Data content caching method for vehicle clusters

Abstract

The present invention discloses a data content caching method for vehicle clusters, comprising: obtaining the relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles, so as to cluster all vehicles within the coverage range of a macro base station according to the relative mobility and content preference similarity; the macro base station constructs a global model so that the cluster head vehicle of each vehicle cluster downloads the global model parameters; the cluster head vehicle of each vehicle cluster trains the local model according to the global model parameters and local training samples to obtain updated local model parameters; the macro base station aggregates all received updated local model parameters and updates the global model parameters until the global model converges; the cluster head vehicle of each vehicle cluster caches corresponding content according to the trained global model; thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle networking, and in particular to a data content caching method for a vehicle cluster, a computer-readable storage medium, and a computer device. Background Art

[0002] With the rapid growth of vehicle scale and the continuous emergence of various emerging Internet of Vehicles (IoV) services, the number of in-vehicle service content request data is constantly increasing, which puts tremendous pressure on IoV communications. Edge caching technology is one of the key technologies to cope with the explosive growth of IoV service content requests. IoV-oriented edge caching can pre-cache and share popular service content data in vehicle nodes or edge servers. Based on edge caching, vehicles can obtain content data through vehicle-to-vehicle or vehicle-to-infrastructure communication, without having to obtain it from remote cloud servers, thereby reducing content acquisition latency and the communication burden of the return link, further ensuring the driving experience and traffic safety. However, the highly dynamic vehicle network topology brings many challenges to content caching in the IoV. How to use limited storage resources to obtain the optimal content caching strategy in a highly dynamic IoV environment has become one of the key research difficulties. Summary of the Invention

[0003] The present invention aims to at least partially address one of the technical problems in the aforementioned technologies. To this end, one object of the present invention is to propose a data content caching method for vehicle clusters that can effectively improve the cache hit rate and cache delivery rate of content while reducing the dynamics of the vehicle network topology.

[0004] A second object of the present invention is to provide a computer-readable storage medium.

[0005] A third object of the present invention is to provide a computer device.

[0006] To achieve the above-mentioned purpose, an embodiment of the first aspect of the present invention proposes a data content caching method for vehicle clusters, comprising the following steps: obtaining the relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles, so as to cluster all vehicles within the coverage of the macro base station according to the relative mobility and the content preference similarity, wherein each vehicle cluster includes a cluster head vehicle; the macro base station constructs a global model so that the cluster head vehicle of each vehicle cluster downloads the global model parameters; the cluster head vehicle of each vehicle cluster trains the local model according to the global model parameters and local training samples to obtain updated local model parameters, and uploads the updated local model parameters to the macro base station; the macro base station aggregates all received local model parameters and updates the global model parameters until the global model converges to complete the training of the global model; the cluster head vehicle of each vehicle cluster obtains the trained global model so as to perform corresponding content caching according to the trained global model.

[0007] According to the data content caching method for vehicle clusters according to an embodiment of the present invention, first, the relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles are obtained, so that all vehicles within the coverage of the macro base station are clustered according to the relative mobility and content preference similarity, wherein each vehicle cluster includes a cluster head vehicle; then, the macro base station constructs a global model so that the cluster head vehicle of each vehicle cluster downloads the global model parameters; then, the cluster head vehicle of each vehicle cluster trains the local model according to the global model parameters and local training samples to obtain updated local model parameters, and uploads the updated local model parameters to the macro base station; then, the macro base station aggregates all received local model parameters and updates the global model parameters until the global model converges to complete the training of the global model; finally, the cluster head vehicle of each vehicle cluster obtains the trained global model so as to perform corresponding content caching according to the trained global model; thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology.

[0008] In addition, the data content caching method for vehicle clusters proposed in the above embodiment of the present invention may also have the following additional technical features:

[0009] Optionally, the relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles are obtained, including: obtaining the relative speed and relative distance between each vehicle within the communication range and its neighboring vehicles, so as to obtain relative mobility based on the relative speed and relative distance; obtaining the type and access duration of historically accessed content of each vehicle within the communication range, so as to obtain the interest preference characteristics of each vehicle based on the type and access duration of historically accessed content, and obtaining the content preference similarity between each vehicle and its neighboring vehicles based on the interest preference characteristics.

[0010] Optionally, all vehicles within the coverage of the macro base station are clustered according to the relative mobility and the content preference similarity, including: obtaining the cluster head factor value of each vehicle according to the relative mobility between each vehicle and its neighboring vehicles and the corresponding number of neighboring vehicles, and selecting the vehicle with the smallest cluster head factor value as the cluster head vehicle; the cluster head vehicle broadcasts a cluster head message to neighboring vehicles within the communication range, so that the vehicles receiving the cluster head message select the cluster head vehicle with the smallest cluster head factor value to send a cluster entry application; the cluster head vehicle obtains the cluster entry factor value of the vehicle according to the relative mobility and the content preference similarity, and clusters the vehicles in order of the cluster entry factor value from small to large to complete vehicle clustering.

[0011] Optionally, the interest preference characteristics of each vehicle are obtained according to the following formula:

[0012]

[0013]

[0014]

[0015] Among them, I k For vehicle u k A preference vector for all content types, for u k Preference for content of type m, C j For any popular content contained in the macro base station content library C, the shortest time required to access the complete content is CT j , C can be divided into N a categories, the content class set can be expressed as cha m is the content set of the mth type, PC k For vehicle u k A collection of accessed content in the past period of time. When C j ∈PC k , For vehicle u k Content Cj The actual visit duration.

[0016] Optionally, the cluster head factor value of each vehicle is obtained according to the following formula:

[0017]

[0018]

[0019]

[0020]

[0021]

[0022] Among them, α1, α2, α2 are weighting coefficients, α1, α2, α2∈(0,1), α1+α2+α2=1; For vehicle u k The velocity vector, For vehicle u q The velocity vector, two-dimensional coordinates (x k ,y k ) represents vehicle u k The instantaneous position, two-dimensional coordinates (x q ,y q ) represents vehicle u q The instantaneous position, DNL k for u k Neighbor node list, |DNL k |=n k , RV(k) is u k The relative mobility of its direct neighbor node, RD(k) is u k The relative distance to its direct neighbor nodes; the smaller M(k), the better k The smaller the relative speed and relative distance to its direct neighbor nodes and the more direct neighbor nodes it has, the higher the u k The more likely it is to become a cluster head.

[0023] Optionally, the content preference similarity between each vehicle and its neighboring vehicles is obtained according to the following formula:

[0024]

[0025] Among them, I k and I o Represents vehicle node u k and vehicle node u o Preference vector for all content types, sim k,o A larger value indicates that the content access preferences of the two nodes are more similar.

[0026] Optionally, obtain the clustering factor value according to the following formula:

[0027]

[0028]

[0029]

[0030]

[0031] Among them, λ1, λ2∈(0,1), λ1+λ2=1, λ1 and λ2 represent the influence of mobility and interest preference on the clustering factor respectively; Vehicle cluster VC o average speed; For VC o The average preference of all nodes in the m-th category for content.

[0032] Optionally, the local training samples include historical content request data, vehicle cache space, and a corresponding optimal vehicle cluster content cache strategy, wherein the optimal vehicle cluster content cache strategy is obtained by solving the following problem of maximizing the intra-cluster vehicle cache revenue:

[0033]

[0034]

[0035]

[0036]

[0037]

[0038] r i,p,r =B V2V log2(1+SNR i,p,r )

[0039] The constraints are:

[0040]

[0041]

[0042]

[0043] k∈{0,1,…,ω i}

[0044] Among them, G i Vehicle cluster VC iBased on the benefits obtained by the cache strategy Θ, VC i Contains cluster head CH i and ω i cluster member nodes, namely In the decision cycle t, the vehicle cluster VC i Cluster head vehicle CH i The caching decision based on the caching strategy Θ is Representatives i,k Cache content C q ;0-1 matrix Represents vehicle cluster VC i The content cache information matrix in decision cycle t, Indicates u i,k Cached content C q , Indicates u i,k Content C is not cached q ; Based on cache decision A i,t , Will enter the next state req i,r,q (t)=1 means that in the decision cycle t, u i,r Content request C was made q , req i,r,q (t)=0, otherwise; reward PR q Cache content for vehicle C q After that, each time you provide this content to other vehicles, you can get rewards, PR q By Ch i According to the content C q In VC i Popularity within pl i,q and content size sc q calculate, is a constant factor; trs i,p,r for u i,p to u i,r Transmission content C q Time required, τ i,r,q for u i,r Request content C q The longest tolerable waiting time, parameter γ i,r,q Used to indicate whether the content request is delivered successfully. i,p,r <τ i,r,q That is γ i,r,q =1, u i,r Request content C q Delivery is successful, otherwise delivery fails; i,p,r Characterization u i,p with u i,r The transmission rate between V2Vis the V2V channel transmission bandwidth, SNR i,p,r Providing a vehicle for content i,p Request vehicle u from content i,r Signal-to-noise ratio when transmitting content through a V2V transmission channel; i,k,t for u i,k The amount of available cache space at time t.

[0045] To achieve the above-mentioned purpose, the second embodiment of the present invention proposes a computer-readable storage medium, on which a data content caching program for a vehicle cluster is stored. When the data content caching program for a vehicle cluster is executed by a processor, the data content caching method for a vehicle cluster as described above is implemented.

[0046] According to the computer-readable storage medium of an embodiment of the present invention, by storing a data content caching program for a vehicle cluster, the processor implements the above-mentioned data content caching method for a vehicle cluster when executing the data content caching program for a vehicle cluster, thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology.

[0047] To achieve the above-mentioned purpose, an embodiment of the third aspect of the present invention proposes a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the data content caching method for vehicle clusters as described above is implemented.

[0048] According to the computer device of an embodiment of the present invention, a data content caching program for a vehicle cluster is stored in a memory, so that when the processor executes the data content caching program for a vehicle cluster, the data content caching method for a vehicle cluster as described above is implemented, thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 1. A schematic flow chart of a data content caching method for a vehicle cluster according to an embodiment of the present invention;

[0050] Figure 2 A schematic diagram of a vehicle networking scenario according to an embodiment of the present invention;

[0051] Figure 3 Schematic diagram of a federated learning architecture according to an embodiment of the present invention. DETAILED DESCRIPTION

[0052] The following describes embodiments of the present invention in detail, examples of which are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present invention, and are not to be construed as limiting the present invention.

[0053] To better understand the above technical solutions, exemplary embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although exemplary embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be limited by the embodiments described herein. Instead, these embodiments are provided to enable a more thorough understanding of the present invention and to fully convey the scope of the present invention to those skilled in the art.

[0054] In order to better understand the above technical solution, the above technical solution will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0055] Figure 1 FIG. 1 is a flow chart of a data content caching method for a vehicle cluster according to an embodiment of the present invention. Figure 1 As shown, the data content caching method for vehicle clusters includes the following steps:

[0056] S101, obtaining relative mobility and content preference similarity between each vehicle within a communication range and its neighboring vehicles, so as to cluster all vehicles within the coverage of a macro base station according to the relative mobility and content preference similarity, wherein each vehicle cluster includes a cluster head vehicle.

[0057] As an embodiment, the relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles are obtained, including: obtaining the relative speed and relative distance between each vehicle within the communication range and its neighboring vehicles, so as to obtain relative mobility based on the relative speed and relative distance; obtaining the type and access duration of historically accessed content of each vehicle within the communication range, so as to obtain the interest preference characteristics of each vehicle based on the type and access duration of historically accessed content, and obtaining the content preference similarity between each vehicle and its neighboring vehicles based on the interest preference characteristics.

[0058] As a specific embodiment, the interest preference characteristics of each vehicle are obtained according to the following formula:

[0059]

[0060]

[0061]

[0062] Among them, Ik For vehicle u k A preference vector for all content types, for u k Preference for content of type m, C j For any popular content contained in the macro base station content library C, the shortest time required to access the complete content is CT j , C can be divided into N a categories, the content class set can be expressed as cha m is the content set of the mth type, PC k For vehicle u k A collection of accessed content in the past period of time. When C j ∈PC k , For vehicle u k Content C j The actual visit duration.

[0063] As a specific embodiment, the content preference similarity between each vehicle and its neighboring vehicles is obtained according to the following formula:

[0064]

[0065] Among them, I k and I o Represents vehicle node u k and vehicle node u o Preference vector for all content types, sim k,o A larger value indicates that the content access preferences of the two nodes are more similar.

[0066] As an embodiment, all vehicles within the coverage of a macro base station are clustered according to relative mobility and content preference similarity, including: obtaining a cluster head factor value for each vehicle according to the relative mobility between each vehicle and its neighboring vehicles and the corresponding number of neighboring vehicles, and selecting the vehicle with the smallest cluster head factor value as the cluster head vehicle; the cluster head vehicle broadcasts a cluster head message to neighboring vehicles within the communication range, so that the vehicles receiving the cluster head message select the cluster head vehicle with the smallest cluster head factor value to send a cluster entry application; the cluster head vehicle obtains the cluster entry factor value of the vehicle according to the relative mobility and content preference similarity, and clusters the vehicles in ascending order of the cluster entry factor value to complete vehicle clustering.

[0067] As a specific embodiment, the cluster head factor value of each vehicle is obtained according to the following formula:

[0068]

[0069]

[0070]

[0071]

[0072]

[0073] Among them, α1, α2, α2 are weighting coefficients, α1, α2, α2∈(0,1), α1+α2+α2=1; For vehicle u k The velocity vector, For vehicle u q The velocity vector, two-dimensional coordinates (x k ,y k ) represents vehicle u k The instantaneous position, two-dimensional coordinates (x q ,y q ) represents vehicle u q The instantaneous position, DNL k for u k Neighbor node list, |DNL k |=n k , RV(k) is u k The relative mobility of its direct neighbor node, RD(k) is u k The relative distance to its direct neighbor nodes; the smaller M(k), the better k The smaller the relative speed and relative distance to its direct neighbor nodes and the more direct neighbor nodes it has, the higher the u k The more likely it is to become a cluster head.

[0074] That is, the cluster head factor value of each vehicle is calculated by weighted summing the relative mobility and the number of neighboring nodes of the vehicle.

[0075] As a specific embodiment, the clustering factor value is obtained according to the following formula:

[0076]

[0077]

[0078]

[0079]

[0080] Among them, λ1, λ2∈(0,1), λ1+λ2=1, λ1 and λ2 represent the influence of mobility and interest preference on the clustering factor respectively; Vehicle cluster VC o average speed; For VC o The average preference of all nodes in the m-th category for content.

[0081] That is, the cluster head vehicle calculates the clustering factor value of the vehicle by weighted summation based on relative mobility and content preference similarity.

[0082] As an example, let N VC The vehicle cluster set is The corresponding cluster head set is In the initial stage, all vehicles are isolated nodes; isolated vehicle u k Periodically broadcast Hello packets within the V2V communication range with a radius of R, and exchange vehicle ID, location, speed, direction and other information with other nodes; when u k When receiving Hello packets from other nodes, the node will be recorded in u k Direct neighbor node list DNL k ;u k The angle between its movement direction and that of each direct neighbor node will be further calculated. k,q >π / 4,u k Will put u q From DNL k , so as to ensure that the nodes of the same vehicle cluster have the same movement direction; each vehicle node u k Each vehicle will calculate its own cluster head factor value and broadcast it to its direct neighbor nodes; if vehicle u k The cluster head factor value is less than DNL k The cluster head factor value of all nodes in the cluster, then u k will be elected as the cluster head; otherwise, DNL k The node with the smallest cluster head factor value becomes the cluster head. After the cluster head election is completed, a cluster head with vehicle node u h Vehicle cluster VC with cluster head m , and u h Recorded as CH m ; CH m The cluster head message will be broadcast to its direct neighbor nodes, including CH m The vehicle node u that receives the cluster head message o Will CH m Send a cluster entry request, if u o Receive cluster head messages sent by multiple cluster head nodes, u o The cluster head node with the smallest cluster head factor value will be selected to send a cluster entry request; mAfter receiving the clustering request from other nodes, the clustering factor of each node will be calculated based on the relative mobility and interest preference characteristics, and the nodes will be added to the cluster member list in order of the clustering factor from small to large, and the request message will be replied. m The number of cluster member lists exceeds N max or u o The clustering factor value exceeds the threshold μ, u o The application of cluster entry may be rejected; isolated nodes that are not currently clustered are considered to be a cluster of their own, but these isolated nodes will continue to try to join the cluster; in the vehicle cluster VC m After generation, in order to maintain the stability of the vehicle cluster network topology, CH m The vehicle cluster needs to be dynamically maintained and updated until the vehicle cluster VC m Disbanded; CH m Periodically check the status of the cluster member list and the direct neighbor cluster head list; when CH m When a cluster member node does not receive periodic messages within a period, it will be removed from the current cluster member list; in addition, CH m The average cluster speed of the current cluster will be calculated. When the cluster member u o Speed ​​and VC m The difference in average speed exceeds the threshold value ρ,u o will also be removed from the cluster member list; when CH m If the direct neighbor cluster head list is not empty, cluster merging will be performed; the cluster head with a lower cluster head factor value will be elected as the new cluster head and send a cluster head message, and other cluster member nodes will re-execute the clustering process.

[0083] Through the mobility-based vehicle clustering mechanism described above, highly dynamic vehicles can be divided into vehicle clusters with relatively stable network topology. This clustering mechanism comprehensively considers the vehicle's location, driving speed, the number of members that the cluster can accommodate, and the vehicle's interest preference characteristics to ensure that the communication load of the cluster head vehicle is within a tolerable range, the vehicle cluster structure tends to be stable, and the content request preferences of vehicles within the cluster tend to be consistent.

[0084] S102: The macro base station constructs a global model so that the cluster head vehicle of each vehicle cluster downloads global model parameters.

[0085] That is to say, if Figure 2 As shown, a data content caching method for vehicle clusters is applied to an Internet of Vehicles edge caching system; the system includes a macro base station and vehicles, and the vehicles are divided into multiple vehicle clusters with similar mobility and similar content preference characteristics. The vehicles in the cluster can be divided into content requesting vehicles, content providing vehicles and other vehicles. The cluster head vehicle and the macro base station have the same structural neural network model.

[0086] S103 , the cluster head vehicle of each vehicle cluster trains the local model according to the global model parameters and the local training samples to obtain updated local model parameters, and uploads the updated local model parameters to the macro base station.

[0087] That is to say, Figure 3 As shown in the figure, each cluster head vehicle participating in the federated learning training iterates and updates the local model parameters based on its local data set and downloaded global model parameters; when the number of iterations is reached, each cluster head vehicle uploads the updated local model parameters to the macro base station.

[0088] As an embodiment, the local training samples include historical content request data, vehicle cache space, and the corresponding optimal vehicle cluster content cache strategy, wherein the optimal vehicle cluster content cache strategy is obtained by solving the following problem of maximizing the intra-cluster vehicle cache revenue:

[0089]

[0090]

[0091]

[0092]

[0093]

[0094] r i,p,r =B V2V log2(1+SNR i,p,r )

[0095] The constraints are:

[0096]

[0097]

[0098]

[0099] k∈{0,1,…,ω i}

[0100] Among them, G i Vehicle cluster VC i Based on the benefits obtained by the cache strategy Θ, VC i Contains cluster head CH i and ω i cluster member nodes, namely In the decision cycle t, the vehicle cluster VC i Cluster head vehicle CH i The caching decision based on the caching strategy Θ is Representatives i,k Cache content C q ;0-1 matrix Represents vehicle cluster VC i The content cache information matrix in decision cycle t, Indicates u i,k Cached content C q , Indicates u i,k Content C is not cached q ; Based on cache decision A i,t , Will enter the next state req i,r,q (t)∈{0,1},req i,r,q (t)=1 means that in the decision cycle t, u i,r Content request C was made q , req i,r,q (t)=0, otherwise; reward PR q Cache content for vehicle C q After that, each time you provide this content to other vehicles, you can get rewards, PR q By CH i According to the content C q In VC i Popularity within pl i,q and content size sc q calculate, is a constant factor; trs i,p,q for u i,p to u i,r Transmission content C q Time required, τ i,r,q for u i,r Request content C q The longest tolerable waiting time, parameter γ i,r,q Used to indicate whether the content request is delivered successfully. i,p,r <τ i,r,q That is γ i,r,q =1, u i,r Request content C q Delivery is successful, otherwise delivery fails; i,p,r Characterization u i,p with u i,r The transmission rate between V2V is the V2V channel transmission bandwidth, SNR i,p,r Providing a vehicle for content i,p Request vehicle u from content i,r Signal-to-noise ratio when transmitting content through a V2V transmission channel; i,k,t for u i,k The amount of available cache space at time t.

[0101] That is, as a specific embodiment, in each round of federated learning (FL) training, each cluster head vehicle uses its local data set and downloaded model parameters W m Its local model parameters W k Perform E iterations of updates; distribute nodes CH k At the tth iteration of the mth round of training, the updated FL parameters are:

[0102]

[0103] in, and Node CH k The model parameters updated after the t-th and t-1-th iterations of the local FL model in the local m-th round of training, η is the learning rate, For node CH k The training sample set for this iteration is is the loss function, for about The gradient, The goal of each cluster head vehicle is to find a loss function The smallest optimal parameter is:

[0104]

[0105] Then when the distribution node CH k When the mth training is completed, the updated local model parameters can be recorded as:

[0106]

[0107] in, is the mth training sample set of the distributed node. When each node completes the update of its local model, it will upload the latest to the macro base station.

[0108] S104: The macro base station aggregates all received local model parameters and updates global model parameters until the global model converges, thereby completing the training of the global model.

[0109] That is, S103 and S104 are repeated until the global model converges to complete the training of the global model.

[0110] It should be noted that the received local model parameters are aggregated according to the following formula and the global model is updated:

[0111]

[0112] Among them, nk For distribution node CH k The number of training samples, Repeat steps 2 to 5 to iteratively update the local and global FL parameters until the model converges, at which point you will get: Complete the model training.

[0113] S105 , the cluster head vehicle of each vehicle cluster obtains the trained global model so as to perform corresponding content caching according to the trained global model.

[0114] That is, each cluster head vehicle copies the trained global model and uses it to make decisions about vehicle cluster content caching. It makes more reasonable content caching decisions based on the historical content access frequency, content popularity, and available cache space of the vehicle cluster, thereby maximizing the utilization of cache space and the cache hit rate.

[0115] Finally, each cluster head vehicle will continuously generate new training samples based on the new content cache status, update the local model parameters based on the new training samples, and dynamically adjust the neural network. The training samples are all stored locally in the cluster head vehicle and do not need to be uploaded to the macro base station.

[0116] In summary, the data content caching method for vehicle clusters according to an embodiment of the present invention can, on the basis of reducing the dynamic changes in the vehicle network topology, jointly train and learn multiple vehicle clusters within the coverage area of ​​a macro base station, thereby reducing data interaction costs and obtaining a better neural network model. In this method, all training samples are stored locally on the cluster head vehicle without having to be uploaded to the macro base station. Based on the trained model, the cluster head vehicle can further realize dynamic caching of popular content in the vehicle cluster. To this end, considering the vehicle mobility and vehicle interest preference characteristics, vehicles with similar movement characteristics and similar content access preferences are divided into a cluster. While reducing the dynamic changes in the vehicle network topology, it is beneficial to improve the content cache hit rate and realize repeated access to popular content. At the same time, it reduces the communication cost during the neural network training and learning process, improves the privacy and security of the vehicle to a certain extent, and can realize dynamic caching and sharing of popular content in the Internet of Vehicles environment, which has certain practical value. It is also applicable to popular content caching scenarios in the dynamic environment of the Internet of Vehicles and has a certain degree of universality. As long as the original optimization goal can be decomposed into a 0-1 knapsack problem, it can be solved using this method.

[0117] In order to implement the above embodiment, an embodiment of the present invention proposes a computer-readable storage medium, on which a data content caching program for a vehicle cluster is stored. When the data content caching program for a vehicle cluster is executed by a processor, the data content caching method for a vehicle cluster as described above is implemented.

[0118] According to the computer-readable storage medium of an embodiment of the present invention, by storing a data content caching program for a vehicle cluster, the processor implements the above-mentioned data content caching method for a vehicle cluster when executing the data content caching program for a vehicle cluster, thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology.

[0119] To implement the above embodiment, an embodiment of the present invention proposes a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned vehicle cluster-oriented data content caching method is implemented.

[0120] According to the computer device of an embodiment of the present invention, a data content caching program for a vehicle cluster is stored in a memory, so that when the processor executes the data content caching program for a vehicle cluster, the data content caching method for a vehicle cluster as described above is implemented, thereby effectively improving the cache hit rate and cache delivery rate of the content on the basis of reducing the dynamics of the vehicle network topology.

[0121] It will be understood by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Thus, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to magnetic disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0122] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0123] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0124] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0125] It should be noted that in the claims, any reference signs placed between parentheses shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of components or steps not listed in the claim. The word "a" or "an" preceding a component does not exclude the presence of a plurality of such components. The invention can be implemented by means of hardware comprising several different components and by means of a suitably programmed computer. In a unit claim enumerating several means, several of these means may be embodied by one and the same item of hardware. The use of the words first, second, third etc. does not indicate any order. These words may be interpreted as names.

[0126] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0127] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

[0128] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood to indicate or imply relative importance or implicitly specify the number of the technical features indicated. Therefore, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, "plurality" means two or more, unless otherwise specifically defined.

[0129] In the present invention, unless otherwise expressly specified or limited, the terms "mounted," "connected," "connect," "fixed," etc. should be understood broadly. For example, they may refer to fixed connection, detachable connection, or integration; mechanical connection or electrical connection; direct connection or indirect connection through an intermediate medium; internal communication between two components or interaction between two components. Those skilled in the art will understand the specific meanings of the above terms in the present invention based on specific circumstances.

[0130] In the present invention, unless otherwise expressly specified or limited, when a first feature is "above" or "below" a second feature, it may mean that the first and second features are in direct contact, or that the first and second features are in indirect contact through an intermediary. Furthermore, when a first feature is "above," "above," or "above" a second feature, it may mean that the first feature is directly above or diagonally above the second feature, or simply means that the first feature is at a higher level than the second feature. When a first feature is "below," "below," or "below" a second feature, it may mean that the first feature is directly below or diagonally below the second feature, or simply means that the first feature is at a lower level than the second feature.

[0131] In the description of this specification, the reference terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic expressions of the above terms should not be understood as necessarily referring to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art can combine and combine different embodiments or examples described in this specification and features of different embodiments or examples without contradiction.

[0132] Although the embodiments of the present invention have been shown and described above, it will be understood that the above embodiments are illustrative and are not to be construed as limitations on the present invention. A person skilled in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A data content caching method for vehicle clusters, characterized in that: The following steps are involved: Obtaining relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles, so as to cluster all vehicles within the coverage of the macro base station according to the relative mobility and the content preference similarity, wherein each vehicle cluster includes a cluster head vehicle; The macro base station constructs a global model so that the cluster head vehicle of each vehicle cluster downloads global model parameters; The cluster head vehicle of each vehicle cluster trains the local model according to the global model parameters and the local training samples to obtain updated local model parameters, and uploads the updated local model parameters to the macro base station; The macro base station aggregates all received local model parameters and updates global model parameters until the global model converges to complete the training of the global model; The cluster head vehicle of each vehicle cluster obtains the trained global model so as to perform corresponding content caching according to the trained global model; The relative mobility and content preference similarity between each vehicle within the communication range and its neighboring vehicles are obtained, including: Obtaining a relative speed and a relative distance between each vehicle and its neighboring vehicles within a communication range, so as to obtain relative mobility based on the relative speed and relative distance; Obtaining the type and duration of historically accessed content of each vehicle within the communication range, so as to obtain the interest preference characteristics of each vehicle based on the type and duration of historically accessed content, and obtaining the content preference similarity between each vehicle and its neighboring vehicles based on the interest preference characteristics; The clustering of all vehicles within the coverage of the macro base station according to the relative mobility and the content preference similarity includes: Obtaining a cluster head factor value for each vehicle according to the relative mobility between each vehicle and its neighboring vehicles and the corresponding number of neighboring vehicles, and selecting a vehicle with the smallest cluster head factor value as a cluster head vehicle; The cluster head vehicle broadcasts the cluster head message to neighboring vehicles within the communication range, so that the vehicles that receive the cluster head message select the cluster head vehicle with the smallest cluster head factor value to send a cluster entry application; The cluster head vehicle obtains the clustering factor value of the vehicle according to the relative mobility and the content preference similarity, and clusters the vehicles in order of the clustering factor value from small to large, so as to complete vehicle clustering; The cluster head factor value of each vehicle is obtained according to the following formula: Among them, α1, α2, α2 are weighting coefficients, α1, α2, α2∈(0,1), α1+α2+α2=1; For vehicle u k The velocity vector, For vehicle u q The velocity vector, two-dimensional coordinates (x k ,y k ) represents vehicle u k The instantaneous position, two-dimensional coordinates (x q ,y q ) represents vehicle u q The instantaneous position, DNL k for u k Neighbor node list, |DNL k |=n k , RV(k) is u k The relative mobility of its direct neighbor node, RD(k) is u k The relative distance to its direct neighbor nodes; the smaller M(k), the better k The smaller the relative speed and relative distance to its direct neighbor nodes and the more direct neighbor nodes it has, the higher the u k The more likely it is to become a cluster head.

2. The data content caching method for vehicle clusters according to claim 1, characterized in that: Obtain the interest preference characteristics of each vehicle according to the following formula: Among them, I k For vehicle u k A preference vector for all content types, for u k Preference for content of type m, C j For any popular content contained in the macro base station content library C, the shortest time required to access the content is CT j , C can be divided into N a categories, the content class set can be expressed as cha m is the content set of the mth type, PC k For vehicle u k A collection of accessed content in the past period of time. When C j ∈PC k , For vehicle u k Content C j The actual visit duration.

3. The data content caching method for vehicle clusters according to claim 2, characterized in that: The content preference similarity between each vehicle and its neighboring vehicles is obtained according to the following formula: Among them, I k and I o Represents vehicle node u k and vehicle node u o Preference vector for all content types, sim k,o A larger value indicates that the content access preferences of the two nodes are more similar.

4. The data content caching method for vehicle clusters according to claim 3, characterized in that: The clustering factor value is obtained according to the following formula: Among them, λ1, λ2∈(0,1), λ1+λ2=1, λ1 and λ2 represent the influence of mobility and interest preference on the clustering factor respectively; Vehicle cluster VC o average speed; For VC o The average preference of all nodes in the m-th category for content.

5. The data content caching method for vehicle clusters according to claim 4, characterized in that: The local training samples include historical content request data, vehicle cache space, and the corresponding optimal vehicle cluster content cache strategy, wherein the optimal vehicle cluster content cache strategy is obtained by solving the following problem of maximizing the intra-cluster vehicle cache benefit: r i,p,r =B V2V log2(1+SNR i,p,r ) The constraints are: k∈{0,1,…,ω i } Among them, G i Vehicle cluster VC i Based on the benefits obtained by the cache strategy Θ, VC i Contains cluster head CH i and ω i cluster member nodes, namely In the decision cycle t, the vehicle cluster VC i Cluster head vehicle CH i The caching decision based on the caching strategy Θ is Representatives i,k Cache content C q ;0-1 matrix Represents vehicle cluster VC i The content cache information matrix in decision cycle t, Indicates u i,k Cached content C q , Indicates u i,k Content C is not cached q ; Based on cache decision A i,t , Will enter the next state req i,r,q (t)∈{0,1},req i,r,q (t)=1 means that in the decision cycle t, u i,r Content request C was made q , req i,r,q (t)=0, otherwise; reward PR q Cache content for vehicle C q After that, each time you provide this content to other vehicles, you can get rewards, PR q By CH i According to the content C q In VC i Popularity within pl i,q and content size sc q calculate, is a constant factor; trs i,p,r for u i,p to u i,r Transmission content C q Time required, τ i,r,q for u i,r Request content C q The longest tolerable waiting time, parameter γ i,r,q Used to indicate whether the content request is delivered successfully. i,p,r <τ i,r,q That is γ i,r,q =1, u i,r Request content C q Delivery is successful, otherwise delivery fails; i,p,r Characterization u i,p with u i,r The transmission rate between V2V is the V2V channel transmission bandwidth, SNR i,p,r Providing a vehicle for content i,p Request vehicle u from content i,r Signal-to-noise ratio when transmitting content through a V2V transmission channel; i,k,t for u i,k The amount of available cache space at time t.

6. A computer-readable storage medium, characterized in that A data content caching program for vehicle clusters is stored thereon, and when the data content caching program for vehicle clusters is executed by a processor, a data content caching method for vehicle clusters according to any one of claims 1 to 5 is implemented.

7. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the data content caching method for vehicle clusters according to any one of claims 1 to 5 is implemented.

Images