Track and resource allocation optimization method and device in unmanned aerial vehicle auxiliary cache system

Abstract

The invention discloses a trajectory and resource allocation optimization method and system in an unmanned aerial vehicle auxiliary cache system, and belongs to the technical field of wireless communication. The trajectory and resource allocation optimization method in the unmanned aerial vehicle auxiliary cache system comprises the steps of constructing the unmanned aerial vehicle auxiliary cache system, deriving a relationship between a user request time delay and an unmanned aerial vehicle hovering position, cache and trajectory based on unmanned aerial vehicle coverage radius and cache capacity constraints, establishing a user average request time delay minimization optimization problem, and optimizing the trajectory and resource allocation in the unmanned aerial vehicle auxiliary cache system. The optimization problem is decomposed into a user clustering sub-problem, a caching strategy sub-problem, a trajectory optimization sub-problem and a bandwidth power joint optimization sub-problem; and a neighbor clustering algorithm, a dynamic programming algorithm, an ant colony algorithm and a block coordinate descent algorithm are utilized to solve a user clustering sub-problem, a cache strategy sub-problem, a trajectory optimization sub-problem and a bandwidth power joint optimization sub-problem respectively, and finally an optimal user clustering, a cache placement strategy, an unmanned aerial vehicle flight trajectory and a bandwidth power distribution result are determined.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of wireless communication, and particularly relates to a trajectory and resource allocation optimization method and device in a UAV-assisted caching system. BACKGROUND

[0002] In the field of wireless communication, UAVs become an important part of the next generation network due to their flexible deployment, high mobility and ability to cover complex areas, providing more efficient content distribution services for mobile users. Mobile edge caching systems can effectively alleviate network congestion and reduce transmission delay by pre-storing content in edge nodes close to users. However, fixedly deployed edge servers have limited coverage and uneven resource utilization in application scenarios with widely distributed and dynamically changing user density. Therefore, researchers combine UAVs with mobile edge caching technology and conduct research on UAV-assisted caching systems.

[0003] However, current UAV-assisted caching systems rarely consider the problem of long average response delay of users. Specifically, for UAV-assisted communication, if there is a lack of effective caching strategies, trajectory planning and bandwidth power allocation mechanisms, ground mobile users may have a problem of excessively long response delay, resulting in reduced user experience quality and system performance.

[0004] To solve the above problems, the application provides a trajectory and resource allocation optimization method in a UAV-assisted caching system. SUMMARY

[0005] To solve the above problems, the application provides a trajectory and resource allocation optimization method in a UAV-assisted caching system.

[0006] The object of the application can be achieved by the following technical solutions:

[0007] The trajectory and resource allocation optimization method in the UAV-assisted caching system comprises the following steps:

[0008] A UAV-assisted caching system is constructed, and based on the constraints of UAV coverage radius and caching capacity, the relationship between user request delay and UAV hovering position, caching and trajectory is derived, an optimization problem of minimizing user average request delay is established, and the optimization problem is divided into user clustering sub-problems, caching strategy sub-problems, UAV trajectory optimization sub-problems and bandwidth power joint optimization sub-problems;

[0009] The near neighbor clustering algorithm, dynamic programming algorithm, ant colony algorithm and block coordinate descent algorithm are used to solve the user clustering subproblem, cache strategy subproblem, unmanned aerial vehicle trajectory optimization subproblem and bandwidth power joint optimization subproblem respectively, and finally the optimal user clustering, cache placement strategy, unmanned aerial vehicle flight trajectory and bandwidth power allocation result are determined.

[0010] Further, the unmanned aerial vehicle assisted cache system comprises one unmanned aerial vehicle, one ground base station and N users; the unmanned aerial vehicle provides assisted mobile edge cache service for ground users in the form of frequency division multiple access; the unmanned aerial vehicle takes off from the ground base station and provides service for N mobile users during flight; the unmanned aerial vehicle stores part of hot content through a cache strategy; when a user initiates a request, if the target content is cached on the unmanned aerial vehicle, the unmanned aerial vehicle directly transmits the target content to the user through the air-to-ground link; if the target content is not cached, the unmanned aerial vehicle relays the request to the ground base station to obtain the target content and then forwards the target content to the user;

[0011] The N users are divided into M user clusters according to spatial distribution, and the users in each cluster are located within the coverage range of a single hovering of the unmanned aerial vehicle.

[0012] Further, the user average request delay minimization optimization problem P1 is:

[0013] P1:

[0014] s.t.C1:

[0015] C2:

[0016] C3:

[0017] C4:

[0018] wherein the constraint C1 indicates that the overall size of the cache content cached by the unmanned aerial vehicle needs to satisfy the upper limit of the capacity; the constraint C2 indicates that the sum of the power and the sum of the bandwidth allocated to the users in the same cluster do not exceed the total power and the total bandwidth of the unmanned aerial vehicle; the constraint C3 indicates that the cache strategy variable c f is a binary variable; and the constraint C4 indicates the distance between the user n and the current position of the unmanned aerial vehicle, i.e. is not greater than the coverage radius L cov of the unmanned aerial vehicle; C represents a cache strategy set, P represents a power allocation set, B represents a bandwidth allocation set, and G represents an unmanned aerial vehicle trajectory; M represents the number of user clusters, N m represents the number of users in the user cluster m, C m represents a user set of the user cluster m, P n represents the transmission power of the unmanned aerial vehicle allocated to the user n, and B ndenotes the bandwidth allocated to user n by the UAV, F denotes the number of files, F denotes the set of files, r n,f denotes the case that user n requests file f, denotes the user transmission delay, denotes the user waiting delay, W f denotes the size of file f; D ave is the average request delay for users.

[0019] Further, the cache strategy sub-problem is:

[0020] P2:

[0021] s.t.C1:

[0022] C3:

[0023] The UAV trajectory optimization sub-problem is:

[0024] P3:

[0025] The bandwidth and power joint optimization sub-problem is:

[0026] P4:

[0027] s.t.C2:

[0028] Further, in solving the user clustering sub-problem by using the near neighbor clustering method, under the constraint of the maximum service radius of the UAV, the near neighbor clustering algorithm is used to divide the ground users into several spatially adjacent clusters, each cluster is sequentially served by a single UAV, and for each clustering cluster, the communication resource allocation of the UAV is optimized.

[0029] Further, the block coordinate descent method is used to solve the bandwidth and power joint optimization sub-problem, including the following steps:

[0030] Under the given UAV flight trajectory G, cache strategy c and bandwidth allocation result B, the Lagrange multiplier method is used to solve the power allocation sub-problem in each cluster;

[0031] Under the given UAV flight trajectory G, cache strategy c and power allocation result P, the convex optimization method is used to solve the bandwidth allocation sub-problem in each cluster;

[0032] The power allocation sub-problem and the bandwidth allocation sub-problem are alternately solved until convergence.

[0033] The trajectory and resource allocation optimization system in the UAV-aided cache system comprises:

[0034] An optimization problem construction decomposition module: construct a UAV-aided cache system, derive the relationship between user request delay and UAV hovering position, cache and trajectory based on UAV coverage radius and cache capacity constraints, establish a user average request delay minimization optimization problem, and decompose the optimization problem into a user clustering subproblem, a cache strategy subproblem, a UAV trajectory optimization subproblem and a bandwidth power joint optimization subproblem;

[0035] An optimization problem solving module: using a nearest neighbor clustering algorithm, a dynamic programming algorithm, an ant colony algorithm and a block coordinate descent algorithm, respectively solving the user clustering subproblem, the cache strategy subproblem, the UAV trajectory optimization subproblem and the bandwidth power joint optimization subproblem, and finally determining the optimal user clustering, cache placement strategy, UAV flight trajectory and bandwidth power allocation result.

[0036] A computer storage medium stores a readable program, when the program runs, the program can instruct a computing device to execute the trajectory and resource allocation optimization method in the UAV-aided cache system as described above.

[0037] An electronic device, comprising: a processor, a memory, a communication interface and a communication bus, the processor, the memory and the communication interface complete the communication among each other through the communication bus;

[0038] The memory is used to store at least one executable instruction, and the executable instruction makes the processor execute the operations corresponding to the trajectory and resource allocation optimization method in the UAV-aided cache system as described above.

[0039] A computer program product comprising computer instructions, the computer instructions instructing a computing device to execute the operations corresponding to the trajectory and resource allocation optimization method in the UAV-aided cache system as described above.

[0040] The beneficial effects of the present application are:

[0041] The trajectory and resource allocation optimization method in the UAV-aided cache system proposed by the present application has a significant reduction in user average request delay compared with existing methods, supports multi-user scenario expansion, has good adaptability to flight height, transmission power and other parameters, provides an effective solution for performance improvement of the UAV-aided cache system, and has flexibility and compatibility in engineering applications. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the drawings needed in the embodiment or prior art description will be briefly introduced as follows, and obviously, other drawings can also be obtained by those skilled in the art without creative labor.

[0043] Figure 1 This is a schematic diagram of the UAV-assisted caching system structure of the present invention;

[0044] Figure 2 This is a flowchart of the optimization method of the present invention;

[0045] Figure 3 This is a graph showing how the average user request latency changes with the number of users, corresponding to the optimization method of this invention.

[0046] Figure 4 This is a comparison chart of the average user request latency and the total bandwidth of the UAV corresponding to the optimization method of the present invention;

[0047] Figure 5 This is a comparison chart of the average user request latency and the total power of the drone corresponding to the optimization method of the present invention;

[0048] Figure 6 This is a comparison chart of the average user request latency and the drone flight altitude corresponding to the optimization method of the present invention;

[0049] Figure 7 This is a comparison chart of the average user request latency and the drone flight speed corresponding to the optimization method of this invention. Detailed Implementation

[0050] 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.

[0051] Example 1

[0052] like Figure 1 As shown, the UAV-assisted caching system includes: one UAV, one ground base station, and N users; the UAV acts as an airborne base station, providing auxiliary mobile edge caching services to ground users in the form of frequency division multiple access; the UAV takes off from the ground base station and provides services to the N mobile users on the ground during flight; the UAV stores some hot content through a caching strategy; when a user initiates a request, if the target content has been cached on the UAV, it is transmitted directly through the air-to-ground link; if it has not been cached, the UAV needs to relay the request to the ground base station to obtain the content before forwarding it to the user.

[0053] To optimize service efficiency, after acquiring user request distribution and implementing a caching strategy, the drone provides services to users at a constant speed v and flight altitude H. Assuming the angle of attack between the drone and the ground terminal is θ, its coverage radius is represented by L. cov=Htanθ. Based on this coverage characteristic, the system divides N users into M user clusters according to their spatial distribution, and the users in each cluster are all within the coverage area of ​​a single hovering of the UAV.

[0054] The set of hovering positions of a drone is represented as CH = {V1, V2, ..., V...} M}and Where V m This represents the hovering position of the drone in the m-th user cluster. Define the variable γ. nm This indicates whether user n belongs to user cluster m. Therefore, the set of users within the m-th user cluster is defined as C. m ={n|γ nm =1, n∈N}, 1≤m≤M, the number of nodes in the set is defined as N. m The drone's trajectory is defined as G. When the drone flies over a user cluster, it hovers and provides services to the users in that cluster. After completing the transmission in that cluster, it flies to the next user cluster and hovers over it to provide services to the users in the next cluster. This continues until the drone has provided services to all clusters and then returns to the ground base station.

[0055] The various model mechanisms involved in this embodiment are as follows:

[0056] 1) Communication model

[0057] Assuming that the channel between the UAV and the ground user is dominated by line-of-sight transmission, with weak influence from non-line-of-sight components and negligible small-scale fading, and assuming that the Doppler effect caused by the UAV's motion can be perfectly compensated for using Doppler effect compensation techniques, the air-to-ground channel quality is mainly affected by the communication distance between the UAV and the ground user.

[0058] Consider using a three-dimensional Cartesian coordinate system, where the position of user n is represented as... Where w n =[x n ,y n ] T Let n represent the horizontal coordinate of user n, and let the drone's position be represented as... Where q B =[x B ,y B ] T H corresponds to the drone's horizontal coordinate, and H corresponds to the drone's flight altitude. Then the distance d between the drone and user n is... U,n It can be represented as:

[0059]

[0060] Therefore, the air-to-ground channel gain h between the drone and user n U,n Then it is represented as:

[0061]

[0062] where η l denotes the power gain when the reference distance is 1m, and a l denotes the channel path loss exponent of the line-of-sight channel.

[0063] Assume that the total communication bandwidth of the UAV is B, and each user in the cluster establishes a communication connection with the UAV through frequency division multiple access, then the bandwidth of the user n in the cluster is B m , and thus we have B n . C m is the user set of the user cluster m. Assume that the total transmission power of the UAV is P, then the power of the user n in the cluster is P m , and thus we have P n . The transmission rate of the UAV for transmitting a file to the user n in the cluster is denoted as .

[0064]

[0065] where is the noise power spectral density.

[0066] The position of the ground base station is denoted as where z b = [x b , y b ] T denotes the horizontal coordinates of the base station, and the position of the UAV is denoted as where q B = [x B , y B ] T corresponds to the horizontal coordinates of the UAV, and H corresponds to the flight height of the UAV. Then the distance d b,U between the base station and the UAV can be denoted as:

[0067]

[0068] Therefore, the channel gain h b,U between the UAV and the base station is denoted as follows:

[0069]

[0070] where η b denotes the power gain when the reference distance is 1m, and a b denotes the channel path loss exponent.

[0071] While the drone hovers above the user cluster, assuming ideal line-of-sight transmission via its backhaul link to the ground base station, its backhaul link transmission rate is... Represented as:

[0072]

[0073] in, For noise power spectral density, B b For base station bandwidth, P b This refers to the base station's transmission power.

[0074] 2) Content Request and Caching Model

[0075] Let the content library be a finite set F containing F files, and let the size of file f be W. f According to the classic theoretical framework of the Zipf distribution, the popularity p of any content file f... f for

[0076]

[0077] in, is the Zipf distribution parameter, representing the degree of skewness in the Zipf distribution. The larger the parameter value, the more significant the concentration of popularity in the head content; j is the summation index variable.

[0078] Consider all users making content requests at the same time, and the requested content is known. This is done using the variable r. n,f To characterize user request behavior, if user n requests content f, then r n,f =1; if not requested, then r n,f =0. Assume each user requests only one file during this service period, and that multiple users are allowed to request the same content file.

[0079] Because drones have limited caching resources, they cannot store all content files; therefore, only a portion of the files can be selectively cached. A variable `c` is introduced. f Indicates cache status: when c f =1 indicates that the drone has cached the content file f requested by the user; otherwise, c f =0. To meet storage constraints, the total size of all cached files must not exceed the drone's cache capacity Q, i.e., the corresponding capacity constraint must be met. Therefore:

[0080]

[0081] 3) Delay Model

[0082] If the drone caches the content the user needs, the time t taken to transmit the content file f to the user n is... U,nis denoted as:

[0083]

[0084] If the UAV does not have the content cached, the UAV needs to request from the ground base station through the wireless backhaul link, and then the content file is transmitted by the UAV to the user. The time t b,U is:

[0085]

[0086] The transmission delay for the intra-cluster user n requesting content f is is denoted as:

[0087]

[0088] The intra-cluster user waiting delay for the mth cluster is It consists of two parts, one is the UAV hovering time, and the other is the UAV flight time. Assuming that there are |C m-1 | request users in cluster m-1, the hovering time of the UAV in this cluster is which can be denoted as Since the UAV flight speed is v, assuming that the distance from cluster m-1 to cluster m is L m-1,m wherein, and is the two-dimensional representation of the UAV hovering at the positions of clusters m-1 and m, so the flight time of the UAV from cluster m-1 to cluster m is is

[0089] Therefore, the cumulative hovering time of the first m-1 clusters is denoted as which is the sum of the time for the users in the first m-1 clusters to complete the transmission request. Accordingly, the cumulative flight time of the first m-1 clusters is denoted as which is the cumulative flight time of the UAV from the base station to the current cluster. Therefore, the user n waiting delay in cluster m is which can be denoted as:

[0090]

[0091] wherein, is the hovering delay of the UAV at the user cluster i, is the flight delay of the UAV to the user cluster i.

[0092] Embodiment 2

[0093] As Figure 2The unmanned aerial vehicle-assisted cache system trajectory and resource allocation optimization method shown includes the following steps:

[0094] S1, constructing an unmanned aerial vehicle-assisted cache system, based on the unmanned aerial vehicle coverage radius and cache capacity constraints, deriving the relationship between user request delay and unmanned aerial vehicle hovering position, cache and trajectory, establishing a user average request delay minimization optimization problem, and decomposing the optimization problem into a user clustering subproblem, a cache strategy subproblem, a trajectory optimization subproblem and a bandwidth power joint optimization subproblem;

[0095] The user average request delay minimization optimization problem P1 is:

[0096] P1:

[0097] s.t.C1:

[0098] C2:

[0099] C3:

[0100] C4:

[0101] Constraint C1 indicates that the overall size of the cache content cached by the unmanned aerial vehicle needs to meet the upper limit of the capacity; constraint C2 indicates that the sum of the power allocated to the users in the same cluster and the sum of the bandwidth are not more than the total power and total bandwidth of the unmanned aerial vehicle; constraint C3 indicates that the cache strategy variable c f is a binary variable; constraint C4 indicates the distance between user n and the current position of the unmanned aerial vehicle, i.e. is not greater than the coverage radius L cov of the unmanned aerial vehicle; C represents the cache strategy set, P represents the power allocation set, B represents the bandwidth allocation set, and G represents the unmanned aerial vehicle trajectory; M represents the number of user clusters, N m represents the number of users in user cluster m, C m represents the user set of user cluster m, P n represents the transmission power of the unmanned aerial vehicle allocated to user n, B n represents the bandwidth of the unmanned aerial vehicle allocated to user n, F represents the number of files, F represents the file set, r n,f represents the case where user n requests file f, represents the user transmission delay, represents the user waiting delay, W f represents the size of file f; D ave is the user average request delay.

[0102] The user clustering subproblem is represented as:

[0103] Under the constraint of the maximum service radius of the UAV, the ground users are divided into several spatially adjacent clusters by using a near neighbor clustering method. Each cluster is sequentially served by a single UAV. For each clustering cluster, the communication resource allocation of the UAV is optimized, so as to minimize the number of user clusters, reduce the average request transmission delay of the ground users, and reduce the delay overhead caused by the frequent switching of service objects by the UAV.

[0104] The cache strategy sub-problem is:

[0105] P2:

[0106] s.t.C1:

[0107] C3:

[0108] The UAV trajectory optimization sub-problem is:

[0109] P3:

[0110] The bandwidth and power joint optimization sub-problem is:

[0111] P4:

[0112] s.t.C2:

[0113] S2, by using a near neighbor clustering algorithm, a dynamic programming algorithm, an ant colony algorithm and a block coordinate descent algorithm, respectively solves the user clustering sub-problem, the cache strategy sub-problem, the trajectory optimization sub-problem and the bandwidth and power joint optimization sub-problem, and finally determines the optimal user clustering, cache placement strategy, UAV flight trajectory and bandwidth and power allocation result.

[0114] 1) The embodiment adopts a near neighbor clustering method to solve the user clustering sub-problem;

[0115] Under the constraint of the maximum service radius of the UAV, the ground users are divided into several spatially adjacent clusters by using a near neighbor clustering method. Each cluster is sequentially served by a single UAV. For each clustering cluster, the communication resource allocation of the UAV is optimized, so as to minimize the number of user clusters, reduce the average request transmission delay of the ground users, and reduce the delay overhead caused by the frequent switching of service objects by the UAV. The specific steps include:

[0116] 1.1) Construct a user similarity matrix s(n, k):

[0117]

[0118] Wherein, L n,k is the distance between users n and k, Lcov = H tan Θ represents the coverage range of the UAV, s cov (n, k) represents the users in the similarity matrix that satisfy the coverage radius constraint, and the specific formula is:

[0119]

[0120] 1.2) Construct the attraction matrix and the belonging matrix: the element r(n, k) in the attraction matrix R is used to measure the degree to which user k is more suitable as the cluster center of user n relative to other candidate points; the element a(n, k) in the belonging matrix A is used to measure the appropriate degree of user n selecting user k as the cluster center.

[0121] 1.3) Construct the update equation of the attraction matrix R and the belonging matrix A:

[0122]

[0123] 1.4) Determine the cluster center according to the update equation: if node k, a(k, k) + r(k, k) > 0, the node is selected as the cluster center. For a non-cluster center node n, select the node k with the maximum value of a(n, k) + r(n, k) as the belonging cluster center. Therefore, for a certain node n, its final belonging cluster center k * is determined by the following rules:

[0124]

[0125] 1.5) According to the more after several iterations, the cluster center is unchanged or the number of iterations exceeds the set number of times, the iteration is terminated.

[0126] 2) The cache strategy sub-problem P2 is a 0-1 integer programming problem, and the dynamic programming method is used to solve it in this embodiment.

[0127] Due to the limited cache capacity, it is necessary to make reasonable decisions on cache objects under limited resources to minimize user request delay. This problem can be modeled as a classic 0-1 knapsack problem, where each cache file corresponds to an item in the knapsack problem, and the cache space it occupies and the user delay it can save correspond to the weight and value of the item, respectively. The specific steps include:

[0128] 2.1) Define and initialize the capacity W required by the cache file f f and the benefits Val f ;

[0129] 2.2) Define the time delay reduced when the file f is cached as , that is, the cache benefit; where C k is the user cluster k, and Vk r is the hovering position of the UAV above the user cluster k n,f for indicating whether the user n has requested the file f, for indicating the UAV serving the cluster C k the latency consumed by the file f transmitted from the base station to the UAV when the UAV is serving.

[0130] 2.3) define the state dp[i][j] representing the maximum profit that can be achieved only considering the first i files and the cache capacity does not exceed j;

[0131] 2.4) build the state transition equation as

[0132]

[0133] where W i is the size of the file i;

[0134] 2.5) build the state transition table according to the state transition equation, and traverse all the files in reverse order from the final state by backtracking algorithm. If the current state dp[i][j]≠dp[i-1][j], it means that the file i is cached, record and update the remaining capacity to obtain the final cache strategy.

[0135] 3) solve the UAV trajectory optimization sub-problem P3 using ant colony algorithm;

[0136] Under the constraint that the UAV maintains a constant speed, solve a flight trajectory starting from the ground base station and sequentially traversing all user clusters to minimize the flight time of the system to complete all user services. That is, the ant starts from the starting point, gradually selects the next node according to the probability, forms a path traversing all nodes, and obtains the optimal trajectory of the UAV; the specific steps include:

[0137] 3.1) initialize the parameters of the ant colony algorithm: for the edge (i,j) connecting the hovering positions V i and V j of the UAV in each cluster, define τ ij (t) as the pheromone concentration on the edge (i,j) in the tth round of circulation, and define η ij as the heuristic information on the edge, which is represented as the reciprocal of the path length between the hovering positions of the UAV.

[0138] 3.2) define the probability of the ant k selecting to move from node i to node j as :

[0139]

[0140] where a is pheromone factor, indicating the degree of influence of pheromone on the path when selecting the next node. β is heuristic factor, indicating the degree of influence of heuristic information. N i k Nk(t) represents the set of neighbor nodes currently available to ant k.

[0141] 3.3) Define the pheromone update formula as:

[0142] τ ij (t+1) = (1- p) τ ij (t) + Aτ ij (t)

[0143] where τ ij (t+1) is the pheromone concentration on the path from node i to node j in the t+1th round of circulation, p is the pheromone evaporation coefficient, Aτ ij (t) is the total amount of pheromone deposited by all ants on the path from node i to node j at time t, which is specifically represented as:

[0144]

[0145] where K a represents the number of ants, represents the pheromone increment of ant k, which is if ant k passes through edge (i, j), otherwise it is 0, where A Q is a preset constant, and L k is the total length of the path of ant k this time.

[0146] 3.4) Through multiple iterations, pheromone continuously accumulates on excellent paths, and ants tend to choose higher quality paths, gradually approaching the global optimal solution.

[0147] 4) Use the block coordinate descent method to solve the UAV bandwidth and power joint optimization subproblem P4 to obtain the optimal UAV bandwidth and power allocation, including the following steps:

[0148] 4.1) Given the UAV flight trajectory G, cache strategy c and bandwidth allocation result B, solve the power allocation subproblem within each cluster by Lagrange multiplier method;

[0149] 4.2) Given the UAV flight trajectory G, cache strategy c and power allocation result P, solve the bandwidth allocation subproblem within each cluster by convex optimization method;

[0150] 4.3) Alternately solve the two subproblems until convergence.

[0151] Specifically:

[0152] For solving the power allocation sub-problem, the optimization objective of the power allocation sub-problem in each cluster is expressed as follows:

[0153] P4.1:

[0154]

[0155] When the UAV flight trajectory, cache strategy and bandwidth allocation result are known, the request delay is only related to the change of transmission delay. Therefore, P4.1 can be further simplified to solve the minimum total transmission delay of the users in a single cluster, and the optimization problem is constructed as follows:

[0156] P4.2

[0157]

[0158] C2:

[0159] Since the above problem is a non-convex optimization problem, it cannot be directly solved by using a convex optimization method. In order to solve the above constrained optimization problem, the Lagrange multiplier is introduced, and the Lagrange function is constructed as follows:

[0160]

[0161] wherein, represents the file size requested by the user n, and λ is the Lagrange multiplier.

[0162] For each p n , the partial derivative is obtained:

[0163]

[0164] Let be equal to 0, and the following equation is obtained:

[0165]

[0166] Since the explicit expression of p n needs to be solved, but the closed-form solution of the equation is difficult to directly obtain, therefore, a numerical method is usually used to solve it. The bisection search method is used to solve the Lagrange multiplier λ, and the optimal value of p n is obtained by solving the nonlinear equation set, that is, the optimal power allocation, and the total power constraint is satisfied.

[0167] For solving the bandwidth allocation sub-problem, the optimization objective of the bandwidth allocation sub-problem in each cluster is expressed as follows:

[0168] P4.3:

[0169]

[0170] When the UAV flight trajectory, caching strategy and transmit power are known, the request latency is only related to the variation of transmission latency. Therefore, P4.3 can be further simplified to solve the total transmission latency minimization of users in a single cluster, and the optimization problem is constructed as follows:

[0171] P4.4

[0172]

[0173] where B n represents the bandwidth allocated to the request user n in the cluster.

[0174] According to the analysis of the problem, the sub-problem is a convex optimization problem, which can be directly solved by using the convex optimization tool CVX. The specific proof is as follows: for the user n in the cluster m, assuming that the transmission power allocation is known, the optimization problem P4.4 can be expressed as a function of B n . Let B n =x, Since the transmission rate of the UAV is , then Define The objective function can be expressed as solving the sum of multiple g(x) and satisfying the constraint that the sum of x is less than the total bandwidth.

[0175] First, analyze the convexity of the function f(x). Since a n is positive, the domain is x>0. The first derivative of f(x) is The second derivative of f'(x) is Since the denominator x(x+a n ) 2 is positive in the domain, and the numerator -a n 2 is negative, f”(x) <0. In the domain, the second derivative of the function is negative, so the function f(x) is a concave function.

[0176] Next, analyze the convexity of the function g(x). It is known that is a concave function, which holds in the domain x>0. The first derivative of g(x) is The second derivative of g'(x) is For the first term -f”(x)f(x) in the numerator, since f”(x) <0 and f(x)>0, the term is positive. For the second term 2[f'(x)] 2 , since the term is a square term, the term is always non-negative. For the denominator [f(x)]3 Since f(x) > 0, the third power of a positive number is positive, so this term is positive. Therefore, in the domain x > 0, the second derivative of the function is always positive, so the function g(x) is a convex function.

[0177] According to the property of the convex function, the non-negative weighted sum of the convex function is still a convex function, so the objective function is a convex function, and the optimization problem P4.4 can be directly solved by using the convex optimization method, and the CVX solver of MATLAB is used for solving.

[0178] By sequentially solving the user clustering sub-problem, the problems P2, P3 and P4, the optimal user clustering result, the cache strategy, the unmanned aerial vehicle flight trajectory and the unmanned aerial vehicle bandwidth power allocation result are obtained.

[0179] Based on the similar inventive concept, the embodiment of the application also provides a computer storage medium, which stores a readable program, when the program is run by a processor, the program can execute the trajectory and resource allocation optimization method in the unmanned aerial vehicle assisted cache system.

[0180] Based on the similar inventive concept, the embodiment of the application provides an electronic device, which comprises a processor, a memory, a communication interface and a communication bus, the processor, the memory and the communication interface complete the communication among each other through the communication bus;

[0181] The memory is used to store at least one executable instruction, and the executable instruction makes the processor execute the operation corresponding to the trajectory and resource allocation optimization method in the unmanned aerial vehicle assisted cache system.

[0182] Based on the similar inventive concept, the embodiment of the application also provides a computer program product, which comprises computer instructions, and the computer instructions instruct a computing device to execute the operation corresponding to the trajectory and resource allocation optimization method in the unmanned aerial vehicle assisted cache system.

[0183] Embodiment 3

[0184] In order to verify the method of the application (embodiment 2), this embodiment considers comparing with the following four methods:

[0185] (1) Comparative method 1: the unmanned aerial vehicle transmission power is not optimized, the unmanned aerial vehicle power allocation adopts average allocation, and other parameters are optimized according to the method of the application;

[0186] (2) Comparative method 2: the unmanned aerial vehicle bandwidth is not optimized, the unmanned aerial vehicle bandwidth allocation adopts average allocation, and other parameters are optimized according to the method of the application;

[0187] (3) Comparative method 3: for trajectory optimization, the unmanned aerial vehicle adopts the nearest neighbor transfer strategy, and other parameters are optimized according to the method of the application;

[0188] (4) Comparison method 4: For the cache strategy, the files are pre-cached according to the popularity of the content, and other parameters are optimized according to the method of the application.

[0189] Figure 3 The trend of the average request delay of users with the number of users is given in the figure. As can be seen from the figure, with the increase of the number of users, the average request delay of users of the three methods shows an upward trend. This is mainly because the increase in the number of users leads to an increase in the number of users in each cluster, so that the bandwidth and power allocated to each user by the unmanned aerial vehicle are correspondingly reduced, and the time required for transmission of the same file is prolonged. Limited by the cache capacity of the unmanned aerial vehicle, with the increase of the number of users, the number of user requests that miss the cache also increases, thereby increasing the backhaul delay. As can be seen from the figure, the performance of the method of the application is better than that of the other four comparison algorithms, and the performance of the average power allocation is the worst, that is, among the four optimization variables, the optimization of power can obtain the maximum gain. This is because when the power is averagely allocated, the difference in channel state of the users in the cluster is ignored, which leads to longer transmission time for users with poor channel state, thereby prolonging the hovering time of the unmanned aerial vehicle above the cluster, thereby affecting the average request delay.

[0190] Figure 4 The influence of the total bandwidth of the unmanned aerial vehicle channel on the average request delay of users is given. As can be seen from the figure, with the increase of the total bandwidth of the unmanned aerial vehicle channel, the average request delay will decrease. This is because, when the number of users is fixed, increasing the total bandwidth of the unmanned aerial vehicle channel, the user gets a higher transmission rate, thereby reducing the user delay, reducing the communication transmission time, and reducing the average request delay of users.

[0191] Figure 5 The influence of the maximum transmission power of the unmanned aerial vehicle on the average request delay of users is given. As can be seen from the figure, with the increase of the maximum transmission power of the unmanned aerial vehicle, the average request delay will decrease. This is because, when the number of users is fixed, increasing the total transmission power of the unmanned aerial vehicle, the user receives more power, the user transmission delay decreases, and the average request delay of users decreases. With the increase of the number of users, the average request delay of users also increases accordingly. This is because, with the increase of the number of users, each user gets less power, so the transmission rate also decreases accordingly, thereby causing the increase of the user transmission delay.

[0192] Figure 6The influence of the flight height of the UAV on the average request delay of the users is given. As can be seen from the figure, the average request delay of the users presents an upward trend as the flight height increases. This is because, with the number of users remaining unchanged, the distance between the UAV and the users increases with the increase of the height, resulting in the corresponding increase of the transmission delay. On the other hand, the increase of the flight height expands the coverage range of the UAV, reducing the number of user clusters. While reducing the flight delay of the users in transferring between clusters, it also increases the number of users in the cluster, resulting in the decrease of the transmission rate of the users in the cluster, thereby causing the increase of the average request delay of the users.

[0193] Figure 7 The influence of the flight speed of the UAV on the average request delay of the users is given. As can be seen from the figure, the average request delay decreases accordingly as the flight speed increases. This is because, with the number of users remaining unchanged, the UAV moves between the hovering points at a higher speed, reducing the time required for transfer, thereby reducing the average request delay of the users.

[0194] Embodiment 4

[0195] Based on the trajectory and resource allocation optimization method of the UAV-aided caching system proposed in Embodiment 2, in this embodiment, the trajectory and resource allocation optimization device in the UAV-aided caching system specifically comprises:

[0196] An optimization problem construction and decomposition module: construct a UAV-aided caching system, based on the constraints of the coverage radius of the UAV and the caching capacity, deduce the relationship between the user request delay and the hovering position of the UAV, caching and trajectory, establish a user average request delay minimization optimization problem, and decompose the optimization problem into a user clustering sub-problem, a caching strategy sub-problem, a trajectory optimization sub-problem and a bandwidth power joint optimization sub-problem;

[0197] An optimization problem solving module: use the nearest neighbor clustering algorithm, the dynamic programming algorithm, the ant colony algorithm and the block coordinate descent algorithm to solve the user clustering sub-problem, the caching strategy sub-problem, the trajectory optimization sub-problem and the bandwidth power joint optimization sub-problem respectively, and finally determine the optimal user clustering, caching placement strategy, UAV flight trajectory and bandwidth power allocation result.

[0198] 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.

[0199] 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 method for optimizing trajectory and resource allocation in an unmanned aerial vehicle (UAV) assisted caching system, characterized in that, Includes the following steps: A drone-assisted caching system is constructed. Based on the constraints of drone coverage radius and cache capacity, the relationship between user request latency and drone hovering position, cache and trajectory is derived. An optimization problem of minimizing average user request latency is established, and the optimization problem is decomposed into user clustering subproblem, caching strategy subproblem, drone trajectory optimization subproblem and bandwidth power joint optimization subproblem. Using nearest neighbor clustering, dynamic programming, ant colony optimization, and block coordinate descent algorithms, we solve the user clustering subproblem, caching strategy subproblem, UAV trajectory optimization subproblem, and bandwidth-power joint optimization subproblem, respectively, and finally determine the optimal user clustering, cache placement strategy, UAV flight trajectory, and bandwidth-power allocation results.

2. The trajectory and resource allocation optimization method in the UAV-assisted caching system according to claim 1, characterized in that, The UAV-assisted caching system includes: one UAV, one ground base station, and N users; the UAV provides assisted mobile edge caching services to ground users in the form of frequency division multiple access; the UAV takes off from the ground base station and provides services to N mobile users during flight; the UAV stores some hot content through a caching strategy; when a user initiates a request, if the target content is already cached on the UAV, it is directly transmitted via the air-to-ground link; if it is not cached, the UAV needs to relay the request to the ground base station to obtain the content before forwarding it to the user; N users are spatially distributed into M user clusters, and the users in each cluster are within the coverage area of ​​a single hovering of the drone.

3. The trajectory and resource allocation optimization method in the UAV-assisted caching system according to claim 2, characterized in that, The optimization problem P1 that minimizes the average user request latency is: P1: s.t.C1: C2: C3: C4: Among them, constraint C1 indicates that the overall size of the drone's cached content must meet the capacity limit; constraint C2 indicates that the sum of power and bandwidth allocated to users within the same cluster cannot exceed the total power and total bandwidth of the drone; constraint C3 indicates the caching strategy variable c f The variables are binary; constraint C4 represents the distance between user n and the current position of the drone, i.e. No greater than the coverage radius L of the drone cov C represents the caching strategy set, P represents the power allocation set, B represents the bandwidth allocation set, G represents the drone trajectory; M represents the number of user clusters, N m C represents the number of users within user cluster m. m Let P represent the set of users in user cluster m. n B represents the transmit power allocated to user n by the drone. n Let F represent the bandwidth allocated to user n by the drone, F represent the number of files, F represent the file set, and r represent the number of files. n,f This indicates the case where user n requests file f. Indicates user transmission latency. W represents the user's waiting time. f Indicates the size of file f; D ave This represents the average latency for user requests.

4. The trajectory and resource allocation optimization method in the UAV-assisted caching system according to claim 3, characterized in that, The sub-problem of the caching strategy is: P2: s.t.C1: C3: The sub-problem of optimizing the drone trajectory is: P3: The bandwidth-power joint optimization subproblem is: P4: s.t.C2:

5. The trajectory and resource allocation optimization method in the UAV-assisted caching system according to claim 2, characterized in that, When using the nearest neighbor clustering method to solve the user clustering subproblem, considering the constraint of the maximum service radius of the UAV, the nearest neighbor clustering algorithm is used to divide the ground users into several spatially adjacent clusters. Each cluster is served sequentially by a single UAV. For each cluster, the communication resource allocation of the UAV is optimized.

6. The trajectory and resource allocation optimization method in the UAV-assisted caching system according to claim 4, characterized in that, Solving the bandwidth-power joint optimization subproblem using the block coordinate descent method includes the following steps: Given the UAV flight trajectory G, caching strategy c, and bandwidth allocation result B, the power allocation subproblem within each cluster is solved using the Lagrange multiplier method. Given the UAV flight trajectory G, caching strategy c, and power allocation result P, the bandwidth allocation subproblem within each cluster is solved using a convex optimization method. Solve the power allocator problem and the bandwidth allocator problem alternately until convergence.

7. A trajectory and resource allocation optimization system in an unmanned aerial vehicle (UAV) assisted caching system, characterized in that, include: The optimization problem decomposition module constructs a drone-assisted caching system. Based on the drone coverage radius and cache capacity constraints, it derives the relationship between user request latency and drone hovering position, cache and trajectory, establishes an optimization problem of minimizing average user request latency, and decomposes the optimization problem into user clustering sub-problems, caching strategy sub-problems, drone trajectory optimization sub-problems and bandwidth-power joint optimization sub-problems. The optimization problem-solving module utilizes nearest neighbor clustering, dynamic programming, ant colony optimization, and block coordinate descent algorithms to solve the user clustering subproblem, caching strategy subproblem, trajectory optimization subproblem, and bandwidth-power joint optimization subproblem, respectively, ultimately determining the optimal user clustering, cache placement strategy, UAV flight trajectory, and bandwidth-power allocation results.

8. A computer storage medium storing a readable program, characterized in that, When the program runs, it can instruct the computing device to execute the trajectory and resource allocation optimization method in the UAV-assisted caching system as described in any one of claims 1-6.

9. 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 trajectory and resource allocation optimization method in the UAV-assisted caching system as described in any one of claims 1-6.

10. A computer program product comprising computer instructions, characterized in that, The computer instructions instruct the computing device to perform the operations corresponding to the trajectory and resource allocation optimization method in the UAV-assisted caching system as described in any one of claims 1-6.

Images