A cache method based on node state and content heat

Abstract

The application discloses a cache method based on node state and content heat, which is applied to NDMANET and comprises the following steps: if there is a PIT table item corresponding to a data packet in an intermediate node, then the cache score S of the intermediate node is calculated based on the connectivity, cache free amount and heat of the node cache and the cache score threshold T score ; S cache is compared with T score ; if S cache is less than T score , the cache process is not executed, and the data packet is directly forwarded; if the cache score S cach is greater than or equal to T score , the content is cached in the CS when the cache capacity is sufficient, and the content is attached with an expiration time; when the cache capacity is insufficient, a cache replacement algorithm is executed first, then the cache free amount is updated, and the data packet is forwarded. The application can cache the content with higher heat to a suitable node, so that the node cache hit rate is improved while reducing the user request time delay.

Description

Technical Field

[0001] This invention relates to the field of network information transmission and communication technology, specifically to a caching method based on node status and content popularity in Named-data Mobile Ad Hoc Network (NDMANET). Background Technology

[0002] In recent years, with the development of network communication technology, users' internet speeds have greatly improved. However, with the continuous development of the internet, especially the emergence of emerging technologies such as cloud computing, big data, and the Internet of Things, the modern internet has long since transformed from a small scientific experimental network into a global information infrastructure network, and from a static network into a network with more mobile devices participating. Although the internet is very successful today, the TCP / IP network, with its content location-based architecture, is not performing satisfactorily in the face of the current trend of content-centric services. In terms of scalability, the rapid increase in the number of internet users has led to problems such as insufficient IPv4 addresses. Regarding security, there are currently many security vulnerabilities in the network, and constantly patching these vulnerabilities makes the network increasingly bloated. In terms of mobility, with the emergence of a large number of mobile devices, traditional TCP / IP networks can no longer adequately support them. The main solution to these problems with traditional TCP / IP networks is currently incremental deployment, continuously adding overlays and network patches. While this improves network utilization efficiency and ensures service quality to some extent, it cannot fundamentally solve the problems facing the internet today. Therefore, abandoning the traditional internet architecture and redesigning the network architecture has become a new approach. Among a series of newly proposed Internet architectures, Named Data Networking (NDN) stands out.

[0003] NDN, as a new type of internet architecture, breaks through the limitations of the current internet system, shifting the focus of current communication methods from data location to the data content itself. NDN communication no longer requires a connection like in TCP / IP networks; it is driven by the receiver. To obtain the data a user wants, the consumer sends an Interest Packet containing the name of the data the user wants to receive. When a node in the network has the requested data, it returns a Data Packet to the consumer node. Each routing node contains three tables: the Pending Interest Table (PIT), the Forwarding Information Base (FIB), and the Content Store (CS). The PIT records which nodes forwarded the Interest Packet from. When an Interest Packet arrives at a routing node, it checks the PIT table. If a corresponding entry exists, it records the source port of the Interest Packet; otherwise, it creates the entry and correctly returns the data packet based on the PIT information. After the data packet returns from the corresponding port, the PIT entry matching that data is deleted. The FIB (Finding Instructions Table) is similar to a routing table, storing the routing addresses of nodes. It uses the longest prefix match of the packet's name attribute value to forward the packet from the corresponding port. The CS (Client / Server) table buffers received packets, improving network response speed.

[0004] With the rapid development of wireless communication technology, wireless networks have expanded from the initial ARPANET military network to civilian wireless networks. The booming development of civilian wireless networks has also accelerated the evolution of PRNET (Radio Packet Networking) into Mobile Ad-hoc Networks (MANETs). Wireless networks are divided into two main categories based on whether they have base stations: Infrastructure Networks (with base stations as wireless controllers) and Mobile Ad-hoc Networking (MANETs) without base stations as wireless controllers. Although the application and research of MANETs are mostly limited to military and emergency environments rather than public networks, MANETs have a promising future in wireless networks as more and more mobile devices access them. As a typical wireless network, MANET performance is susceptible to node mobility. In current TCP / IP networks, IP addresses are used as unique identifiers to locate hosts in the network and establish end-to-end link connections, thus enabling reliable data transmission. Considering the mobility of MANET nodes, when a node moves, the connection established via TCP will fail. Therefore, MANET networks based on the TCP / IP protocol suite suffer from short point-to-point connection durations and frequent interruptions when providing connection-oriented services; and high packet loss rates when providing best-effort transmission services. Thus, the TCP / IP protocol, originating from fixed-line networks, is not suitable for the highly dynamic MANET application scenarios.

[0005] Compared to traditional TCP / IP network protocols, NDN is data-centric and introduces node caching technology. NDN's characteristics allow consumer terminals to retrieve data from nearby nodes that have cached the requested content, rather than constantly establishing a communication link with a specific content server. These mechanisms also reduce latency and network traffic for consumer terminals. NDN employs a receiver-driven model, enabling asynchronous communication between two nodes in the network without the need to specifically build and maintain point-to-point connection states. These advantages of NDN make it highly suitable for the data transmission needs of medium- to high-dynamic MANET networks. Therefore, researchers have applied the NDN architecture to MANETs, ​​proposing the Named Data Mobile Ad-hoc Networking (NDMANET). However, current caching strategies used in NDMANETs suffer from low cache hit rates, high user request latency, high bandwidth consumption, and low cache diversity. Summary of the Invention

[0006] To address the shortcomings of existing technologies, this invention provides a caching method based on node status and content heat (NSCH) for Named-data Mobile Ad Hoc Networks (NDMANETs). This caching method fully considers both node status and content heat, selecting nodes with high connectivity and ample cache reserves in the wireless network as caching nodes. It also incorporates the heat of the requested content into the caching decision, thereby reducing network cache redundancy, improving cache hit rate, and reducing user request latency.

[0007] This invention discloses a caching method based on node status and content popularity, applied to named data movement ad hoc networks; comprising:

[0008] The intermediate node receives the data packet and determines whether there is a PIT entry in the intermediate node corresponding to the data packet.

[0009] If a PIT entry exists, then calculate the intermediate node. Cache rating With cache score threshold ;in,

[0010]

[0011]

[0012] In the formula:

[0013] For normalized nodes The connectivity;

[0014] For normalized nodes The amount of free cache;

[0015] For nodes Content popularity;

[0016] Weight for content requests;

[0017] For normalized clusters The average connectivity of the nodes in the middle;

[0018] For normalized clusters Average cache free space;

[0019] By node With nodes active neighbor set composition;

[0020] Compare cache scores With cache score threshold ;

[0021] If cache score Less than the cache score threshold If the caching process is not executed, the data packet will be forwarded directly.

[0022] If cache score Greater than or equal to the cache score threshold Then determine whether the cache capacity is sufficient;

[0023] If the cache capacity is sufficient, cache the content in CS and attach an expiration time to the content; update the remaining cache space according to the cache space occupied by the cached content, and forward the data packet;

[0024] If the cache capacity is insufficient, the cache replacement algorithm is executed first, then the cache free space is updated, and the data packet is forwarded.

[0025] As a further improvement to the present invention, it also includes:

[0026] If no PIT entry corresponding to the data packet exists in the intermediate node, the data packet is discarded.

[0027] As a further improvement to the present invention, the normalized nodes connectivity for:

[0028]

[0029] In the formula:

[0030] For nodes The connectivity;

[0031] For clusters The minimum connectivity of the nodes in the middle;

[0032] For clusters The maximum connectivity of the nodes.

[0033] As a further improvement to the present invention, the normalized nodes cache free space for:

[0034]

[0035] In the formula:

[0036] For nodes The amount of free cache;

[0037] For clusters Minimum free cache size for a node;

[0038] For clusters The maximum amount of free cache in the middle node.

[0039] As a further improvement to the present invention, the normalized cluster Average connectivity of nodes for:

[0040]

[0041] In the formula:

[0042] For clusters The average connectivity of the nodes in the middle;

[0043] For clusters The minimum connectivity of the nodes in the middle;

[0044] For clusters Maximum connectivity of nodes in the middle;

[0045] in,

[0046] cluster Average connectivity of nodes for:

[0047]

[0048] In the formula:

[0049] For nodes The connectivity;

[0050] For nodes The sum of the connectivity of all active neighbors;

[0051] Cluster The total number of nodes in the system.

[0052] As a further improvement to the present invention, the normalized cluster Average cache free space for:

[0053]

[0054] In the formula:

[0055] For clusters Average cache free space in the middle node;

[0056] For clusters Minimum free cache size for a node;

[0057] For clusters Maximum free cache size in the middle node;

[0058] in,

[0059] cluster Average cache free space of middle nodes for:

[0060]

[0061] In the formula,

[0062] For nodes The amount of free cache;

[0063] For nodes The sum of the cache free space of all active neighbors;

[0064] Cluster The total number of nodes in the system.

[0065] As a further improvement of the present invention, when the content is cached in CS, the expiration time is related to the popularity of the content. The higher the popularity of the content, the longer the cache expiration time. At the same time, when a user requests the cached content, the expiration time will be reset. When the cached content reaches the expiration time, CS will delete it.

[0066] As a further improvement of the present invention, the method for determining the number of active neighbors of the current node includes:

[0067] A PreHopNode field is added to both interest packets and data packets to store the nodes traversed by the previous hop of the transmission packet;

[0068] Before forwarding an interest packet or data packet, a node updates its node name in the PreHopNode field. In the next hop, the node that receives the interest packet or data packet will identify the active neighbor nodes of the node based on the PreHopNode field. Based on the number of PreHopNode types received, the number of active neighbors of the current node is determined.

[0069] As a further improvement of the present invention, each node is also used to maintain the number of active neighbors and the remaining cache amount for all its neighbors; the specific maintenance method includes:

[0070] Add an active neighbor number field and a cache remaining field to both interest packets and data packets to transmit the number of active neighbors and the remaining cache of the current node to the next hop node;

[0071] The node that receives the data obtains the number of active neighbors and the remaining cache for a given neighbor through the active neighbor number field and the cache remaining field, and sets the corresponding expiration time.

[0072] As interest packets and data packets are transmitted, each node can maintain the number of active neighbors and the remaining cache for all its neighbors.

[0073] As a further improvement of the present invention, it also includes: recording the node ID, number of active neighbors, and remaining cache in the active neighbor information table of each node.

[0074] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0075] This invention calculates a cache score for arriving content on the current node by combining the current node's state and the popularity of the arriving content. This score is then compared to a threshold. Content is cached on the current node only when the cache score reaches the threshold. The cache score fully considers node connectivity, cache availability, and content popularity. The threshold is dynamically adjusted based on network topology and node cache availability to meet the needs of NDMANETs with dynamically changing node locations. This invention, through joint node and content decision-making, allows highly popular content to be cached on appropriate nodes, thereby improving node cache hit rate while reducing user request latency. Attached Figure Description

[0076] Figure 1 This is a flowchart of the caching method based on node status and content popularity disclosed in this invention.

[0077] Figure 2 This is a schematic diagram of the modified interest packet and data packet structure of the present invention.

[0078] Figure 3 A diagram illustrating the process of maintaining neighboring properties.

[0079] Figure 4 To maintain the intention of representing active neighbor information.

[0080] Figure 5 This is a schematic diagram of the initial network topology for the cache simulation experiment.

[0081] Figure 6This is a statistical chart showing cache hit rate and request latency at different node movement speeds.

[0082] Figure 7 This is a statistical chart showing network traffic and cached content types at different node movement speeds.

[0083] Figure 8 This is a statistical chart showing cache hit rate and request latency for different cache sizes.

[0084] Figure 9 This is a statistical chart showing network traffic and cache content types for different cache sizes. Detailed Implementation

[0085] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0086] The present invention will now be described in further detail with reference to the accompanying drawings:

[0087] A caching method based on Node Status and Content Heat (NSCH) for Named-data Mobile Ad Hoc Network (NDMANET) is proposed. The core idea is to calculate a caching score by measuring the connectivity and cache space limit of a node and combining it with the content's popularity on the node. Then, a score threshold is calculated by comprehensively considering the status of the current cache node and its active neighbor nodes. Finally, a caching decision is made by comparing the caching score with the score threshold.

[0088] Specifically:

[0089] like Figure 1 As shown, this invention provides a caching method based on node status and content popularity, applied in named data movement ad hoc networks; including:

[0090] Step 1: The intermediate node receives the data packet and determines whether there is a PIT entry in the intermediate node that corresponds to the data packet.

[0091] Step 2-1: If it does not exist, discard the data packet.

[0092] Step 2-2: If a PIT entry exists, calculate the intermediate node. Cache rating With cache score threshold ;

[0093] The specific calculation method is as follows:

[0094] Consider a node and all its active neighbor nodes as a cluster. The active neighbor set is , and All nodes in the cluster form a group ; cluster The maximum connectivity of a node is defined as ,cluster The minimum connectivity of a node is defined as After normalization, the nodes The connectivity is:

[0095]

[0096] Similarly, clusters The maximum free cache size of a middle node is defined as follows: ,cluster The minimum free cache size of a middle node is defined as After normalization, the nodes The amount of free cache space is:

[0097]

[0098] In summary, the scoring formula for the NSCH caching strategy is as follows:

[0099]

[0100] In the formula:

[0101] For nodes The popularity of the content.

[0102] To distribute the content that cannot be stored on nodes with high connectivity to their neighboring nodes, it is necessary to comprehensively evaluate the node information within the entire cluster to calculate a scoring threshold. This invention considers a node and all its neighboring nodes as a cluster. The active neighbor set is , and All nodes in the cluster form a group .

[0103] Define cluster Total number of nodes for:

[0104]

[0105] cluster The formula for calculating the average connectivity of nodes is:

[0106]

[0107] In the formula, For clusters The average connectivity of the nodes in the middle For nodes connectivity, For nodes The sum of the connectivity of all active neighbors, Cluster The total number of nodes in the system.

[0108] normalized clusters Average connectivity of nodes for:

[0109]

[0110] cluster The formula for calculating the average free cache space of a node is:

[0111]

[0112] In the formula, For clusters Average cache free space For nodes The amount of free cache space, For nodes The sum of the cache free space of all active neighbors. Cluster The total number of nodes.

[0113] normalized clusters Average cache free space for:

[0114]

[0115] In summary, nodes Cache rating threshold The calculation formula is:

[0116]

[0117] In the formula:

[0118] This sets the content request weight, with a default value of 0.05.

[0119] For normalized clusters The average connectivity of the nodes in the middle;

[0120] For normalized clusters The average amount of free cache.

[0121] Step 3: Compare cache scores With cache score threshold .

[0122] Step 4-1, if cached score Less than the cache score threshold If the caching process is not executed, the data packet will be forwarded directly.

[0123] Step 4-2, if cached score Greater than or equal to the cache score threshold Then determine whether the cache capacity is sufficient.

[0124] Step 5-1: If the cache capacity is sufficient, cache the content in CS and attach an expiration time to the content; update the remaining cache space according to the cache space occupied by the cached content, and forward the data packet; the expiration time is related to the popularity of the content, the higher the popularity of the content, the longer the cache expiration time; at the same time, when a user requests cached content, the expiration time will be reset; when the cached content reaches the expiration time, CS will delete it.

[0125] Step 5-2: If the cache capacity is insufficient, first execute the cache replacement algorithm, then update the cache free space, and forward the data packet.

[0126] To implement the NSCH strategy described above, it is necessary to obtain information such as the number of neighboring nodes of a node. Therefore, this method modifies the interest packets and data packets, and realizes the interaction of information between nodes through the transmission of interest packets and data packets.

[0127] Specifically:

[0128] like Figure 2 As shown, the information exchange mechanism of this invention, such as the number of active neighbor nodes, relies on interest packets and data packets transmitted in the network. This ensures both the timeliness of information and effectively saves the bandwidth consumption of NDMANET. Therefore, this invention... Figure 1The shown interest packets and data packets each include an additional field (PreHopNode) to store the nodes traversed in the previous hop of the transmission. Specifically, before forwarding an interest packet or data packet, a node updates its own node name in this field. At the next hop, the node receiving the interest packet or data packet identifies active neighbor nodes based on the PreHopNode field. The number of received PreHopNode types determines the current node's number of active neighbors. Thus, through the continuous transmission of interest packets and data packets, each node updates its neighbor count in real time, achieving the effect of maintaining neighbors with relatively low bandwidth consumption.

[0129] The specific process of discovering the neighbor is as follows: Figure 3 As shown, Figure 3 In diagram (a), the process after consumer C broadcasts an interest packet begins. Initially, the interest packet carries the PreHopNode of node C. When nodes M1 and M2 receive the interest packet, they have discovered neighbor C. Then, the interest packet updates its PreHopNode based on the nodes reached at each hop, until the interest packet reaches the producer. Through the broadcast of the interest packet, the intermediate nodes have discovered some neighbors. Next... Figure 3 In (b), the process of producers P1, P2, and P3 returning response data packets is shown. In the returned data packets, PreHopNode will also be continuously updated according to the node reached by each hop. New neighbor nodes will be added to the active neighbor table of the current intermediate node and an expiration time will be set. Finally, each intermediate node successfully maintains its own active neighbor nodes.

[0130] Similarly, this method also adds an Active NeighborNum field and a CacheRemaining field to the interest packets and data packets. These two fields transmit the number of active neighbors and the remaining cache of the current node to the next-hop node, respectively. The receiving node can obtain the number of active neighbors and the remaining cache of a certain neighbor and set the corresponding expiration time. With the transmission of interest packets and data packets, each node can maintain the number of active neighbors and the remaining cache of all its neighbors. This information will be recorded in the Active Neighbor Information Table (ANIT) of each node, which records the node ID, the number of active neighbors, and the remaining cache.

[0131] by Figure 4 For example, through information transmission between nodes, Figure 4 In (a) of the diagram, each node has already maintained its own number of active neighbors and remaining cache in ANIT. The following is... Figure 4In (b) of the diagram, each node shares its active neighbor count and cache availability with its neighbors via interest packets and data packets, thus completing the dynamic maintenance of ANIT. If a node and all its active neighbors are considered a cluster in the network, then the average connectivity and average cache availability of nodes in its cluster can be calculated for each node, providing the NSCH caching strategy with a threshold for caching decision scoring.

[0132] Example:

[0133] In the NSCH caching strategy, the structure of interest packets and data packets also needs to be modified. This step is similar to the forwarding strategy and needs to be implemented by modifying the tags.hpp file under / ndn-cxx / src / lp. The simulation experiment was conducted in an Ubuntu 20.04 operating system environment, using ndnSIM version 2.8. The main simulation experimental work included writing the main program of the simulation scenario, initializing the network topology used in this simulation experiment, creating and configuring consumer and producer nodes using ns3::ndn::ConsumerCbr, and setting the parameters in the simulation experiment.

[0134] This experiment simulated an 802.11g mobile ad hoc network within a 1000m × 1000m rectangular space, using an ErpOfdmRate of 54Mbps data transmission mode. Sixty-four nodes were deployed in the network, moving randomly with a maximum speed of 20m / s. Communication employed the ndnSIM server-consumer model, with each content packet size set to 4MB. The CS table cache replacement strategy used an LRU policy, with each node having a cache capacity of 1024MB. The content popularity model followed a Zipf distribution, and there were 100,000 content types. The experimental parameters are recorded in Table 1.

[0135] Table 1 Simulation Experiment Parameters

[0136]

[0137] In the experiment, 20% of the nodes were selected as consumers and 20% as producers. Each producer could satisfy all the consumer's requests. Figure 5 This demonstrates an initial network topology randomly generated in the experiment, where circles represent intermediate nodes, pentagons represent consumers, and squares represent producers. Nodes within communication range are connected by line segments, indicating that communication is possible between two nodes.

[0138] Computer simulations were used to evaluate the performance of the proposed NSCH caching strategy. Performance metrics included: Average request delay (ARD), cache hit rate (CHR), total network traffic (TNT), and types of network cache content (TNCC).

[0139] (1) Average Request Latency (ARD). This measures the average latency from when a consumer sends a request to when they receive the requested content, thus measuring the overall data transmission efficiency of the network. A lower average request latency indicates that the consumer receives the requested data faster, and the network service quality is higher. Definition The sum of the latency for each request. This represents the total number of requests. The average request latency is calculated using the following formula:

[0140]

[0141] (2) Cache hit rate (CHR) is one of the important metrics for measuring NDN performance. It reflects the proportion of requests satisfied by routing nodes rather than servers, thus balancing content requests among available cache resources. When the content requested by a user is in the cache of a routing node, a request to the origin server can be avoided, thereby reducing the burden on the origin server and network traffic. Therefore, a high cache hit rate can improve network performance and efficiency and reduce the waste of network resources. In NDN networks, cache hit rate typically refers to the proportion of data that is hit in the cache of a routing node after a consumer makes a request. Definition The number of requests that hit the cache. This represents the total number of requests. The cache hit rate is calculated using the following formula:

[0142]

[0143] (3) Total Network Traffic (TNT). This calculates the total data transmission traffic in the network and measures the impact of each caching strategy on network link load. Lower network traffic indicates less bandwidth consumed by consumers requesting data, and a more reasonable content caching location.

[0144] (4) Network Cache Content Types (TNCC). This calculates the number of cached content types at each node in the network, measuring the overall cache diversity. A higher TNCC indicates greater cache diversity in the network, which better meets diverse user requests and improves network service quality.

[0145] The experiment used node movement speed and node cache capacity as variables, conducting two experiments. Four different caching strategies were compared: LCE, LCD, ProbCache, and MPC. The LCE caching strategy caches data on all nodes along the data packet's path, leading to significant cache redundancy. When cache space is limited, the node's cache hit rate decreases. The LCD caching strategy uses downstream caching, caching content on nodes closer to the user while also considering content popularity. As the number of requests for the same content increases, it gradually becomes popular and is cached on nodes closer to the user at a slower rate. In wireless networks, nodes consist of mobile terminals, each acting as both a routing node and a user. Therefore, using the LCD caching strategy may reduce the cache hit rate. ProbCache is a probabilistic caching method that does not consider node state and content popularity; therefore, the most needed content may not be cached, resulting in a low node cache hit rate and high user request latency. MPC is a content popularity-based caching strategy, but it only considers popularity metrics and not node state information, thus failing to effectively reduce user request latency.

[0146] (1) Using node movement speed as a variable

[0147] In this experiment, the cache capacity of each node was fixed at 1GB, and the node movement speed was set to 0, 5, 10 and 20 m / s respectively. Figure 6 (a) shows a comparison of cache hit rate performance for each caching strategy. LCE performs the worst in terms of caching performance compared to other methods. This is because LCE blindly caches content on all nodes through which data packets flow, leading to severe content redundancy and frequent content replacement within nodes, which greatly reduces the system's caching performance. ProbCache can improve the utilization of cache resources to some extent, so its cache hit rate is slightly higher than LCE. Under long-term operation, MPC can cache popular content on various nodes in the network, making it easier for user requests to hit the node cache, so MPC performs the second best. This solution combines node status and content popularity information, storing cached content on nodes with high connectivity and prioritizing the caching of more popular content, thus achieving the best performance. Specifically, when the parameter node movement speed is 20m / s, the cache hit rate of this solution is 29.4%, which is 6% higher than the cache hit rate of MPC. When the node movement speed varies between 0m / s and 20m / s, the average cache hit rate of this solution is 28.48%, which is 4.12% higher than MPC's 24.36%, 5.73% higher than LCD's 22.75%, 10.25% higher than ProbCache's 18.23%, and 11.66% higher than LCE's 16.82%.

[0148] NSCH can gradually cache more popular content on nodes with higher connectivity based on the number of user requests. Therefore, in Figure 6 In (b) of the above, the NSCH proposed in this patent has the lowest request latency. MPC caches popular content to nodes but does not consider the location of the cache nodes, resulting in higher request latency. LCD can gradually cache content in nodes closer to the user, achieving the second-best latency performance when the node movement speed is 0 m / s, but the latency begins to increase as the node moves. ProbCache's random caching cannot ensure that content is cached in nodes close to the user, thus its latency performance is also poor. LCE blindly caches content in all routing nodes, resulting in the worst latency performance. In terms of average latency at the four node movement speeds, this scheme is 754ms, which is 14.7% lower than LCD's 884ms, 15.66% lower than LCE's 894ms, 7.59% lower than ProbCache's 816ms, and 4.92% lower than MPC's 793ms. Therefore, NSCH outperforms other caching strategies in terms of latency.

[0149] NSCH considers the available cache space of each node when making caching decisions, making efficient use of each node's cache space and distributing the cache across neighboring nodes. This makes it easier for consumers to find the requested content in the caches of intermediate nodes, reducing the number of transmissions of interest packets and data packets, and reducing network traffic usage. Therefore, in Figure 7 In (a) of the data, the network traffic consumption of the NSCH caching strategy is significantly lower than that of LCD, LCE, ProbCache, and MPC. Similarly, in NSCH, cache nodes with more idle space have higher cache scores, allowing for a more even distribution of cache across nodes. This results in a greater variety of cached content in the network, improving cache diversity. Therefore, as... Figure 7 As shown in (b), in networks using the NSCH caching strategy, the number of cached content types is significantly higher than other caching strategies. When node movement speed varies between 0 m / s and 20 m / s, the average number of cached content types in this scheme is 174% higher than LCE, 132% higher than LCD, 28% higher than ProbCache, and 88% higher than MPC. This demonstrates the advantages of NSCH in both saving network bandwidth and improving cache diversity.

[0150] (2) Using node cache capacity as a variable

[0151] In this experiment, the movement speed of each node was fixed at 10m / s, and the cache capacity was set to 0.1, 0.5, 1, and 2GB respectively. Figure 8The comparison in (a) shows that the caching strategy proposed in this patent performs best in terms of average cache hit rate. With the increase in cache capacity, the average cache hit rate of all five caching strategies significantly improves, especially the NSCH strategy. This is because with increased cache space, nodes can cache more content, thereby improving the cache hit rate. Combined with... Figure 8 As shown in (b), the average request latency of NSCH is also relatively low. This is because as the node cache capacity increases, more popular content is cached in nodes with higher connectivity, reducing the average number of hops for user requests and thus lowering the overall average request latency of the network. Therefore, the NSCH caching strategy proposed in this patent can improve cache hit rate and reduce average request latency, demonstrating good performance.

[0152] exist Figure 9 In (a) of the data, the NSCH caching strategy exhibits significantly lower network traffic consumption than LCD, LCE, ProbCache, and MPC, and the network traffic consumption gradually decreases as the node cache capacity increases. This is because NSCH comprehensively considers the node's cache availability when making caching decisions, making reasonable use of each node's cache space and distributing the cache across neighboring nodes. This makes it easier for consumers to find the requested content in the caches of intermediate nodes, reducing the number of transmissions of interest packets and data packets, thereby reducing network traffic usage. As the node cache capacity increases, nodes with higher connectivity can cache more frequently accessed content, further reducing network traffic consumption. Therefore, the NSCH caching strategy demonstrates good performance in terms of network traffic consumption.

[0153] Similarly, in the NSCH caching strategy, cache nodes with more free space receive higher cache scores, resulting in a more even distribution of cached content and increasing the diversity of cached content in the network. For example... Figure 9 As shown in (b), in networks using the NSCH caching strategy, the number of cached content types is significantly higher than other caching strategies. When the node cache capacity is 2GB, the number of cached content types using the NSCH strategy across the entire network is 2.13 times that of LCE, 1.72 times that of LCD, 1.13 times that of ProbCache, and 1.22 times that of MPC. This demonstrates the advantages of NSCH in both saving network bandwidth and improving cache diversity.

[0154] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A caching method based on node state and content popularity, applied to named data movement ad hoc networks; characterized in that, include: The intermediate node receives the data packet and determines whether there is a PIT entry in the intermediate node corresponding to the data packet. If a PIT entry exists, then calculate the intermediate node. Cache rating With cache score threshold ;in, In the formula: For normalized nodes The connectivity; For normalized nodes The amount of free cache; For nodes Content popularity; Weight for content requests; For normalized clusters The average connectivity of the nodes in the middle; For normalized clusters Average cache free space; By node With nodes active neighbor set composition; Compare cache scores With cache score threshold ; If cache score Less than the cache score threshold If the caching process is not executed, the data packet will be forwarded directly. If cache score Greater than or equal to the cache score threshold Then determine whether the cache capacity is sufficient; If the cache capacity is sufficient, cache the content in CS and attach an expiration time to the content; update the remaining cache space according to the cache space occupied by the cached content, and forward the data packet; If the cache capacity is insufficient, the cache replacement algorithm is executed first, then the cache free space is updated, and the data packet is forwarded.

2. The caching method based on node status and content popularity as described in claim 1, characterized in that, Also includes: If no PIT entry corresponding to the data packet exists in the intermediate node, the data packet is discarded.

3. The caching method based on node status and content popularity as described in claim 1, characterized in that, Normalized nodes connectivity for: In the formula: For nodes The connectivity; For clusters The minimum connectivity of the nodes in the middle; For clusters The maximum connectivity of the nodes.

4. The caching method based on node status and content popularity as described in claim 1, characterized in that, Normalized nodes cache free space for: In the formula: For nodes The amount of free cache; For clusters Minimum free cache size for a node; For clusters The maximum amount of free cache in the middle node.

5. The caching method based on node status and content popularity as described in claim 1, characterized in that, normalized clusters Average connectivity of nodes for: In the formula: For clusters The average connectivity of the nodes in the middle; For clusters The minimum connectivity of the nodes in the middle; For clusters Maximum connectivity of nodes in the middle; in, cluster Average connectivity of nodes for: In the formula: For nodes The connectivity; For nodes The sum of the connectivity of all active neighbors; Cluster The total number of nodes in the system.

6. The caching method based on node status and content popularity as described in claim 1, characterized in that, normalized clusters Average cache free space for: In the formula: For clusters Average cache free space in the middle node; For clusters Minimum free cache size for a node; For clusters Maximum free cache size in a middle node; in, cluster Average cache free space of middle nodes for: In the formula, For nodes The amount of free cache; For nodes The sum of the cache free space of all active neighbors; Cluster The total number of nodes in the system.

7. The caching method based on node status and content popularity as described in claim 1, characterized in that, When caching this content in CS, the expiration time is related to the popularity of the content; the higher the popularity of the content, the longer the cache expiration time. At the same time, when a user requests the cached content, the expiration time will be reset. When the cached content reaches its expiration time, CS will delete it.

8. The caching method based on node state and content popularity as described in any one of claims 1 to 7, characterized in that, The methods for determining the number of active neighbors of the current node include: A PreHopNode field is added to both interest packets and data packets to store the nodes traversed by the previous hop of the transmission packet; Before forwarding an interest packet or data packet, a node updates its node name in the PreHopNode field. In the next hop, the node that receives the interest packet or data packet will identify the active neighbor nodes of the node based on the PreHopNode field. Based on the number of PreHopNode types received, the number of active neighbors of the current node is determined.

9. The caching method based on node state and content popularity as described in any one of claims 1 to 7, characterized in that, Each node is also used to maintain the number of active neighbors and the amount of cache remaining for all its neighbors; Specific maintenance methods include: Add an active neighbor number field and a cache remaining field to both interest packets and data packets to transmit the number of active neighbors and the remaining cache of the current node to the next hop node; The node that receives the data obtains the number of active neighbors and the remaining cache for a given neighbor through the active neighbor number field and the cache remaining field, and sets the corresponding expiration time. As interest packets and data packets are transmitted, each node can maintain the number of active neighbors and the remaining cache for all its neighbors.

10. The caching method based on node state and content popularity as described in claim 9, characterized in that, Also includes: Record the node ID, number of active neighbors, and remaining cache in the active neighbor information table for each node.

Images