A low-power consumption security cache method for internet of vehicles and medium

Abstract

The application discloses a low-power consumption security cache method and medium for Internet of Vehicles, and the method comprises the following steps: 1) building an edge cache system; 2) acquiring vehicle request data and determining vehicle request content preference; the request content preference comprises cache content and cache priority; 3) determining a content placement scheme of the edge cache system according to the vehicle request content preference; 4) acquiring content data from a content server by a macro base station and a roadside unit according to the content placement scheme; and 5) determining a content distribution scheme based on resource allocation according to the vehicle request data. The medium stores a computer program. The application significantly improves the power consumption of the system by jointly considering the cache scheme of vehicle preference, vehicle activity, content size and power and bandwidth allocation.

Description

TECHNICAL FIELD

[0001] The application relates to edge cache optimization, resource allocation and data security in the Internet of Vehicles, and in particular to a low-power safe cache method for the Internet of Vehicles and a medium. BACKGROUND

[0002] With the rapid development of the Internet of Vehicles, various vehicle-mounted applications based on the Internet of Vehicles are emerging, which greatly improves the driving experience of drivers, but this directly leads to an explosive growth of data traffic in the Internet of Vehicles, which greatly affects the latency and stability of the Internet of Vehicles system, so that the vehicle cannot obtain a large amount of data within an effective time, and at the same time, the safety and comfort of the driver are also threatened.

[0003] At present, edge cache technology can solve system cost and backhaul pressure, reduce content access latency and improve network capacity. Edge cache usually includes two stages: content placement and content distribution. Content placement usually places user-interested content into edge network devices in advance through a formulated cache strategy during off-peak hours, and content distribution distributes the pre-cached content to the corresponding user during the user request peak. In the Internet of Vehicles, the user is generally a vehicle, and the edge cache node that caches content is usually a macro base station (MBS) and a roadside unit (RSU) close to the vehicle. Considering some factors on cost, the cache resources of MBS and RSU are usually limited. Therefore, it is a great challenge to design a suitable cache strategy for the Internet of Vehicles scenario and solve how to reasonably cache content under limited cache capability. In addition, in the content placement stage, the content needs to be transmitted according to the cache strategy, and some private data may be included in the transmitted content, so how to effectively protect these contents through data encryption method is also a research difficulty.

[0004] Based on the above problems, relevant research institutions have proposed a collaborative caching scheme based on mobile prediction for the Internet of Vehicles to reduce the access delay of content; some research institutions have proposed a hierarchical end-edge framework that deeply combines data communication, computing offloading and content caching to minimize network overhead; and some research institutions have considered the optimal cache and computing allocation problem by combining the limited storage resources and computing resources of RSU as a constraint, while considering the mobility of vehicles to minimize the system overhead. However, these researches do not consider the influence of user's differentiated request, limited vehicle-end cache capacity, wireless resource (such as spectrum, power) limitation and other factors on edge computing, which will most likely lead to unreasonable content placement and content access problems, thereby reducing the quality of user experience. SUMMARY

[0005] The application aims to provide a low-power consumption security cache method for vehicle networking, comprising the following steps:

[0006] 1) Building the edge cache system.

[0007] The edge cache system comprises a system network model, an edge cache model and a power consumption model of the edge cache system;

[0008] The system network model comprises a content server, a macro base station, a roadside unit and a vehicle;

[0009] When the roadside unit detects a vehicle requesting communication in the communication coverage range, the channel state information of the vehicle is obtained through the network and transmitted to the macro base station;

[0010] The macro base station transmits the channel state information to the content server;

[0011] The content server is used to determine the content size requested by the vehicle;

[0012] The edge cache model is used to determine the vehicle activity, the vehicle preference for content and the probability density function of content size;

[0013] The power consumption model of the edge cache system is used to calculate the transmission power consumption and backhaul power consumption of the edge cache system;

[0014] 2) Obtain the vehicle request data and determine the vehicle request content preference; the request content preference comprises cached content and cache priority;

[0015] 3) According to the vehicle request content preference, determine the content placement scheme of the edge cache system;

[0016] 4) According to the content placement scheme, the macro base station and the roadside unit obtain content data from the content server;

[0017] 5) According to the vehicle request data, determine the content distribution scheme based on resource allocation.

[0018] Further, the roadside unit is installed on the roadside and is a logical service unit communicating with the vehicle-mounted unit and the roadside traffic control equipment.

[0019] Further, in the system network model, the macro base station and the roadside unit are both equipped with edge servers with cache capability;

[0020] The macro base station has a transmission power p0 and a bandwidth b0 for downlink content distribution,

[0021] The set of roadside units is denoted as

[0022] The available transmission power of the mth road side unit is p m , and the bandwidth is b m ;

[0023] The set of vehicles is denoted as

[0024] The communication coverage radius of the road side unit and the macro base station is r1 and r2, and N vehicles and M road side units are randomly distributed in the area served by the macro base station; in the communication coverage range of the road side unit, each vehicle can only be associated with one road side unit, and one road side unit is associated with vehicles;

[0025] The set of contents available in the content server is The size of the content f is denoted as s f ; F is the number of contents.

[0026] Further, the steps of building the edge cache model include:

[0027] a1) defining the preference vector θ n of the vehicle n for the content, θ n = {θ n，1 ,..., θ n，f ..., θ n，F}, wherein θ n，f represents the request probability of the vehicle n for the content f;

[0028] a2) defining the cache vector of the road side unit and the macro base station; when the content f is cached in the mth road side unit, the cache vector y m，f of the road side unit is 1, otherwise, y m，f = 0;

[0029] When the content f is cached in the macro base station, the cache vector y 0，f of the macro base station is 1, otherwise, y 0，f = 0,

[0030] a3) representing the preference of the vehicle n for the content f as ε n represents the preference distribution parameter of the vehicle n;

[0031] a4) using the Pareto distribution to represent the content size s f , and the probability density function of the content size under the Pareto distribution is represented as wherein α represents the minimum content size; β represents the variance of the content size distribution;

[0032] a5) setting the vehicle activity wherein d0，n denotes the distance between the vehicle n and the m-th road side unit.

[0033] Further, the step of establishing the edge caching system power consumption model comprises:

[0034] b1) defining the power and bandwidth allocated by the macro base station to the vehicle n as p 0，n and b 0，n , respectively, and the power and bandwidth allocated by the m-th road side unit to the vehicle n as p m，n and b m，n , respectively, and the constraint on the transmission delay of the vehicle n as τ n ;

[0035] b2) calculating the rate r m，n allocated by the m-th road side unit to the vehicle n, i.e.:

[0036]

[0037] wherein G denotes the channel gain between the m-th road side unit and the vehicle n; μ denotes a system parameter, denotes the fast fading gain between the m-th road side unit and the vehicle n, subject to an exponential distribution, d m，n denotes the distance between the m-th road side unit and the vehicle n, and p1 denotes the path loss exponent between the m-th road side unit and the vehicle n; σ2 denotes the additive white Gaussian noise power;

[0038] b3) calculating the rate r 0，n allocated by the macro base station to the vehicle n, i.e.:

[0039]

[0040] wherein G denotes the channel gain between the vehicle n and the macro base station; p2 denotes the path loss exponent between the macro base station and the vehicle n, denotes the fast fading gain between the macro base station and the vehicle n, subject to an exponential distribution; d 0，n denotes the distance between the macro base station and the vehicle n;

[0041] b4) calculating the delay of the vehicle n associated with the m-th road side unit requesting the content f, i.e.:

[0042]

[0043] wherein d bh denotes the backhaul delay; τ1 denotes the delay of the vehicle n to retrieve the content from the m-th road side unit; τ2 denotes the delay of the vehicle n to retrieve the content from the macro base station; τ3 denotes the delay of the vehicle n to retrieve the content from a remote content server.​

[0044] b5) calculating the circuit power consumption δ of the edge caching system cc and the cache power consumption δ ca i.e.

[0045]

[0046]

[0047] wherein λm represents the circuit power consumption of the mth road side unit, wherein λm represents the circuit power consumption of the mth road side unit, wherein λBS represents the circuit power consumption of the macro base station; wherein αm represents the cache power consumption coefficient of the mth road side unit; wherein αm represents the cache power consumption coefficient of the mth road side unit;

[0048] b6) calculating the transmission power consumption δ of the edge caching system t and the backhaul power consumption δ bh i.e.

[0049]

[0050]

[0051] wherein λ1 represents the transmission power consumption coefficient of the road side unit; λ2 represents the transmission power consumption coefficient of the macro base station; ω bh wherein ω represents the backhaul power consumption coefficient.

[0052] Further, the step of obtaining vehicle request data and determining vehicle request content preference comprises:

[0053] 2.1) determining the popularity of the stored content of each macro base station and road side unit using a machine learning algorithm, and ranking the content according to the popularity, and caching the top h contents in the edge server;

[0054] 2.2) calculating the caching probability of content f in the mth road side unit according to the content preference of the vehicle, the vehicle activity, and the content size a n is the vehicle activity;

[0055] 2.3) calculating the similarity of the vehicle content preference using the Pearson correlation coefficient, and establishing a vehicle similarity matrix W; each row or column of the vehicle similarity matrix W corresponds to a specific vehicle, and each element of the matrix represents the similarity score between the corresponding vehicles;

[0056] generating a vehicle degree matrix D using a nearest neighbor graph according to the vehicle similarity matrix W; the vehicle degree matrix D is a diagonal matrix, and each element represents the degree of a vehicle; the degree of a vehicle represents the network connection of the vehicle with the road side unit;

[0057] 2.4) Generate the Laplacian matrix L = D based on the vehicle similarity matrix W and the vehicle degree matrix D. -1 / 2 (DW)D -1 / 2 ), and calculate the eigenvectors corresponding to the K smallest eigenvalues ​​in the Laplacian matrix;

[0058] 2.5) Use the K-means++ algorithm to divide N vehicles into K classes and output the vehicle clustering results;

[0059] In a macro base station, the caching probability of the request content f of the k-th type of vehicle is expressed as: The caching priority of the request content for the k-th type of vehicle is represented as follows: a i This refers to vehicle activity level.

[0060] Furthermore, the steps to determine the content placement scheme for the edge caching system include:

[0061] 3.1) Obtain the cache size, cache probability, cache priority, and content size of macro base stations and roadside units, and initialize y. * =0;

[0062] 3.2) Based on the buffer probability ξ of the m-th roadside unit m Determine the sorted roadside unit cache content set φ m Based on the cache probability of macro base stations Determine the sorted macro base station cache content set Based on the cache priority η, determine the sorted cache priority set η′;

[0063] 3.3) Traverse the roadside cell cache content set φ m Macro base station cache content set All content in the cache is cached based on the relationship between the remaining space and the content size, until there is no remaining cache space;

[0064] 3.4) Generate the optimal buffer vector y for macro base stations and roadside units. * The optimal cache vector y * As a content placement solution for edge caching systems.

[0065] Furthermore, according to the content placement scheme, the steps for macro base stations and roadside units to obtain content data from the content server include:

[0066] 4.1) Construct the elliptic curve E for each macro base station and roadside unit. i And select a point B on the elliptic curve. i As a generator; calculate the order g of the generator;

[0067] 4.2) For each macro base station and roadside unit, randomly generate an integer v as the private key. And calculate the public key and key 0 < v < n;

[0068] 4.3) Send the public key set of each macro base station and roadside unit to the content server; the public key set includes elliptic curve E i Generator B i Public key and key S i ;

[0069] 4.4) The content server periodically generates the AES content key P required for encrypting the content. c and use the public key set to pair the AES content key P c Encryption is performed, and the encrypted content {C1, C2} is sent to the corresponding macro base station and roadside unit;

[0070] The encryption formula is as follows:

[0071] C1=(R*B i )+S i (8)

[0072]

[0073] In the formula, R is from 1 to... n Integers between -1;

[0074] 4.5) After receiving the encrypted content {C1, C2}, the macro base station and roadside unit, according to the private key... and key S i = to decrypt the ciphertext and obtain the content key required for decryption.

[0075] 4.6) The content server uses the content key P c The content to be transmitted is encrypted using the AES encryption algorithm, and the encrypted content is then transmitted to the corresponding macro base station and roadside unit.

[0076] Furthermore, the steps to determine a content distribution scheme based on resource allocation include:

[0077] 5.1) The buffer vector y obtained during the input content placement stage * Available power and bandwidth of macro base stations and roadside units;

[0078] 5.2) Initialize parameter m=1, content size f=1, and iterate through the request content of vehicles under each roadside unit;

[0079] 5.3) judge whether the parameter m is equal to M, if yes, jump to step 5.7), otherwise, let m = m + 1, and enter step 5.4);

[0080] 5.4) judge whether the content size f is equal to F, if yes, return to step 5.3), otherwise, the transmission and backhaul power consumption optimization target function min(δ t +δ bh ) is expressed as a Lagrange function L(b m，n , b 0，n );

[0081] The Lagrange function L(b m，n , b 0，n ) is solved by using the bisection method, the power and bandwidth minimum value of vehicle allocation is obtained, and the transmission power consumption δ t and the backhaul power consumption δ bh are calculated;

[0082] 5.5) calculate the circuit power consumption δ cc and the cache power consumption δ ca ;

[0083] 5.6) calculate the system power consumption δ all = δ cc + δ ca + δ t + δ bh

[0084] 5.7) return the optimal total power consumption δ all generated by the system;

[0085] 5.8) select the content distribution scheme corresponding to the optimal total power consumption δ all .

[0086] A computer readable storage medium, having stored thereon a computer program;

[0087] When the computer program is invoked, the steps of the above method are executed.

[0088] The technical effect of the present application is self-evident. According to the existing research results on content size, the present application uses the Pareto distribution to represent the content size distribution, considers the differentiated requests of users and the limited cache capacity, uses the personalized preferences of users for the cache of content, and reasonably allocates the limited wireless resources, so that the users can obtain content from the nearest MBS or RSU as much as possible, thereby improving the network capacity while reducing the power consumption of the system. In addition, in the content placement stage, the present application uses the improved elliptic curve algorithm (IECC) and the advanced encryption standard (AES) algorithm to encrypt the content for secure transmission and protect the security of data. The present application studies the power consumption minimization problem in the edge network under the constraints of limited wireless resources and cache resources. Considering the heterogeneous vehicle speed, limited cache and wireless resources, the present application models the cache optimization problem as a power consumption minimization problem, uses the relaxation technique to convert the original problem into two stages of content placement and content distribution for solving, and further uses the Pareto distribution to analyze the influence of content size on the cache performance. Considering that the cache can be divided into two key stages (content placement and content distribution) according to the time sequence, the present application converts the power consumption optimization problem into two corresponding stages for solving. In the content placement stage, a content placement algorithm based on content popularity and size is proposed by using the personalized preferences of vehicles, different content sizes and clustering methods. In the content distribution stage, the Lagrange multiplier method is used for effective allocation of power and bandwidth, and then a content distribution algorithm based on the cache strategy is proposed by using the allocated bandwidth and power resources.

[0089] Meanwhile, the present application uses the improved elliptic curve encryption algorithm IECC and the advanced encryption standard AES algorithm to safely place the content, effectively protecting the security of the content. The influence of content size, content number, vehicle number and cache space change on the network performance is studied. Numerical results show that under the heterogeneous vehicle delay requirement, the cache scheme considering the vehicle preference, vehicle activity, content size and power and bandwidth allocation can significantly improve the power consumption of the system. BRIEF DESCRIPTION OF DRAWINGS

[0090] Fig. 1 is a system model diagram;

[0091] Fig. 2 is a system strategy flowchart. DETAILED DESCRIPTION

[0092] The application will be further described in connection with the following examples, which should not be construed as limiting the above-mentioned subject matter of the application to the examples described below. Various replacements and modifications can be made according to ordinary technical knowledge and conventional means in the art without departing from the above-mentioned technical idea of the application, and all such replacements and modifications should be included in the scope of protection of the application.

[0093] Example 1

[0094] Referring to Figs. 1-2 A low-power consumption security cache method for Internet of Vehicles, comprising the following steps:

[0095] 1) Building the edge cache system.

[0096] The edge cache system comprises a system network model, an edge cache model and a power consumption model of the edge cache system.

[0097] The system network model comprises a content server, a macro base station, a roadside unit and a vehicle.

[0098] When the roadside unit detects that a vehicle requests communication in the communication coverage range, the channel state information of the vehicle is obtained through the network and transmitted to the macro base station.

[0099] The macro base station transmits the channel state information to the content server.

[0100] The content server is used to determine the content size requested by the vehicle.

[0101] The edge cache model is used to determine the vehicle activity, the vehicle preference for content and the probability density function of content size.

[0102] The power consumption model of the edge cache system is used to calculate the transmission power consumption and backhaul power consumption of the edge cache system.

[0103] 2) Obtain the vehicle request data and determine the vehicle request content preference; the request content preference comprises cached content and cache priority.

[0104] 3) According to the vehicle request content preference, determine the content placement scheme of the edge cache system.

[0105] 4) According to the content placement scheme, the macro base station and the roadside unit obtain content data from the content server.

[0106] 5) According to the vehicle request data, determine the content distribution scheme based on resource allocation.

[0107] The roadside unit is installed on the roadside and is a logical service unit communicating with the vehicle-mounted unit and the roadside traffic control equipment.

[0108] In the system network model, the macro base station and the roadside unit are both equipped with edge servers with cache capability;

[0109] The macro base station has a transmission power p0 and a bandwidth b0 for downlink content distribution,

[0110] The set of roadside units is denoted as

[0111] The transmission power of the mth roadside unit is p m , and the bandwidth is b m ;

[0112] The set of vehicles is denoted as

[0113] The communication coverage radii of the roadside unit and the macro base station are r1 and r2, and N vehicles and M roadside units are randomly distributed in the area served by the macro base station; within the communication coverage range of the roadside unit, each vehicle can only be associated with one roadside unit, and one roadside unit is associated with vehicles;

[0114] The set of contents available in the content server is The size of the content f is denoted as s f ; F is the number of contents.

[0115] The steps for building the edge cache model include:

[0116] a1) Define the preference vector θ n of vehicle n for content, θ n = {θ n，1 ,..., θ n，f ,..., θ n，F}, where θ n，f represents the request probability of vehicle n for content f;

[0117] a2) Define the cache vector of the roadside unit and the macro base station; when content f is cached in the mth roadside unit, the cache vector y m，f of the roadside unit is 1, otherwise, y m，f = 0;

[0118] When content f is cached in the macro base station, the cache vector y 0，f of the macro base station is 1, otherwise, y 0，f = 0,

[0119] a3) The preference of vehicle n for content f is represented as ε n represents the preference distribution parameter of vehicle n;

[0120] a4) Representing the content size s using a Pareto distribution f The probability density function of the content size under the Pareto distribution is represented as Where a represents the minimum content size; β represents the variance of the content size distribution;

[0121] a5) Setting the vehicle activity Where d 0，n represents the distance between the vehicle n and the MBS.

[0122] The steps for establishing the power consumption model of the edge caching system include:

[0123] b1) Defining the power p 0，n and bandwidth b 0，n allocated by the macro base station to the vehicle n, the power p m，n and bandwidth b m，n allocated by the mth roadside unit to the vehicle n, and the constraint of the transmission delay of the vehicle n represented as τ n ;

[0124] b2) Calculating the rate r m，n allocated by the mth roadside unit to the vehicle n, that is:

[0125]

[0126] In the formula, represents the channel gain between the mth roadside unit and the vehicle n; μ represents the system parameter, represents the fast fading gain between the mth roadside unit and the vehicle n subject to an exponential distribution, d m，n represents the distance between the mth roadside unit and the vehicle n, and ρ1 represents the path attenuation index between the mth roadside unit and the vehicle n; σ 2 represents the additive white Gaussian noise power;

[0127] b3) Calculating the rate r 0，n allocated by the macro base station to the vehicle n, that is:

[0128]

[0129] In the formula, represents the channel gain between the vehicle n and the macro base station; ρ2 represents the path attenuation index between the macro base station and the vehicle n, represents the fast fading gain between the macro base station and the vehicle n subject to an exponential distribution; d 0，n represents the distance between the macro base station and the vehicle n;

[0130] b4) Calculating the delay of the vehicle n requesting the content f associated with the mth roadside unit , that is:

[0131]

[0132] where d bh denotes the backhaul delay; τ1 denotes the delay for the vehicle to obtain content from the mth roadside unit; τ2 denotes the delay for the vehicle to obtain content from the macro base station; τ3 denotes the delay for the vehicle to obtain content from the remote content server;

[0133] b5) calculating the circuit power consumption δ cc of the edge caching system ca i.e.

[0134]

[0135]

[0136] wherein denotes the circuit power consumption of the mth roadside unit, denotes the circuit power consumption of the macro base station; denotes the cache power consumption coefficient of the macro base station, denotes the cache power consumption coefficient of the mth roadside unit;

[0137] b6) calculating the transmission power consumption δ t and the backhaul power consumption δ bh of the edge caching system

[0138]

[0139]

[0140] wherein λ1 denotes the transmission power consumption coefficient of the roadside unit; λ2 denotes the transmission power consumption coefficient of the macro base station; ω bh denotes the backhaul power consumption coefficient.

[0141] The step of obtaining the vehicle request data and determining the vehicle content preference comprises:

[0142] 2.1) determining the popularity of each macro base station and roadside unit storing content by using a machine learning algorithm, and ranking the content according to the popularity, and caching the first h contents in the edge server;

[0143] 2.2) calculating the cache probability of content f in the mth roadside unit according to the content preference of the vehicle, the vehicle activity, and the content size a n is the vehicle activity;

[0144] 2.3) Calculate the similarity of vehicle content preference using Pearson correlation coefficient, and establish the vehicle similarity matrix W; each row or column of the vehicle similarity matrix W corresponds to a specific vehicle, and each element of the matrix represents the similarity score between the corresponding vehicles;

[0145] According to the vehicle similarity matrix W, generate the vehicle degree matrix D using the nearest neighbor graph; the vehicle degree matrix D is a diagonal matrix, and each element represents the degree of the vehicle; the degree of the vehicle represents the network connection of the vehicle with the road side unit;

[0146] 2.4) According to the vehicle similarity matrix W and the vehicle degree matrix D, generate the Laplacian matrix L = D -1 / 2 (D-W)D -1 / 2 ), and calculate the eigenvectors corresponding to the K smallest eigenvalues in the Laplacian matrix;

[0147] 2.5) Divide the N vehicles into K classes using the K-means++ algorithm, and output the vehicle clustering result;

[0148] The cache probability of the kth class of vehicles requesting content f in the macro base station is represented as The cache priority of the kth class of vehicles requesting content is represented as a i is the vehicle activity.

[0149] The steps for determining the content placement scheme of the edge caching system include:

[0150] 3.1) Obtain the cache size, cache probability, cache priority and content size of the macro base station and the road side unit, and initialize y * = 0;

[0151] 3.2) According to the cache probability of the mth road side unit m , determine the sorted road side unit cache content set m ; according to the cache probability of the macro base station , determine the sorted macro base station cache content set ; according to the cache priority , determine the sorted cache priority set

[0152] 3.3) Traverse all contents in the road side unit cache content set m and the macro base station cache content set , and determine whether the current content is cached according to the relationship between the remaining space and the content size, until there is no remaining cache space;

[0153] 3.4) Generate the optimal cache vector y * of the macro base station and the road side unit, and the optimal cache vector y * is taken as the content placement scheme of the edge caching system.

[0154] According to the content placement scheme, the steps for macro base stations and roadside units to obtain content data from the content server include:

[0155] 4.1) Construct the elliptic curve E for each macro base station and roadside unit. i And select a point B on the elliptic curve. i As a generator; calculate the order g of the generator;

[0156] 4.2) For each macro base station and roadside unit, randomly generate an integer v as the private key. And calculate the public key and key 0 < v < n;

[0157] 4.3) Send the public key set of each macro base station and roadside unit to the content server; the public key set includes elliptic curve E i Generator B i Public key and key S i ;

[0158] 4.4) The content server periodically generates the AES content key P required for encrypting the content. c and use the public key set to pair the AES content key P c Encryption is performed, and the encrypted content {C1, C2} is sent to the corresponding macro base station and roadside unit;

[0159] The encryption formula is as follows:

[0160] C1=(R*B i )+S i (8)

[0161]

[0162] In the formula, R is an integer between 1 and n-1;

[0163] 4.5) After receiving the encrypted content {C1, C2}, the macro base station and roadside unit, according to the private key... and key S i = to decrypt the ciphertext and obtain the content key required for decryption.

[0164] 4.6) The content server uses the content key P c The content to be transmitted is encrypted using the AES encryption algorithm, and the encrypted content is then transmitted to the corresponding macro base station and roadside unit.

[0165] The steps to determine a content distribution scheme based on resource allocation include:

[0166] 5.1) The cache vector y obtained in the input content placement stage * , the available power and bandwidth of the macro base station and the road side unit;

[0167] 5.2) Initialize the parameter m = 1, the content size f = 1, and traverse the requested content of each vehicle under the road side unit;

[0168] 5.3) Determine whether the parameter m is equal to M. If yes, jump to step 5.7), otherwise, let m = m + 1, and enter step 5.4);

[0169] 5.4) Determine whether the content size f is equal to F. If yes, return to step 5.3), otherwise, express the transmission and backhaul power consumption optimization objective function min (δ t + δ bh ) as a Lagrangian function L(b m，n , b 0，n );

[0170] Solve the Lagrangian function L(b m，n , b 0，n ) by using the bisection method, obtain the power and bandwidth minimum value of the vehicle allocation, and calculate the transmission power consumption δ t and the backhaul power consumption δ bh ;

[0171] 5.5) Calculate the circuit power consumption δ cc and the cache power consumption δ ca ;

[0172] 5.6) Calculate the system power consumption δ all = δ cc + δ ca + δ t + δ bh

[0173] 5.7) Return the optimal total power consumption δ all generated by the system.

[0174] 5.8) Select the content distribution scheme corresponding to the optimal total power consumption δ all .

[0175] A computer readable storage medium having a computer program stored thereon;

[0176] When the computer program is invoked, the steps of the above method are executed.

[0177] Embodiment 2:

[0178] A low-power safe caching method for vehicle networking, comprising the following steps:

[0179] 1) Build a system network model, mainly including content server, MBS, RSU and vehicle, wherein, the MBS and RSU are equipped with edge servers with cache capability.

[0180] The signal state information of the vehicle passing through the RSU is routed to the MBS, and then the MBS routes all the channel state information to the content server through a backhaul link;

[0181] The coverage radii of the MBS and RSU are r1 and r2 respectively, and the cache capacities of the RSU and MBS are assumed to be c1 and c2 respectively;

[0182] The number of RSUs and MBSs is assumed to be M and N respectively, and the vehicles are randomly distributed in the area served by the MBS;

[0183] The available transmission power of the MBS is p0, and the available transmission power of the RSU m is p m The bandwidth used by the MBS for downlink content distribution is represented as b0, and the bandwidth of the RSU m is represented as b m The set of RSUs is represented as The set of vehicles is represented as It is specified that each vehicle can only be associated with one RSU. The set of available contents in the content server is represented as The size of the content f is represented as The set of vehicles served by the RSU m is represented as

[0184] 2) Build an edge cache model according to the network model constructed in step 1), and the specific construction steps are as follows:

[0185] 2.1) Define the preference vector θ of vehicle n for content n , θ n ={θ n，1 ,..., θ n，f ..., θ n，F}, wherein θ n，f represents the request probability of vehicle n for content f, and it is assumed that the RSU and MBS can only cache part of the content, and the vehicle can directly obtain the content f from the associated RSU, MBS, content server of the backhaul link which caches the content f;

[0186] 2.2) Define the cache vector of RSU and MBS, when the content f is cached in the RSU m, the cache vector y m，f of the RSU m is 1, when other conditions occur, y m，f = 0, and When the content f is cached in the MBS, the cache vector y 0，f of the MBS is 1, when other conditions occur, y 0，f = 0, and

[0187] 2.3) The preference of vehicle n to content f is represented as

[0188] where ε n represents the preference distribution parameter of vehicle n.

[0189] 2.4) The content size s is represented using a Pareto distribution f The probability density function of content size under the distribution is represented as where α represents the minimum content size, and β represents the variance of content size distribution. The larger the β, the more stable the distribution of content size.

[0190] 2.5) The vehicle activity is set as where d 0，n represents the distance between vehicle n and MBS.

[0191] 3) According to the network model and the cache model built in steps 1) and 2), a power consumption model of the edge cache system is constructed, and the specific steps are as follows:

[0192] Define the power and bandwidth allocated by MBS to vehicle n as p 0，n and b 0，n , the power and bandwidth allocated by RSUm to vehicle n as p m，n and b m，n , and the constraint of vehicle n to transmission delay as τ n .

[0193] 3.1) The rate allocated by RSU m to vehicle n is represented as where g n，m represents the channel gain between RSU m and vehicle n, and σ 2 represents the additive white Gaussian noise power.

[0194] 3.2) The rate allocated by MBS to vehicle n is represented as where g 0，n represents the channel gain between vehicle n and MBS.

[0195] 3.3) The channel gain between RSU m and vehicle n is represented as where μ represents a system parameter, represents the fast fading gain between RSU m and vehicle n following an exponential distribution, d m，n represents the distance between RSU m and vehicle n, and ρ1 represents the path attenuation exponent between RSU m and vehicle n.

[0196] 3.4) The channel gain between MBS and vehicle n is denoted as where, ρ2 denotes the path loss exponent between MBS and vehicle n, denotes the fast fading gain between MBS and vehicle n following an exponential distribution, and ρ2 denotes the path loss exponent between MBS and vehicle n.

[0197] 3.5) The latency of vehicle n requesting content f associated with RSU m is denoted as where, d bh denotes the backhaul latency. τ1 denotes the latency of vehicle retrieving content from RSU, τ2 denotes the latency of vehicle retrieving content from MBS, and τ3 denotes the latency of vehicle retrieving content from remote content server.

[0198] 3.6) The circuit power consumption of the system is denoted as where, denotes the circuit power consumption of RSU m, denotes the circuit power consumption of MBS.

[0199] 3.7) The cache power consumption of the system is denoted as where denotes the cache power consumption coefficient of MBS, denotes the cache power consumption coefficient of RSU m.

[0200] 3.8) The transmission power consumption is denoted as where λ1 denotes the transmission power consumption coefficient of RSU, and λ2 denotes the transmission power consumption coefficient of MBS.

[0201] 3.9) The backhaul power consumption is denoted as where, ω bh denotes the power consumption coefficient of backhaul.

[0202] 4) Retrieving vehicle request data and analyzing vehicle content preference

[0203] 4.1) Using machine learning algorithm to analyze vehicle request data from popular websites to obtain content popularity in each MBS and RSU, and pre-cache popular content in edge server close to vehicle;

[0204] 4.2) Regulating content popularity as statistical characteristics of multiple vehicles, not equal to the preference of single vehicle, and only a small part of active vehicles generate requests;

[0205] 4.3) Predicting local content popularity according to vehicle content preference, vehicle activity and content size, and the cache probability of content f in RSU m is denoted as

[0206] 4.4) Calculate the similarity of vehicle content preference using Pearson correlation coefficient, generate vehicle similarity matrix W using the calculated vehicle similarity, generate vehicle degree matrix D using the proximity graph; vehicle activity a n The preference of vehicle n to content f is denoted as θ n，f The size of content f is denoted as s f

[0207] 4.5) Generate Laplacian matrix L using vehicle similarity matrix and vehicle degree matrix (L = D -1 / 2 (D-W)D -1 / 2 ), and calculate the eigenvector corresponding to the K smallest eigenvalues in the Laplacian matrix;

[0208] 4.6) Divide N vehicles into K classes using K-means++ algorithm, and output the final clustering result. The cache probability of content f in MBS belonging to class k is denoted as

[0209] The cache priority of class k in MBS is denoted as

[0210] 5) Determine the content placement strategy of edge caching system

[0211] 5.1) Determine the content placement strategy based on content popularity and content size awareness, the specific steps are as follows:

[0212] 5.1.1) Start inputting the cache size, cache probability, cache priority and content size of MBS and RSU, initialize y * = 0;

[0213] 5.1.2) Use the cache probability of RSU m m to obtain the sorted cache content set φ m , and use the cache probability of MBS to obtain the sorted cache content set Use the cache priority to obtain the sorted cache content set η′;

[0214] 5.1.3) Traverse each content in the content set φ m , , and determine whether the content is cached by judging the relationship between the remaining space and the content size;

[0215] 5.1.4) If there is cache, return to step 5.1.3) for execution, and if there is no remaining cache space, stop traversal.

[0216] 5.1.5) Return the optimal cache vector y * of MBS and RSU.

[0217] ​6) Content security encryption transmission

[0218] After obtaining the optimal content placement strategy, the content is transmitted securely in combination with the IECC and AES, and the specific steps are as follows:

[0219] 6.1) Construct an elliptic curve E of each MBS and RSU i , and select a point B on the elliptic curve i as a generator, and obtain B i by calculation. The order of B i is g.

[0220] 6.2) Randomly select an integer v (0 < v < n) as the private key of each MBS and RSU , and calculate the public key and the secret key

[0221] 6.3) According to the key obtained in step 6.2), each MBS and RSU sends the public key group E i , B i , S i to the content server.

[0222] 6.4) The content server regularly generates the AES content key P c required for encrypted content, and uses the public key group of each MBS and RSU to encrypt P c , and sends it to the corresponding MBS and RSU, and the encryption formula is as follows: C1 = (R * B i ) + S i , where R is an integer between 1 and n-1.

[0223] 6.5) After receiving the ciphertext C1 and C2, the MBS and RSU combine their respective private keys secret key S i , and use to decrypt the ciphertext to obtain the content key required for decryption content;

[0224] 6.6) The content server uses P c to encrypt the content to be transmitted according to the AES encryption algorithm, and transmits the encrypted content to the corresponding MBS and RSU;

[0225] 6.7) After receiving the encrypted content, the MBS and RSU can decrypt the content according to the P c they decrypted in advance to obtain the real content and provide access to vehicles.

[0226] 7) Determine the content distribution strategy based on resource allocation

[0227] After the completion of content placement, the content distribution strategy based on resource allocation needs to be further determined, mainly through effective allocation of the limited power and bandwidth resources of MBS and RSU to obtain the specific power and bandwidth resources of vehicles. The specific steps are as follows:

[0228] 7.1) Assuming that the association of vehicles and RSUs is completed before content distribution;

[0229] 7.2) Start inputting the cache vector y obtained in the content placement stage * , the available power and bandwidth of MBS and RSU;

[0230] 7.3) Initialize m = 1, f = 1, and traverse the requested content of vehicles under each RSU;

[0231] 7.4) Determine whether m is equal to M

[0232] 7.5) If the result of step 7.4) is no, execute m++, and if the result of step 7.4) is yes, jump to step 7.7) for execution;

[0233] 7.6) Determine whether f is equal to F;

[0234] 7.7) If the result of step 7.6) is no, set the transmission and backhaul power consumption optimization objective function min(δ t + δ bh ) as L(b m，n， b 0，n ), and then use the bisection method to solve the first-order partial derivative of the Lagrange function with respect to b m，n , b 0，n , to obtain the minimum value of the allocated power and bandwidth of vehicles, and calculate the transmission power consumption δ t according to step 3.8), and calculate the backhaul power consumption δ bh according to step 3.9);

[0235] 7.8) If the result of step 7.6) is yes, jump to step 7.4) for continuous execution;

[0236] 7.9) Calculate the circuit power consumption δ cc according to step 3.6), and calculate the cache power consumption δ ca according to step 3.7);

[0237] 7.10) Calculate the system power consumption δ all = δ cc + δ ca + δ t + δ bh

[0238] 7.11) return the optimal total power consumption delta generated by the system all .

[0239] Embodiment 3

[0240] A low-power safe cache method for Internet of Vehicles, comprising the following steps:

[0241] 1) Build the edge cache system.

[0242] The edge cache system includes a system network model, an edge cache model, and a power consumption model of the edge cache system.

[0243] The system network model includes a content server, a macro base station, a roadside unit, and a vehicle.

[0244] When the roadside unit detects a vehicle requesting communication within the communication coverage range, the channel state information of the vehicle is obtained through the network and transmitted to the macro base station.

[0245] The macro base station transmits the channel state information to the content server.

[0246] The content server is used to determine the content size requested by the vehicle.

[0247] The edge cache model is used to determine the vehicle activity, the vehicle's preference for content, and the probability density function of the content size.

[0248] The power consumption model of the edge cache system is used to calculate the transmission power consumption and backhaul power consumption of the edge cache system.

[0249] 2) Obtain vehicle request data and determine vehicle request content preference; the request content preference includes cached content and cache priority.

[0250] 3) Determine the content placement scheme of the edge cache system according to the vehicle request content preference.

[0251] 4) According to the content placement scheme, the macro base station and the roadside unit obtain content data from the content server.

[0252] 5) Determine the content distribution scheme based on resource allocation according to the vehicle request data.

[0253] Embodiment 4

[0254] A low-power safe cache method for Internet of Vehicles, the main content is seen in embodiment 3, wherein the roadside unit is installed on the roadside and is a logical service unit communicating with the vehicle-mounted unit and the roadside traffic control equipment.

[0255] Embodiment 5

[0256] A low-power safe caching method for vehicle networking, the main content of which is seen in embodiment 3, wherein in the system network model, the macro base station and the roadside unit are both equipped with edge servers with caching capability;

[0257] The macro base station has a transmission power p0 and a bandwidth b0 for downlink content distribution,

[0258] The set of roadside units is denoted as

[0259] The available transmission power of the mth roadside unit is p m , and the bandwidth is b m ;

[0260] The set of vehicles is denoted as

[0261] The communication coverage radii of the roadside units and the macro base station are r1 and r2, and N vehicles and M roadside units are randomly distributed in the area served by the macro base station; within the communication coverage range of the roadside units, each vehicle can only be associated with one roadside unit, and one roadside unit is associated with vehicles;

[0262] The set of available contents in the content server is The size of content f is denoted as s f ; F is the number of contents.

[0263] Embodiment 6:

[0264] A low-power safe caching method for vehicle networking, the main content of which is seen in embodiment 3, wherein the step of building an edge caching model comprises:

[0265] a1) defining the preference vector θ n of vehicle n for content, θ n = {θ n，1 , …, θ n，f , …, θ n，F}, wherein θ n，f represents the request probability of vehicle n for content f;

[0266] a2) defining the caching vector of the roadside unit and the macro base station; when content f is cached in the mth roadside unit, the caching vector y m，f of the roadside unit is 1, otherwise, y m，f = 0;

[0267] When content f is cached in the macro base station, the caching vector y 0，f of the macro base station is 1, otherwise, y 0，f = 0,

[0268] a3) representing the preference of vehicle n for content f as ε n representing the preference distribution parameter of vehicle n;

[0269] a4) representing the content size s using a Pareto distribution f , the probability density function of content size under the Pareto distribution is represented as where α represents the minimum content size; β represents the variance of content size distribution;

[0270] a5) setting the vehicle activity where, d 0，n represents the distance between vehicle n and MBS.

[0271] Embodiment 7:

[0272] A low-power security cache method for Internet of Vehicles, the main content of which is shown in Embodiment 3, wherein the step of establishing an edge cache system power consumption model comprises:

[0273] b1) defining the power and bandwidth allocated by the macro base station to vehicle n as p 0，n and b 0，n , the power and bandwidth allocated by the mth roadside unit to vehicle n as p m，n and b m，n , and the constraint of vehicle n on transmission delay as τ n ;

[0274] b2) calculating the rate r m，n allocated by the mth roadside unit to vehicle n, that is:

[0275]

[0276] In the formula, represents the channel gain between the mth roadside unit and vehicle n; μ represents a system parameter, represents the fast fading gain between the mth roadside unit and vehicle n obeying an exponential distribution, d m，n represents the distance between the mth roadside unit and vehicle n, and ρ1 represents the path attenuation index between the mth roadside unit and vehicle n; σ 2 represents the additive white Gaussian noise power;

[0277] b3) calculating the rate r 0，n allocated by the macro base station to vehicle n, that is:

[0278]

[0279] In the formula, denotes the channel gain between vehicle n and the macro base station; p2denotes the path loss exponent between the macro base station and vehicle n, denotes the fast fading gain between the macro base station and vehicle n obeying exponential distribution; d 0，n denotes the distance between the macro base station and vehicle n;

[0280] b4) calculating the latency of vehicle n requesting content f associated with the mth road side unit That is,

[0281]

[0282] where d bh denotes the backhaul latency; t1denotes the latency of vehicle obtaining content from the mth road side unit; t2denotes the latency of vehicle obtaining content from the macro base station; t3denotes the latency of vehicle obtaining content from a remote content server;

[0283] b5) calculating the circuit power consumption d of the edge caching system cc and the cache power consumption d ca That is,

[0284]

[0285]

[0286] wherein, denotes the circuit power consumption of the mth road side unit, denotes the circuit power consumption of the macro base station; denotes the cache power consumption coefficient of the macro base station, denotes the cache power consumption coefficient of the mth road side unit;

[0287] b6) calculating the transmission power consumption d of the edge caching system t and the backhaul power consumption d bh That is,

[0288]

[0289]

[0290] wherein, l1denotes the transmission power consumption coefficient of the road side unit; l2denotes the transmission power consumption coefficient of the macro base station; w bh denotes the power consumption coefficient of the backhaul.

[0291] Embodiment 8:

[0292] A low-power safe caching method for Internet of Vehicles, the main content of which is shown in Embodiment 3, wherein the step of obtaining vehicle requested data and determining vehicle content preference preference includes:

[0293] 2.1) Use machine learning algorithm to determine the popularity of each macro base station and roadside unit storage content, and sort the content according to the popularity, and cache the first h content in the edge server;

[0294] 2.2) According to the content preference of the vehicle, the vehicle activity, and the content size, calculate the cache probability of the content f in the mth roadside unit a n For vehicle activity;

[0295] 2.3) Calculate the similarity of vehicle content preference using Pearson correlation coefficient, and establish vehicle similarity matrix W; each row or column of the vehicle similarity matrix W corresponds to a specific vehicle, and each element of the matrix represents the similarity score between the corresponding vehicles;

[0296] According to the vehicle similarity matrix W, generate the vehicle degree matrix D using the neighbor graph; the vehicle degree matrix D is a diagonal matrix, and each element represents the degree of the vehicle; the degree of the vehicle represents the network connection of the vehicle with the roadside unit;

[0297] 2.4) According to the vehicle similarity matrix W and the vehicle degree matrix D, generate the Laplacian matrix L = D -1 / 2 (D-W)D -1 / 2 ), and calculate the eigenvector corresponding to the K smallest eigenvalue in the Laplacian matrix;

[0298] 2.5) Use K-means++ algorithm to divide N vehicles into K classes, and output the vehicle clustering result;

[0299] The cache probability of the kth class of vehicles requesting content f in the macro base station is represented as The cache priority of the kth class of vehicles requesting content is represented as a i For vehicle activity.

[0300] Embodiment 9:

[0301] A low-power safe cache method for Internet of Vehicles, the main content of which is shown in Embodiment 3, wherein the step of determining the content placement scheme of the edge cache system comprises:

[0302] 3.1) Obtain the cache size, cache probability, cache priority and content size of the macro base station and the roadside unit, and initialize y * = 0;

[0303] 3.2) According to the cache probability of the mth roadside unit m , determine the sorted roadside unit cache content set m ; according to the cache probability of the macro base station , determine the sorted macro base station cache content set According to the cache priority η, determine the sorted cache priority set η';

[0304] 3.3) Traverse the roadside unit cache content set φ m , the macro base station cache content set All contents, according to the relationship between the remaining space and the content size, determine whether the current content is cached, until there is no remaining cache space;

[0305] 3.4) Generate the optimal cache vector y of the macro base station and the roadside unit * , the optimal cache vector y * As the content placement scheme of the edge cache system.

[0306] Embodiment 10:

[0307] A low-power safe cache method for Internet of Vehicles, the main content of which is seen in Embodiment 3, wherein according to the content placement scheme, the step of the macro base station and the roadside unit obtaining content data from the content server includes:

[0308] 4.1) Construct an elliptic curve E i for each macro base station and roadside unit i Select a point B i as the generator; calculate the order g of the generator;

[0309] 4.2) For each macro base station and roadside unit, randomly generate an integer v as a private key And calculate the public key And the secret key 0

[0310] 4.3) Send the public key group of each macro base station and roadside unit to the content server; the public key group includes the elliptic curve E i , the generator B i , the public key And the secret key S i ;

[0311] 4.4) The content server periodically generates the AES content key P required for encrypted content c , and uses the public key group to encrypt the AES content key P c , and sends the encrypted content {C1, C2} to the corresponding macro base station and roadside unit;

[0312] Wherein, the encryption formula is as follows:

[0313] C1 = (R * B i ) + S i ; (8)

[0314]

[0315] wherein R is an integer between 1 and n-1;

[0316] 4.5) After the macro base station and the road side unit receive the encrypted content {C1, C2}, the private key and the key S i are used to decrypt the ciphertext to obtain the content key required for decrypting the content

[0317] 4.6) The content server uses the content key P c to encrypt the content to be transmitted according to the AES encryption algorithm, and transmits the encrypted content to the corresponding macro base station and road side unit;

[0318] Embodiment 11:

[0319] A low-power secure caching method for Internet of Vehicles, the main content of which is shown in Embodiment 3, wherein the step of determining a content distribution scheme based on resource allocation comprises:

[0320] 5.1) Input the caching vector y * obtained in the content placement stage, the available power and bandwidth of the macro base station and the road side unit;

[0321] 5.2) Initialize the parameters m = 1 and content size f = 1, and traverse the requested content of each vehicle under the road side unit;

[0322] 5.3) Determine whether the parameter m is equal to M, if yes, jump to step 5.7), otherwise, let m = m + 1, and enter step 5.4);

[0323] 5.4) Determine whether the content size f is equal to F, if yes, return to step 5.3), otherwise, express the transmission and backhaul power consumption optimization objective function min(δ t + δ bh ) as a Lagrangian function L(b m，n , b 0，n );

[0324] Solve the Lagrangian function L(b m，n , b 0，n ) by using the bisection method, obtain the power and bandwidth minimum value allocated to the vehicle, and calculate the transmission power consumption δ t and the backhaul power consumption δ bh ;

[0325] 5.5) Calculate the circuit power consumption δ cc and the caching power consumption δ ca ;

[0326] 5.6) Calculate the system power consumption δ all = δ cc + δca + δ t + δ bh

[0327] 5.7) return the optimal total power consumption delta all ;

[0328] 5.8) select the content distribution scheme corresponding to the optimal total power consumption delta all .

[0329] Embodiment 12:

[0330] A computer readable storage medium having stored thereon a computer program;

[0331] When the computer program is invoked, it performs the steps of the method of embodiments 3-11.

Claims

1. A low-power secure caching method for vehicle-to-everything (V2X) networks, characterized in that, Includes the following steps: 1) Build an edge caching system; The edge caching system includes a system network model, an edge caching model, and a power consumption model for the edge caching system. The system network model includes content servers, macro base stations, roadside units, and vehicles; When the roadside unit detects a vehicle requesting communication within its coverage area, it obtains the vehicle's channel status information through the network and transmits it to the macro base station. The macro base station transmits channel status information to the content server; The content server is used to determine the size of the content requested by the vehicle. The edge caching model is used to determine the probability density function of vehicle activity, vehicle preference for content, and content size; The power consumption model of the edge caching system is used to calculate the transmission power consumption and backhaul power consumption of the edge caching system. 2) Obtain vehicle request data and determine vehicle request content preferences; the request content preferences include cached content and cache priority; 3) Determine the content placement scheme for the edge caching system based on the vehicle's requested content preferences; 4) Based on the content placement plan, macro base stations and roadside units obtain content data from the content server; 5) Based on vehicle request data, determine a content distribution scheme based on resource allocation; The steps to determine the content placement scheme for an edge caching system include: 3.1) Obtain the cache size, cache probability, cache priority, and content size of macro base stations and roadside units, and initialize y. * =0; 3.2) Based on the buffer probability ξ of the m-th roadside unit m Determine the sorted roadside unit cache content set φ m Based on the cache probability of macro base stations Determine the sorted macro base station cache content set Based on the cache priority η, determine the sorted cache priority set η′; 3.3) Traverse the roadside cell cache content set φ m Macro base station cache content set All content in the cache is cached based on the relationship between the remaining space and the content size, until there is no remaining cache space; 3.4) Generate the optimal buffer vector y for macro base stations and roadside units. * The optimal cache vector y * As a content placement solution for edge caching systems; The steps to determine a content distribution scheme based on resource allocation include: 5.1) The buffer vector y obtained during the input content placement stage * Available power and bandwidth of macro base stations and roadside units; 5.2) Initialize parameter m' = 1, content f = 1, and iterate through the request content of vehicles under each roadside unit; 5.3) Determine if parameter m' is equal to M'. If yes, proceed to step 5.7); otherwise, set m' = m' + 1 and proceed to step 5.

4. 5.4) Determine if the content f is equal to F. If yes, return to step 5.3); otherwise, optimize the transmission and backhaul power consumption objective function min(δ). t +δ bh ) is represented as the Lagrangian function L(b) m,n ,b 0,n F represents the number of contents; Solving the Lagrangian function L(b) using the bisection method m,n ,b 0,n This allows us to obtain the minimum power and bandwidth allocated to the vehicle, and calculate the transmission power consumption δ. t and return power consumption δ bh ; 5.5) Calculate the circuit power consumption δ cc and cache power consumption v ca ; 5.6) Calculate the system power consumption δ all =δ cc +δ ca +δ t +δ bh 5.7) Return the optimal total power consumption δ generated by the system. all ; 5.8) Select the optimal total power consumption δ all The corresponding content distribution scheme.

2. The low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, The roadside unit is installed on the roadside and is a logical service unit that communicates with the vehicle-mounted unit and the roadside traffic control equipment.

3. The low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, In the system network model, both macro base stations and roadside units are equipped with edge servers with caching capabilities; The macro base station has an available transmission power of p0 and a bandwidth of b0 for downlink content distribution. The set of roadside units is denoted as The available transmission power of the m-th roadside unit is p m The bandwidth is b m ; The set of vehicles is denoted as The communication coverage radii of roadside units and macro base stations are r1 and r2, respectively. N vehicles and M roadside units are randomly distributed within the area served by the macro base station. Within the communication coverage area of ​​a roadside unit, each vehicle can only be associated with one roadside unit, and one roadside unit can only be associated with... Individual vehicle associations; The collection of content available in the content server is The size of the content f is represented by s. f ; F represents the amount of content.

4. The low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, The steps to build an edge caching model include: 1) Define the content preference vector θ for vehicle n. n θ n ={θ n,1 ,…,θ n,f …,θ n,F }, where θ m,f This represents the probability that vehicle n requests content f; 2) Define the cache vectors for roadside units and macro base stations; when content f is cached in the m-th roadside unit, the cache vector y of the roadside unit is... m,f =1, otherwise, y m,f =0; When content f is cached in the macro base station, the macro base station's cache vector y 0,f =1, otherwise, y 0,f =0, 3) Represent the preference of vehicle n for content f as: ε n The parameter representing the preference distribution of vehicle n; 4) Use Pareto distribution to represent content size s f The probability density function of content size under the Pareto distribution is expressed as: Where α represents the minimum content size; β represents the variance of the content size distribution; 5) Set vehicle activity level Where, d 0,n This represents the distance between vehicle n and MBS.

5. A low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, The steps to establish a power consumption model for an edge caching system include: 1) Define the power and bandwidth allocated by the macro base station to vehicle n as p, respectively. 0,n and b 0,n The power and bandwidth allocated to vehicle n by the m-th roadside unit are p, respectively. m,n and b m,n The constraint of vehicle n on transmission delay is expressed as τ. n ; 2) Calculate the speed r assigned to vehicle n by the m-th roadside unit. m,n ,Right now: In the formula, μ represents the channel gain between the m-th roadside unit and vehicle n; μ represents the system parameters. d represents the fast fading gain that follows an exponential distribution between the m-th roadside unit and vehicle n. m,n ρm represents the distance between the m-th roadside unit and vehicle n, and ρ1 represents the path decay exponent between the m-th roadside unit and vehicle n; σm 2 This represents the power of additive white Gaussian noise; 3) Calculate the rate r allocated to vehicle n by the macro base station. 0,n ,Right now: In the formula, ρn represents the channel gain between vehicle n and the macro base station; ρ2 represents the path attenuation exponent between the macro base station and vehicle n. d represents the fast fading gain between the macro base station and vehicle n, which follows an exponential distribution; 0,m This represents the distance between the macro base station and vehicle n; 4) Calculate the latency of the request content f of vehicle n associated with the m-th roadside unit. Right now: In the formula, d bh τ1 represents the latency of the vehicle obtaining content from the m-th roadside unit; τ2 represents the latency of the vehicle obtaining content from the macro base station; τ3 represents the latency of the vehicle obtaining content from the remote content server. 5) Calculate the circuit power consumption δ of the edge buffer system. cc and cache power consumption δ ca ,Right now: In the formula, This represents the circuit power consumption of the m-th roadside unit. Indicates the circuit power consumption of the macro base station; This represents the cache power consumption coefficient of the macro base station. This represents the cache power consumption coefficient of the m-th roadside unit; 6) Calculate the transmission power δ of the edge buffer system. t and return power consumption δ bh ,Right now: In the formula, λ1 represents the transmission power consumption coefficient of the roadside unit; λ2 represents the transmission power consumption coefficient of the macro base station; ω bh This represents the power consumption factor for the return trip.

6. A low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, The steps for obtaining vehicle request data and determining vehicle request content preferences include: 1) Use machine learning algorithms to determine the popularity of the content stored in each macro base station and roadside unit, sort the content according to the popularity, and cache the top h content on the edge server; 2) Calculate the caching probability of content f in the m-th roadside unit based on the vehicle's content preferences, vehicle activity level, and content size. a n Vehicle activity level; s f θ represents the size of the content f; n,f This represents the probability that vehicle n requests content f; 3) Calculate the similarity of vehicle content preferences using the Pearson correlation coefficient and establish a vehicle similarity matrix W; each row or column of the vehicle similarity matrix W corresponds to a specific vehicle, and each element of the matrix represents the similarity score between the corresponding vehicles. Based on the vehicle similarity matrix W, a vehicle degree matrix D is generated using a nearest neighbor graph; the vehicle degree matrix D is a diagonal matrix, and each element represents the degree of a vehicle; the degree of a vehicle characterizes the network connectivity between the vehicle and the roadside unit. 4) Generate the Laplacian matrix L = D based on the vehicle similarity matrix W and the vehicle degree matrix D. -1 / 2 (DW)D -1 / 2 ), and calculate the eigenvectors corresponding to the K smallest eigenvalues ​​in the Laplacian matrix; 5) Use the K-means++ algorithm to divide N vehicles into K classes and output the vehicle clustering results; In a macro base station, the caching probability of the request content f of the k-th type of vehicle is expressed as: The caching priority of the request content for the k-th type of vehicle is represented as follows: a i This refers to vehicle activity level.

7. A low-power secure caching method for vehicle-to-everything (V2X) communication as described in claim 1, characterized in that, According to the content placement scheme, the steps for macro base stations and roadside units to obtain content data from the content server include: 1) Construct the elliptic curve E for each macro base station and roadside unit. i And select a point B on the elliptic curve. i As a generator; calculate the order g of the generator; 2) For each macro base station and roadside unit, randomly generate an integer v as the private key. And calculate the public key and key 0 <v<g; 3) Send the public key set of each macro base station and roadside unit to the content server; the public key set includes elliptic curve E i Generator B i Public key and key S i ; 4) The content server periodically generates the AES content key P required for encrypting the content. c and use the public key set to pair the AES content key P c Encryption is performed, and the encrypted content {C1, C2} is sent to the corresponding macro base station and roadside unit; The encryption formula is as follows: C1=(R*B i )+S i ; (8) In the formula, R is an integer between 1 and g-1; 5) After receiving the encrypted content {C1, C2}, the macro base station and the roadside unit use the private key... and key S i The ciphertext is decrypted to obtain the content key required for decryption. 6) The content server uses the content key P c The content to be transmitted is encrypted using the AES encryption algorithm, and the encrypted content is then transmitted to the corresponding macro base station and roadside unit.

8. A computer-readable storage medium, characterized in that, It contains computer programs; When the computer program is invoked, it performs the steps of the method according to any one of claims 1-7.

Images