0188] (2) Clustering vehicles and optimizing the hovering position of drones based on vehicle locations and vehicle user activity;

[0189] (3) Optimize the flight trajectory and hovering time of the UAV according to the activity of the vehicle cluster and the tolerance time of the vehicle users;

[0190] (4) Obtaining global content popularity based on the activity of the vehicle cluster and the popularity of the content within the cluster;

[0191] (5) In the present invention, only drones and base stations can cache content. Different content acquisition methods will produce different delays. The content request delay is analyzed;

[0192] (6) Construct an optimization problem to minimize content request latency;

[0193] (7) The particle swarm algorithm is used to solve the problem. First, the cache placement scheme is initialized and the content request latency of each scheme is calculated;

[0194] (8) updating the particle velocity and its corresponding cache placement scheme according to the individual extreme value of the particle and the group extreme value of the particle group;

[0195] (9) When the maximum number of iterations is reached, the iteration is stopped, the cache placement solution with the minimum latency is found, and the content request latency at this time is recorded;

[0196] In summary, in order to make full use of the cache resources in the vehicle network and reduce the delay of vehicle request content, the present invention discloses a UAV-assisted vehicle network collaborative caching method. Specifically, the mobile vehicles are first clustered, and the cluster center closer to the active users is found according to the activity of the vehicle users, that is, the optimal two-dimensional position of the drone. Then, the flight trajectory and hovering time of the drone are optimized according to the activity of the vehicle cluster, so that the drone can better serve ground vehicle users. Finally, the global content popularity is obtained based on the cluster activity and the popularity of the content within the cluster. Since the cache capacity of the drone and the base station is limited, an optimization problem of delay minimization is constructed, and then the particle swarm algorithm is used to solve the problem to obtain the optimal cache placement solution.

[0197] Example 2

[0198] In a second aspect, based on Embodiment 1, this embodiment provides a UAV-assisted vehicle network cooperative caching device, including a processor and a storage medium;

[0199] The storage medium is used to store instructions;

[0200] The processor is configured to operate according to the instructions to execute the method according to embodiment 1.

[0201] Example 3

[0202] In a third aspect, based on embodiment 1, this embodiment provides a device, including:

[0203] Memory;

[0204] processor;

[0205] as well as

[0206] Computer programs;

[0207] The computer program is stored in the memory and is configured to be executed by the processor to implement the method described in Example 1.

[0208] Example 4

[0209] In a fourth aspect, based on Embodiment 1, this embodiment provides a storage medium on which a computer program is stored, and when the computer program is executed by a processor, the method described in Embodiment 1 is implemented.

[0210] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0211] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0212] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0213] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0214] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0215] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A UAV-assisted vehicle network collaborative caching method, characterized in that: include: Step S1: Cluster the vehicles to obtain N C vehicle clusters; Step S2: For each vehicle cluster, the hovering position of the UAV in the vehicle cluster is optimized according to the vehicle positions in the cluster and the activity of the vehicle users; Step S3, optimizing the flight trajectory and hovering time of the UAV according to the activity of the vehicle cluster and the tolerance time of the vehicle users; Step S4, optimizing the global content popularity according to the activity of the vehicle cluster and the popularity of the content within the cluster; Step S5: Based on the optimized hovering positions of the drones in each vehicle cluster, the flight trajectories and hovering times of the drones, and the global content popularity, a particle swarm algorithm is used to solve the pre-constructed target optimization problem to obtain a cache placement solution; The target optimization problem is represented as P1; T=α·T1+β·T2 s.t.:a :,: ={0,1} x min ≤x UAV ≤x max and min ≤y UAV ≤y max h＞0 Among them, A Q,2 represents the cache placement matrix, T is the total content request latency of vehicle users; α represents the number of contents directly obtained from the drone, β represents the number of contents obtained from the base station, T1 represents the content request latency directly obtained from the drone, T2 represents the content request latency obtained from the base station, and the requested content is q, where q∈{1,2,3,...,Q}, Q is the total number of contents, and the content size is s q , is the transmission rate between the UAV and the vehicle user, ζ BS,UAV is the transmission rate between the drone and the base station, st represents the constraint condition, a :,: ={0,1} means that the elements in the cache placement matrix can only be 0 or 1, a q,: =1 means that the drone or base station has content q, a cached. q,: = 0 means that there is no content q cached in the drone or base station. The cache capacity of the base station and the drone is S respectively. BS and S UAV , the drone position is [x UAV ,y UAV ,h],x UAV ,y UAV 、x min 、x max ,y min ,y max They represent the two-dimensional horizontal and vertical coordinates of the drone and the corresponding minimum and maximum values, and h represents the altitude of the drone. The first two constraints indicate that the content cannot be cached repeatedly in the drone and the base station. The third and fourth constraints indicate that the size of the cached content in the drone and the base station cannot exceed their own cache capacity. The last three constraints restrict the location of the drone. The method for constructing the target optimization problem includes: The communication link between the drone and the ground is modeled by a probabilistic loss model. The path loss of the wireless backhaul link is determined by the line-of-sight LoS and non-line-of-sight NLoS link states. The path losses of LoS and NLoS are PL and , respectively. LoS , PL NLoS : PL LoS =20log(4πfd / c)+η LoS PL NLoS =20log(4πfd / c)+η NLoS Where d is the distance between the drone and the requesting vehicle or base station, f is the carrier frequency, c is the speed of light, and η LoS and η NLoS are the attenuation factors of LoS and NLoS links respectively; The probability of LoS and NLoS links is denoted as P Los , P NLoS : P NLoS =1-P Los Where θ is the elevation angle between the drone and the vehicle or base station, X and Y are constants that depend on the environment; the average path loss PL between the drone and the vehicle or base station is: PL=PL LoS ×P Los +PL NLoS ×P NLoS Transmission rate between drone and vehicle users for: in, is the bandwidth between the drone and the vehicle, is the received power of the vehicle, σ 2 is the noise power; The transmission rate between the drone and the base station is BS,UAV for: in, is the bandwidth between the base station and the drone, P UAV is the received power of the UAV; After a mobile vehicle sends a content request, different content acquisition methods will produce different request delays. The total content request delay of the vehicle user is T, the requested content is q, where q∈{1,2,3,...,Q}, and the content size is s q ; When a mobile vehicle obtains content directly from a drone, the content request latency T1 obtained directly from the drone is: When the content requested by the vehicle user is not cached in the drone, the drone needs to send a content request to the base station. The base station checks the local content library and sends the requested content to the drone, which then forwards it to the requesting vehicle. The content request latency T2 obtained from the base station is: The total content request delay T of vehicle users is: T = α·T1+β·T2; Step S6: According to the obtained cache placement solution, the drone and the base station cache the vehicle request content.

2. The UAV-assisted vehicle network collaborative caching method according to claim 1, characterized in that: Step S1: Cluster the vehicles to obtain N C Vehicle clusters, including: In the urban road scene, N V There are N mobile vehicles, and the mobile vehicles are divided into N C vehicle clusters, the jth (j∈{1,2,3,...,n}) vehicle in the i-th vehicle cluster is denoted by V i,j .

3. The UAV-assisted vehicle network collaborative caching method according to claim 1, characterized in that: Step S2: For each vehicle cluster, the hovering position of the UAV in the vehicle cluster is optimized according to the vehicle positions in the cluster and the activity of the vehicle users, including: Among them, l i is the hovering position of the UAV in the i-th vehicle cluster, For vehicle V i,j activity, For vehicle V i,j The position of V i,j represents the jth vehicle in the i-th vehicle cluster, and n represents the total number of vehicles in the i-th vehicle cluster.

4. The UAV-assisted vehicle network collaborative caching method according to claim 1, characterized in that: Step S3, optimizing the flight trajectory and hovering time of the drone according to the activity of the vehicle cluster and the tolerance time of the vehicle users, including: Sort the activity of all vehicle clusters to get the activity of the vehicle cluster The flight trajectory of the drone includes: based on the activity of the vehicle clusters, the drone first flies to the most active vehicle cluster, hovers at the hovering position of the vehicle cluster to provide content services to the vehicle, and then flies to the second most active vehicle cluster, and so on, until the tolerance time t of the vehicle user is exceeded; The hovering time of the drone in each vehicle cluster is allocated according to the activity of the vehicle cluster to better serve the vehicle users. The hovering time of the drone in each vehicle cluster meets the following requirements: in, are the hovering time of the UAV in each vehicle cluster according to the activity of the vehicle cluster; t UAV The total hovering time of the drone.

5. The UAV-assisted vehicle network collaborative caching method according to claim 1, characterized in that: Step S4, optimizing the global content popularity according to the activity of the vehicle cluster and the popularity of the content in the cluster, including: Among them, P is the global content popularity, R k is the activity of the vehicle cluster, O k The content popularity of each cluster.

6. The UAV-assisted vehicle network collaborative caching method according to claim 1, characterized in that: The particle swarm algorithm is used to solve the pre-constructed target optimization problem, including: (1) Initialize the cache placement scheme and calculate the content request latency for each scheme; (2) updating the particle velocity and its corresponding cache placement scheme according to the individual extreme value of the particle and the group extreme value of the particle group; (3) Iterate step (2) until a preset number of times is reached, and find the cache placement solution with the minimum latency as the solved cache placement solution.

7. An electronic device, characterized in that: include: Memory; processor; as well as Computer programs; The computer program is stored in the memory and is configured to be executed by the processor to implement the method according to any one of claims 1 to 6.

8. A storage medium, characterized in that: A computer program is stored thereon, and when the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images