Cooperative caching scheme for high-security private network MIN-V2X

Abstract

The invention discloses a cooperative caching scheme for a high-security private network MIN-V2X, and relates to the field of information center networks and Internet of Vehicles. Comprising the steps that in a designed multi-identification network compatible with TCP / IP and ICN, nodes of the multi-identification network are divided into a management layer and a cache layer, the management layer is composed of road side units, and the cache layer is composed of vehicles; the vehicles are used as nodes of a cache layer to be clustered through a weight-based clustering algorithm; a cooperative cache search algorithm is put forward, cooperative search between clusters can prevent a large number of redundant data packets from being transmitted in the network, network transmission consumption is reduced, the characteristics of the MIN and the Internet of Vehicles are combined, a comprehensive access prefetching strategy, a vehicle node access-based prefetching strategy and an identification information-based prefetching strategy are put forward, and through a cache replacement strategy, the network access efficiency is improved. The efficiency is improved and the resource consumption is reduced. According to hot content and habits and preferences of vehicle nodes, the content is cached in advance according to the clustering characteristics of the MIN-identifier, the access delay is reduced, and the user experience is improved.

Description

Technical Field

[0001] The present invention relates to the field of vehicle networking information caching technology, and specifically to a collaborative caching solution for a high-security private network MIN-V2X. Background Art

[0002] The early development of the Internet of Vehicles began in 1962, when General Motors designed an inertial guidance and navigation system for the Apollo program, which laid the foundation for the subsequent advancement of vehicle navigation and communication technology. The Internet of Vehicles has also developed rapidly in China, going through the 2G Internet of Vehicles stage (2009-2013), the 4G Internet of Vehicles stage (2014-2019) and the current 5G Internet of Vehicles stage (2020-2025). Since the development of the Internet of Vehicles originated from mobile wireless self-organizing networks, all of which are based on TCP / IP, there are the following problems:

[0003] First, IP addresses are limited. Faced with the growing number of vehicle nodes, IP addresses will eventually be exhausted. Despite the existence of network address translation (NAT), dynamic allocation of IP addresses, and IPv6, it is complex to allocate IP addresses to various devices and nodes in the Internet of Vehicles and to implement routing aggregation.

[0004] Secondly, it is the ossification of the TCP / IP system. The TCP / IP system is an end-to-end network architecture, but it is built on a relatively stable network topology. When the network topology changes due to vehicle movement and network instability, the TCP / IP network structure and routing links are often greatly affected, resulting in packet loss, network delay fluctuations and other phenomena. In addition, this end-to-end transmission is all returned by the data source. Due to changes in the network topology, the transmission path often cannot be reused, which will bring heavy network traffic overhead and content retrieval delays.

[0005] Thirdly, the existing network information forwarded to the IP layer based on IP address routing is too stressful, and the upper layer message content information is not perceived. The information distribution is not accurate, and the content information cannot be what you see is what you get.

[0006] Finally, there is the security of the TCP / IP system. On the one hand, most nodes use unlimited shared media, which makes them vulnerable to DoS (Denial of service) attacks. On the other hand, the Internet of Vehicles has a large and complex network structure, which makes it vulnerable to network traffic injection from fake nodes.

[0007] In order to solve the above challenges, the future network will develop towards Information-Centric Networking (ICN), and Named Data Network (NDN) is one of the most popular ICN models. Although ICN has opened up a new path for the development of future networks, this design, which is very different from the existing TCP / IP architecture, means a high migration cost. Under the premise of reducing costs as much as possible, a network architecture solution that is compatible with IP and ICN systems has become the primary goal of related research. To achieve this goal, researchers proposed the Multi-Identifier Network (MIN) solution.

[0008] Although MIN's identity module has certain advantages over other ICN solutions, directly labeling the identity and using MIN's default cache algorithm will result in high costs and waste of resources. In order to improve cache efficiency and save vehicle node resources, a node management and cache management solution based on MIN in the Internet of Vehicles is urgently needed.

[0009] In the prior art, publication number CN111556531B discloses a collaborative cache optimization method in a microcellular wireless network, which models the microcellular wireless network as a two-dimensional directed topology graph, characterizes the collaborative transmission relationship between small base stations by designing a collaborative transmission matrix, and constructs an average cache hit rate expression from the collaborative transmission matrix and the file request probability matrix, and solves the collaborative cache solution that optimizes the average cache hit rate by a convex optimization method. However, this method may be more complicated to implement using a convex optimization method, especially in large-scale networks, where the computational cost of solving the optimal solution may be very high, resulting in limited real-time performance. In addition, the construction of the file request probability matrix depends on accurate modeling of user request behavior. If the user's request behavior changes or is difficult to predict, the cache strategy may no longer be effective, thereby reducing the cache hit rate.

[0010] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0011] The purpose of the present invention is to provide a collaborative caching solution for a high-security private network MIN-V2X to solve the problems raised in the above background technology.

[0012] To achieve the above object, the present invention provides the following technical solutions:

[0013] A collaborative caching solution for a high-security private network MIN-V2X, the specific steps of which include:

[0014] A hierarchical clustering model of network nodes is established, and nodes in the network are divided into two parts: a management layer and a cache layer. The management layer is composed of a number of roadside units, and the cache layer is composed of a number of vehicle nodes. The cache layer is clustered to form a number of clusters, and each cluster includes a cluster head and a number of cluster members.

[0015] Based on the network node state parameters, a cluster head in each cluster is selected by a weight-based clustering algorithm, specifically, each vehicle node is scored, and a certain vehicle node is selected as the cluster head of each cluster according to the score of the vehicle node. The network node state parameters include the average transmission time from the vehicle node to all adjacent vehicle nodes within the range, the shortest distance from the vehicle node to the management layer node, the remaining energy of the vehicle node, and the remaining storage space of the vehicle node;

[0016] Based on the selected cluster head, among the remaining vehicle nodes in the cache layer, a judgment condition is generated according to the longest prefix matching algorithm, the upper limit of the current cluster size and the selected state of the vehicle node, and the vehicle node that meets the judgment condition is selected as the cluster member of the cluster head;

[0017] Based on the pre-fetching strategy of comprehensive access, the pre-fetching strategy of vehicle node access and the pre-fetching strategy of identification information, cache pre-fetching is achieved through the combination of three strategies. At the same time, in view of the characteristics of large traffic and strong mobility in the vehicle environment, a multi-level probabilistic cache strategy is set to complete cache replacement;

[0018] The network is layered based on the network node hierarchical clustering model, and the nodes are clustered to achieve cache prefetching and cache replacement. The collaborative cache search algorithm is used to find the node that can respond to the request within the cluster or adjacent clusters and distribute the node content. If it cannot be found quickly from the cache layer, the content is obtained directly from the cloud server.

[0019] Furthermore, the nodes in the network are divided into two parts: the management layer and the cache layer. The management layer is responsible for managing and recording metadata, coordinating data transmission between the server and the cache layer, implementing content pre-fetching strategies, and scheduling the forwarding of cache requests. The cache block metadata includes the location of the cluster head in the network, MIN identifier, size, and storage time;

[0020] The cache layer is composed of vehicle nodes, which are responsible for actually storing the cache blocks within the cluster. The cache layer forms multiple clusters through a weight-based clustering algorithm (WCA). Each cluster contains a cluster head and several cluster members. The cluster head not only stores the cache blocks within the cluster, but also manages the cache block metadata information.

[0021] Furthermore, based on the network node state parameters, the cluster head in each cluster is selected by a weight-based clustering algorithm. The logic for scoring each vehicle node is as follows: the average transmission time from the calibrated vehicle node to all adjacent vehicle nodes within the range is , the shortest distance from the vehicle node to the management node is , the remaining energy of the vehicle node is and the remaining storage space of the vehicle node is , the formula for calculating the score of each vehicle node is:

[0022]

[0023] In the formula represents the vehicle node score, , , and The average transmission time , closest distance , the remaining energy of the vehicle node and the remaining storage space of the vehicle node The weight coefficient of and , , and All are greater than 0;

[0024] According to the vehicle node score, a certain vehicle node is selected as the cluster head of each cluster. The logic is as follows: first, the remaining energy exceeds the predetermined limit. The vehicle nodes are used to form a cluster head pre-selection set. In the cluster head pre-selection set, the vehicle node with the highest vehicle node score is selected as the cluster head according to the score of each vehicle node.

[0025] Furthermore, based on the selected cluster heads, the logic for selecting cluster members in the remaining vehicle nodes in the cache layer is as follows: after the cluster heads are selected, each cluster head is responsible for allocating the remaining vehicle nodes to their respective clusters, setting the upper limit of the number of nodes in each cluster to , and at the same time determine the selected state of the vehicle node, the selected state of the vehicle node includes not being included in any cluster and being included in the cluster, for any neighbor node m within the transmission range of the cluster head, and the neighbor node has not joined any cluster before, and the current cluster size is less than the upper limit of the number of nodes , and the result of the longest prefix matching algorithm between node m and the cluster head node cannot be an empty set. If the node meets the above conditions, it will be selected as a member of the cluster by the cluster head node;

[0026] The specific logic of the longest prefix matching algorithm of node m and cluster head node is: all network nodes including node m and cluster head node need to support MIN-V2X multi-identification protocol, and all message formats contain MIN-V2X prefix information. When matching, starting from the starting bit of the prefix information, the prefix of the identification information of the cluster head node and node m is compared bit by bit. If no matching prefix is ​​found, the default value 0 is returned, that is, the result of the longest prefix matching algorithm of node m and cluster head node is an empty set, otherwise the longest prefix matching length result is returned.

[0027] Furthermore, based on the characteristics of MIN-V2X and the Internet of Vehicles, cache pre-fetching is achieved through a combination of three strategies: pre-fetching strategy based on comprehensive access, pre-fetching strategy based on vehicle node access, and pre-fetching strategy based on identification information.

[0028] The comprehensive access-based pre-fetching strategy uses comprehensive access data to predict future data request patterns. This strategy identifies potential popular content and caches it in advance by analyzing access data from the entire network. The comprehensive access data includes historical access records of users, access popularity during a certain period of time, and geographic location factors.

[0029] The vehicle node access-based pre-fetching strategy uses the access content of a single vehicle node to predict future data requests, and monitors and records the access behavior of individual vehicle nodes to identify their habits and preferences, thereby performing personalized cache pre-fetching;

[0030] The pre-fetching strategy based on identification information relies on the specific identification information of the user or content to predict and provide identified data services. The specific identification information includes the user's ID, the unique identifier of the content and the MIN-identifier. The MIN-identifier is the identity of the vehicle node in the multi-identification network system MIN. By analyzing these identification information, the access needs of specific users or specific content are predicted, and related data are cached in advance.

[0031] Furthermore, in view of the characteristics of large traffic volume and high mobility in the vehicle environment, a multi-level probabilistic caching strategy is set to complete the cache replacement, wherein the process of replacing the cache node includes: when a cache block is newly added to the storage space of the node, it is placed in a new area. If the space in the new area is full at this time, the longest prefix match shortest threshold is set, and the cache block below the longest prefix match shortest threshold is selected to form a first pre-elimination set. In the first pre-elimination set, the cache block that has not been accessed for the longest time is selected and eliminated. When the newly added cache block is accessed again, the cache block is migrated to the old area, and the position of the cache block is adjusted to the head of the queue. If the space in the old area is full at this time, the same method as the elimination in the new area is adopted to select the cache block below the longest prefix match shortest threshold to form a second pre-elimination set, filter out the cache block that has not been accessed for the longest time in the second pre-elimination set, and move the selected cache block to the new area. If the new area is also full at this time, the cache block with the shortest matching prefix in the first pre-elimination set and closest to the end of the queue is selected again in the new area and eliminated.

[0032] Furthermore, by using the collaborative cache search algorithm, when there is a user request in the cluster, the logic for quickly locating the node in the cluster that responds to the request is as follows: in the collaborative cache design of MIN-V2X, the cluster head manages the data cache blocks cached in each of its cluster members, and the members of the cluster cache no redundant data. The communication within the entire cluster is only carried out through the cluster head. When a cluster member receives any content cache block request, it first uses the full matching algorithm to check whether its own node has cached the content cache block request. If not found, the longest prefix matching algorithm is used to find the metadata of the corresponding node in the cluster according to the MIN-identifier of the target data, and the corresponding node in the cluster is found according to the metadata and the full matching algorithm is used again to find the cache block content and return the data packet; if there is no match in the request of this cluster head, the cluster head forwards the request to the roadside unit, and the roadside unit determines to send the request to a neighboring cluster head, where the basis for determining the target neighboring cluster head is still the longest prefix matching algorithm to match the cluster head with the most relevant MIN-identifier through the MIN-identifier, and the same search process is followed after forwarding to the neighboring cluster head. If the roadside unit has not yet matched, the request is forwarded to the cloud server through the callback function to obtain the content from the cloud server.

[0033] Each content cache block request determines a limited number of hops. Each time a content cache block request is forwarded, the limited number of hops decreases by 1. Only when the limited number of hops is greater than 0, a collaborative cache search is performed. Otherwise, if the search result is considered to have failed, a negative acknowledgment (NACK) is used to notify the initial requester, and the request is forwarded to the roadside unit, which then forwards the request to the cloud server through a callback function to obtain the content from the cloud server.

[0034] Furthermore, after executing the collaborative cache search algorithm, if the cluster can be successfully matched, the data will be returned; if the match fails, the request will be forwarded to the roadside unit, which will locate the cluster and then forward it to the corresponding cluster member through the neighbor cluster head. This cycle will continue until the limited number of hops drops to 0, and then the node that responds to the request will be found, and finally the node will distribute the content. If the node that responds to the request cannot be found in the cache layer, the requested content will be obtained from the cloud service.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] Based on the network layering and node clustering, a collaborative cache search algorithm is proposed. This algorithm makes excellent use of the clustering characteristics and gives full play to the efficiency of collaborative search. In this algorithm, the locally uncached request content will not be broadcast to the entire network. Instead, the request will be sent to the cluster head node. The management function of the cluster head node is used to quickly locate the node in the cluster that can respond to the request, or the request will be sent to the roadside unit, which will locate the cluster and then find the target node, and finally the node will distribute the content. It effectively reduces the network load caused by broadcasting in highly dense vehicle node scenarios, and the collaborative search between clusters can avoid the transmission of a large number of redundant data packets in the network, reducing the transmission consumption in the network;

[0037] Combining the characteristics of MIN and IoV, a comprehensive cache pre-fetching strategy is proposed, which takes into account the popular content of the entire network, the access preference of vehicle nodes and the clustering characteristics based on MIN-identifier. It can pre-cache the content to be accessed to a large extent, thereby improving the user experience.

[0038] Improvements have been made to the ordinary LRU to suit MIN-V2X. The cache space is partitioned into new and old partitions, and the new area content is given priority during replacement. On the other hand, the cache content is probabilistically scored based on the MIN identifier and popularity. Only content that reaches a certain score can be stored, which largely retains the hot content that users frequently visit. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0040] Figure 2 The flow chart of the cluster head selection algorithm of the present invention;

[0041] Figure 3 A flow chart of the cluster member selection algorithm of the present invention;

[0042] Figure 4 This is a schematic diagram of the cache replacement process of the present invention;

[0043] Figure 5This is a flow chart of the collaborative cache search algorithm of the present invention. DETAILED DESCRIPTION

[0044] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0045] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0046] Example:

[0047] See also Figure 1-5 , the present invention provides a technical solution:

[0048] A collaborative caching solution for a high-security private network MIN-V2X, the specific steps of which include:

[0049] Step 1: Establish a hierarchical clustering model of network nodes, divide the nodes in the network into two parts: management layer and cache layer. The management layer is composed of several roadside units, and the cache layer is composed of several vehicle nodes. The cache layer is clustered to form several clusters, each cluster contains a cluster head and several cluster members;

[0050] The nodes in the network are divided into two parts: the management layer and the cache layer. The management layer is responsible for recording the cache block metadata and managing the data transmission and content pre-fetching from the server to the cache layer. The cache block data includes the location of the cluster head in the network, MIN identifier, size and storage time;

[0051] However, the management layer does not directly participate in the cache operations within the cluster. Specifically, the main responsibilities of the management layer include the following aspects:

[0052] Cluster head recording and management: The management layer records the location, MIN identifier, size, storage time and other information of the cluster head in the network. These metadata are crucial for the management and scheduling of cache resources and can help quickly locate and retrieve cache content;

[0053] Data transfer management: The management layer is responsible for coordinating data transfer between the server and the cache layer. This includes the implementation of content pre-fetching strategies to ensure that the vehicle nodes in the cache layer can obtain the required data in a timely manner;

[0054] Resource scheduling: By scheduling the forwarding of cache requests at the management level, the utilization of network resources can be optimized, the redundancy and delay of data transmission can be reduced, and the operating efficiency of the entire network can be improved.

[0055] The cache layer is composed of vehicle nodes, which are responsible for actually storing cache blocks in the cluster. The cache layer forms multiple clusters through clustering. Each cluster contains a cluster head and several cluster members. The cluster head not only stores the cache blocks in the cluster, but also manages the cache block metadata information. Specifically, the responsibilities of the cache layer include:

[0056] Actual storage: The vehicle nodes in the cache layer store the actual data blocks. These data blocks may include videos, map data, software updates, etc.

[0057] Cluster structure management: Through clustering algorithms, vehicle nodes are divided into multiple clusters. Each cluster has a cluster head, which is responsible for managing the data transmission and cache block metadata within the cluster. These metadata information include but are not limited to the cache block identification, size, storage time, etc. This structure helps to reduce data transmission delays and improve access speed;

[0058] Cache resource sharing: Nodes within the cluster can share cache resources and achieve efficient data access and transmission through the coordination of the cluster head. This sharing mechanism can significantly reduce the frequency of obtaining data from the server and reduce network load.

[0059] The vehicle nodes are divided into a cache layer, while the roadside unit is in the management layer. The callback server and content server are cloud devices, and the vehicle nodes in the cache layer are divided into clusters. The cluster head in each cluster manages and coordinates the use of cache resources.

[0060] Step 2: Based on the network node state parameters, the cluster head in each cluster is selected through a weight-based clustering algorithm. Specifically, each vehicle node is scored and a certain vehicle node is selected as the cluster head of each cluster according to the score of the vehicle node. The network node state parameters include the average transmission time from the vehicle node to all adjacent vehicle nodes within the range, the shortest distance from the vehicle node to the management node, the remaining energy of the vehicle node, and the remaining storage space of the vehicle node;

[0061] Based on the network node state parameters, the cluster head in each cluster is selected through a weight-based clustering algorithm. The logic for scoring each vehicle node is as follows: the average transmission time from the calibrated vehicle node to all adjacent vehicle nodes within the range is , the shortest distance from the vehicle node to the management node is , the remaining energy of the vehicle node is and the remaining storage space of the vehicle node is , the formula for calculating the score of each vehicle node is:

[0062]

[0063] In the formula represents the vehicle node score, , , and The average transmission time , closest distance , the remaining energy of the vehicle node and the remaining storage space of the vehicle node The weight coefficient of and , , and All are greater than 0;

[0064] Since the smaller the transmission time, the better, its reciprocal is used for scoring. The shorter the transmission time, the higher the score. Similar to the transmission time, the shorter the distance to the management layer node, the better, and the reciprocal is used for scoring. The higher the remaining energy, the better, so it is directly used. Score; the remaining storage space should be as high as possible, so use it directly At the same time, in order to reduce the average transmission time, the shortest distance and the remaining energy and storage space of the vehicle node are small, so the setting .

[0065] According to the vehicle node score, a certain vehicle node is selected as the cluster head of each cluster. The logic is as follows: first, the remaining energy exceeds the predetermined limit. The vehicle nodes are used to form a cluster head pre-selection set. In the cluster head pre-selection set, the vehicle node with the highest vehicle node score is selected as the cluster head according to the score of each vehicle node.

[0066] Step 3: Based on the selected cluster head, among the remaining vehicle nodes in the cache layer, a judgment condition is generated according to the longest prefix matching algorithm, the upper limit of the current cluster size and the selected state of the vehicle node, and the vehicle node that meets the judgment condition is selected as the cluster member of the cluster head;

[0067] Based on the selected cluster heads, the logic for selecting cluster members in the remaining vehicle nodes in the cache layer is as follows: after the cluster heads are selected, each cluster head is responsible for allocating the remaining vehicle nodes to their respective clusters, setting the upper limit of the number of nodes in each cluster to , and at the same time determine the selected state of the vehicle node, the selected state of the vehicle node includes not being included in any cluster and being included in the cluster, for any neighbor node m within the transmission range of the cluster head, and the neighbor node has not joined any cluster before, and the current cluster size is less than the upper limit of the number of nodes , and the result of the longest prefix matching algorithm between node m and the cluster head node cannot be an empty set. If the node meets the above conditions, it will be selected as a member of the cluster by the cluster head node;

[0068] The specific logic of the longest prefix matching algorithm of node m and cluster head node is: all network nodes including node m and cluster head node need to support MIN-V2X multi-identification protocol, and all message formats contain MIN-V2X prefix information. When matching, starting from the starting bit of the prefix information, the prefix of the identification information of the cluster head node and node m is compared bit by bit. If no matching prefix is ​​found, the default value 0 is returned, that is, the result of the longest prefix matching algorithm of node m and cluster head node is an empty set, otherwise the longest prefix matching length result is returned.

[0069] Newly added MIN-V2X nodes and free nodes will also be divided into appropriate clusters by the algorithm. After selecting the cluster head, cluster members will be selected from the remaining nodes in the storage layer to further divide the clusters. In order to avoid a large number of sparse node clusters in the cache layer, the sparse node clusters have no cluster members or only a few cluster members. After dividing a cluster, it is necessary to determine whether its size has reached the minimum standard. If not, this cluster will be added to Middle, Last That is, the cluster is formed by merging multiple sparse node clusters, and a cluster head node is selected according to the weight-based clustering algorithm. , this cluster head node will be added to the cluster head list, but the set of cluster heads and cluster members is not static. Due to the instability of the vehicle environment, some vehicle nodes will leave their clusters after failing to obtain the requested content for many times, becoming free nodes and waiting to be re-divided. Similarly, when the number of nodes in a cluster is lower than the threshold, we can assume that the cluster head node no longer has the ability to manage the cluster, the cluster will be disbanded, all nodes will become free nodes, and the algorithm will be re-executed on them.

[0070] Step 4: Based on the prefetching strategy of comprehensive access, the prefetching strategy of vehicle node access and the prefetching strategy of identification information, cache prefetching is achieved through a combination of the three strategies. At the same time, in view of the characteristics of large traffic and strong mobility in the vehicle environment, a multi-level probabilistic cache strategy is set to complete cache replacement.

[0071] The pre-fetching strategy based on comprehensive access uses comprehensive access data to predict future data request patterns. This strategy identifies potential hot content by analyzing access data in the entire network and caches it in advance. The comprehensive access data includes the user's historical access records, access popularity in time periods, and geographical location factors. This strategy identifies potential hot content by analyzing access data in the entire network and caches it in advance. In this way, the network can respond quickly when demand is high, reducing delays and network congestion. Comprehensive access data includes factors such as the user's historical access records, access popularity in time periods, and geographical location. Through big data analysis, this strategy can accurately predict future access needs, thereby effectively performing cache pre-fetching.

[0072] The pre-fetching strategy based on vehicle node access uses the access content of a single vehicle node to predict future data requests, and identifies its habits and preferences by monitoring and recording the access behavior of individual vehicle nodes, so as to perform personalized cache pre-fetching; This strategy monitors and records the access behavior of individual vehicle nodes, identifies its habits and preferences, and performs personalized cache pre-fetching. For example, if a vehicle frequently accesses certain types of content, the strategy will cache these contents in advance to ensure that the vehicle node can quickly access the required data. A Markov chain is also constructed for the vehicle node to predict the user's access trend, and to perform ranking and pre-fetching. This method not only optimizes the allocation and use of cache resources, but also improves the experience of individual users.

[0073] The identification information-based pre-fetching strategy relies on the specific identification information of the user or content to predict and provide the identified data service. The specific identification information includes the user's ID, the unique identifier of the content and the MIN-identifier. The MIN-identifier is the identity of the vehicle node in the multi-identification network system MIN. By analyzing these identification information, the access needs of specific users or specific content are predicted, and the relevant data is cached in advance. This strategy can predict the access needs of specific users or specific content and cache the relevant data in advance. For example, we will consider some data content with the same identification as the node as more needed by the user, so as to cache these contents in advance to reduce access delays.

[0074] It is worth noting that these cache prefetching strategies are not limited to specific types of network nodes. They can be widely used in a variety of network environments such as vehicle nodes, roadside units (RSUs), and edge computing nodes. Each strategy plays a specific role in different node environments: in vehicle nodes, these strategies provide customized cache services by analyzing the vehicle's MIN-identifier, driving path, and access records to improve data access speed and user experience; in roadside units (RSUs), these strategies use a large amount of traffic data and comprehensive access records to optimize traffic flow management and content distribution efficiency; in edge computing nodes, these strategies use the powerful processing power of edge computing resources to perform large-scale data analysis and prediction, and support efficient cache prefetching and data transmission. Furthermore, the information exchange and collaborative work between these strategies can significantly improve the prediction accuracy and efficiency of the entire network system. For example, the comprehensive access-based strategy can provide macro access trend data for the vehicle node access-based strategy, while the vehicle node access-based prefetching strategy can provide accurate user preference data, and the identification information-based strategy can provide highly relevant content and content that is private through the MIN-identifier. Through the coordination and information sharing of multiple strategies, the network system can achieve more accurate and efficient cache prefetching, reduce redundant data transmission, and optimize the overall utilization of network resources.

[0075] In view of the characteristics of large traffic and high mobility in the vehicle environment, a multi-level probabilistic caching strategy is set to complete the cache replacement, wherein the cache node replacement process includes: when a cache block is newly added to the storage space of the node, it is placed in a new area. If the space in the new area is full at this time, the longest prefix match shortest threshold is set, and the cache blocks below the longest prefix match shortest threshold are selected to form a first pre-elimination set. In the first pre-elimination set, the cache blocks that have not been accessed for the longest time are selected and eliminated. When the newly added cache block is accessed again, the cache block is migrated to the old area, and the position of the cache block is adjusted to the head of the queue. If the space in the old area is full at this time, the same method as the elimination in the new area is adopted to select the cache blocks below the longest prefix match shortest threshold to form a second pre-elimination set, and the cache blocks that have not been accessed for the longest time in the second pre-elimination set are screened out, and the selected cache blocks are moved to the new area. If the new area is also full at this time, the cache block with the shortest matching prefix in the first pre-elimination set and closest to the end of the queue is selected again in the new area and eliminated.

[0076] Step 5: Based on the network node hierarchical clustering model, the network is layered and the nodes are clustered to implement cache pre-fetching and cache replacement. The collaborative cache search algorithm is used to find the node that can respond to the request within the cluster or adjacent clusters and distribute the node content. If it cannot be found quickly from the cache layer, the content is obtained directly from the cloud server.

[0077] Using the collaborative cache search algorithm, when there is a user request in the cluster, the logic of quickly locating the node in the cluster that responds to the request is as follows: In the collaborative cache design of MIN-V2X, the cluster head manages the data cache blocks cached in each of its cluster members. The members of the cluster cache no redundant data. The communication within the entire cluster is only carried out through the cluster head. When a cluster member receives any content cache block request, it first checks whether its own node has cached the content cache block request through the full matching algorithm. If it is not found, it will request the cluster head of the cluster. Due to the limited cache space of the cluster head, it is impossible to store the cache data of all its members. The cluster head only caches the metadata of the cached data of its cluster members and the corresponding MIN-identifier. Use the longest prefix matching algorithm to find the metadata of the corresponding cluster node according to the MIN-identifier of the target data. If the corresponding cluster node is found according to the metadata, the full matching algorithm is used again to find the cache block content and return the data packet; if there is no match in the request of this cluster head, the cluster head will forward the request to the road side unit, and the road side unit will determine to send the request to a neighbor cluster head. The basis for determining the target neighbor cluster head is still the longest prefix matching algorithm to match the cluster head with the most relevant MIN-identifier through the MIN-identifier. After forwarding to the neighbor cluster head, the same search process is followed. If there is no match through the road side unit, the request is forwarded to the cloud server through the callback function to obtain the content from the cloud server.

[0078] Each content cache block request determines a limited number of hops. Each time the content cache block request is forwarded, the limited number of hops decreases by 1. Only when the limited number of hops is greater than 0, the collaborative cache search is performed. Otherwise, if the search result is considered to have failed, a negative acknowledgment (NACK) is used to notify the initial requester, and the request is forwarded to the roadside unit, and then the request is forwarded to the cloud server through a callback function to obtain the content from the cloud server.

[0079] The specific steps of the complete match algorithm include: determining the object to be completely matched, such as two strings, database records or data structures, ensuring that the data to be matched are comparable, using appropriate comparison methods to determine whether the objects are completely matched; performing subsequent operations based on the comparison results, such as returning matching records. The complete match algorithm is a mature existing technology and will not be described in detail here.

[0080] After executing the collaborative cache search algorithm, if a match is successful, the data will be returned or forwarded to the cluster member corresponding to the cluster where the neighbor cluster head is located. If the match fails, the request will be forwarded to the roadside unit, which will locate the cluster. This cycle will be repeated until the limited number of hops drops to 0, and then the node that responds to the request is found, and finally the node will distribute the content.

[0081] In this process, the request is not broadcast to all neighboring cluster heads, but the MIN-identifier of the cluster stored in the roadside unit is used for routing guidance, which reduces the traffic load in the entire network. On the other hand, for content matching in the scene environment of the cluster head, the corresponding member MIN-identifier is used to respond to the request locally within the cluster. This reduces the cost of transmitting the request to all cluster members or neighboring clusters, thereby reducing the overall cost.

[0082] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0083] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0084] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0085] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A collaborative caching solution for a highly secure private network MIN-V2X, characterized in that: The specific steps include: A hierarchical clustering model of network nodes is established, and nodes in the network are divided into two parts: a management layer and a cache layer. The management layer is composed of a number of roadside units, and the cache layer is composed of a number of vehicle nodes. The cache layer is clustered to form a number of clusters, and each cluster includes a cluster head and a number of cluster members. Based on the network node state parameters, a cluster head in each cluster is selected by a weight-based clustering algorithm, specifically, each vehicle node is scored, and a certain vehicle node is selected as the cluster head of each cluster according to the score of the vehicle node. The network node state parameters include the average transmission time from the vehicle node to all adjacent vehicle nodes within the range, the shortest distance from the vehicle node to the management layer node, the remaining energy of the vehicle node, and the remaining storage space of the vehicle node; Based on the selected cluster head, among the remaining vehicle nodes in the cache layer, a judgment condition is generated according to the longest prefix matching algorithm, the upper limit of the current cluster size and the selected state of the vehicle node, and the vehicle node that meets the judgment condition is selected as the cluster member of the cluster head; Based on the pre-fetching strategy of comprehensive access, the pre-fetching strategy of vehicle node access and the pre-fetching strategy of identification information, cache pre-fetching is achieved through the combination of three strategies. At the same time, in view of the characteristics of large traffic and strong mobility in the vehicle environment, a multi-level probabilistic cache strategy is set to complete cache replacement; The network is layered based on the network node hierarchical clustering model, and the nodes are clustered to achieve cache prefetching and cache replacement. The collaborative cache search algorithm is used to find the node that can respond to the request within the cluster or adjacent clusters and distribute the node content. If it cannot be found quickly from the cache layer, the content is obtained directly from the cloud server.

2. According to claim 1, a collaborative caching solution for a high-security private network MIN-V2X is characterized in that: The nodes in the network are divided into two parts: the management layer and the cache layer. The management layer is responsible for managing and recording metadata, coordinating data transmission between the server and the cache layer, implementing content pre-fetching strategies, and scheduling the forwarding of cache requests. The metadata includes the location of the cluster head in the network, MIN identification, size, and storage time; The cache layer is composed of vehicle nodes, which are responsible for actually storing the cache blocks within the cluster. The cache layer forms multiple clusters through a weight-based clustering algorithm. Each cluster contains a cluster head and several cluster members. The cluster head not only stores the cache blocks within the cluster, but also manages the cache block metadata information.

3. According to claim 2, a collaborative caching solution for a high-security private network MIN-V2X is characterized in that: Based on the network node state parameters, the cluster head in each cluster is selected through a weight-based clustering algorithm. The logic for scoring each vehicle node is as follows: the average transmission time from the calibrated vehicle node to all adjacent vehicle nodes within the range is , the shortest distance from the vehicle node to the management node is , the remaining energy of the vehicle node is and the remaining storage space of the vehicle node is , the formula for calculating the score of each vehicle node is: In the formula represents the vehicle node score, , , and The average transmission time , closest distance , the remaining energy of the vehicle node and the remaining storage space of the vehicle node The weight coefficient of and , , and All are greater than 0; According to the vehicle node score, a certain vehicle node is selected as the cluster head of each cluster. The logic is as follows: first, the remaining energy exceeds the predetermined limit. The vehicle nodes are used to form a cluster head pre-selection set. In the cluster head pre-selection set, the vehicle node with the highest vehicle node score is selected as the cluster head according to the score of each vehicle node.

4. According to claim 1, a collaborative caching solution for a high-security private network MIN-V2X is characterized in that: Based on the selected cluster heads, the logic for selecting cluster members in the remaining vehicle nodes in the cache layer is as follows: after the cluster heads are selected, each cluster head is responsible for allocating the remaining vehicle nodes to their respective clusters, setting the upper limit of the number of nodes in each cluster to , and at the same time determine the selected state of the vehicle node, the selected state of the vehicle node includes not being included in any cluster and being included in the cluster. For any neighbor node m within the transmission range of the cluster head, the neighbor node has not joined any cluster before, and the size of the current cluster is less than the upper limit of the number of nodes , and the result of the longest prefix matching algorithm between node m and the cluster head node cannot be an empty set. If the node meets the above conditions, it will be selected as a member of the cluster by the cluster head node; The specific logic of the longest prefix matching algorithm of node m and cluster head node is as follows: all network nodes including node m and cluster head node need to support the MIN-V2X multi-identification protocol, and all message formats contain MIN-V2X prefix information. When matching, start from the starting bit of the prefix information and compare the prefix of the identification information of the cluster head node and node m bit by bit. If no matching prefix is ​​found, the default value 0 is returned, that is, the result of the longest prefix matching algorithm of node m and cluster head node is an empty set, otherwise the longest prefix matching length result is returned.

5. According to claim 1, a collaborative caching solution for a high-security private network MIN-V2X is characterized in that: Based on the characteristics of MIN-V2X and the Internet of Vehicles, cache pre-fetching is achieved through a combination of three strategies: comprehensive access pre-fetching strategy, vehicle node access pre-fetching strategy, and identification information pre-fetching strategy. The comprehensive access-based pre-fetching strategy uses comprehensive access data to predict future data request patterns. This strategy identifies potential popular content and caches it in advance by analyzing access data from the entire network. The comprehensive access data includes historical access records of users, access popularity during a certain period of time, and geographic location factors. The vehicle node access-based pre-fetching strategy uses the access content of a single vehicle node to predict future data requests, and monitors and records the access behavior of individual vehicle nodes to identify their habits and preferences, thereby performing personalized cache pre-fetching; The pre-fetching strategy based on identification information relies on the specific identification information of the user or content to predict and provide identified data services. The specific identification information includes the user's ID, the unique identifier of the content and the MIN-identifier. The MIN-identifier is the identity of the vehicle node in the multi-identification network system MIN. By analyzing these identification information, the access needs of specific users or specific content are predicted, and related data are cached in advance.

6. The collaborative caching solution for a high-security private network MIN-V2X according to claim 1, characterized in that: In view of the characteristics of large traffic and high mobility in the vehicle environment, a multi-level probabilistic caching strategy is set to complete the cache replacement, wherein the cache node replacement process includes: when a cache block is newly added to the storage space of the node, it is placed in a new area. If the space in the new area is full at this time, the longest prefix match shortest threshold is set, and the cache blocks below the longest prefix match shortest threshold are selected to form a first pre-elimination set. In the first pre-elimination set, the cache blocks that have not been accessed for the longest time are selected and eliminated. When the newly added cache block is accessed again, the cache block is migrated to the old area, and the position of the cache block is adjusted to the head of the queue. If the space in the old area is full at this time, the same method as the elimination in the new area is adopted to select the cache blocks below the longest prefix match shortest threshold to form a second pre-elimination set, and the cache blocks that have not been accessed for the longest time in the second pre-elimination set are screened out, and the selected cache blocks are moved to the new area. If the new area is also full at this time, the cache block with the shortest matching prefix in the first pre-elimination set and closest to the end of the queue is selected again in the new area and eliminated.

7. The collaborative caching solution for a high-security private network MIN-V2X according to claim 5 is characterized in that: By using the collaborative cache search algorithm, when there is a user request in the cluster, the logic for quickly locating the node in the cluster that responds to the request is as follows: In the collaborative cache design of MIN-V2X, the cluster head manages the data cache blocks cached in each of its cluster members. The members of the cluster cache no redundant data. The communication within the entire cluster is only carried out through the cluster head. When a cluster member receives any content cache block request, it first uses the full match algorithm to check whether its own node has cached the content. If not found, it will request to search for the target content from its cluster head, use the longest prefix matching algorithm to find the metadata of the corresponding node in the cluster according to the MIN-identifier of the target data, find the corresponding node in the cluster according to the metadata, and use the full match algorithm again to find the cache block content, and return the data packet; If there is no match in the request of this cluster head, the cluster head will forward the request to the road side unit, and the road side unit will determine to send the request to a neighboring cluster head. The basis for determining the target neighboring cluster head is still the longest prefix matching algorithm to match the cluster head with the most relevant MIN-identifier through the MIN-identifier. After forwarding to the neighboring cluster head, the same search process is followed. If the road side unit still fails to complete the match, the request will be forwarded to the cloud server through the callback function to obtain the content from the cloud server; Each content cache block request determines a limited number of hops. Each time a content cache block request is forwarded, the limited number of hops decreases by 1. Only when the limited number of hops is greater than 0, the collaborative cache search is performed. Otherwise, if the collaborative cache search result is considered to have failed, a negative acknowledgment (NACK) is used to notify the initial requester, and the request is forwarded to the roadside unit, and then the request is forwarded to the cloud server through a callback function to obtain the content from the cloud server.

8. The collaborative caching solution for a high-security private network MIN-V2X according to claim 7 is characterized in that: After executing the collaborative cache search algorithm, if a match is successful within the cluster, the data is returned; if the match fails, the request is forwarded to the roadside unit, which locates the neighbor cluster head and then forwards it to the corresponding cluster member through the neighbor cluster head. This cycle continues until the limited number of hops drops to 0, and then the node that responds to the request is found, and finally the node distributes the content; if the node that responds to the request cannot be found in the cache layer, the requested content is obtained through the cloud service.

Images