An energy-efficient collaborative caching and forwarding method for information-centric low-Earth orbit satellite networks
By dividing autonomous domains in low-Earth orbit (LEO) satellite networks and adopting a cooperative caching and forwarding method, and by utilizing eigenvector centrality and virtual node topology, the energy waste and latency problems caused by resource constraints and aperiodic requests in LEO satellite networks are solved, achieving more efficient energy management and data transmission.
Patent Information
- Application Number
- CN202311189295.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2022-12-14
- Filing Date
- 2023-09-14
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2043-09-14
AI Technical Summary
Low-Earth orbit satellite networks in information-centric networks suffer from energy waste and latency issues due to node resource constraints and non-periodic data requests, especially in scenarios with sudden requests. Existing energy-saving methods have failed to effectively coordinate node collaborative caching.
By dividing autonomous domain networks into low-Earth orbit satellite networks, a cooperative caching and forwarding method is adopted. The optimal cooperative range is determined by eigenvector centrality. Cooperative caching decisions are made by combining energy consumption and data acquisition latency. A domain-based topology structure based on virtual nodes is adopted to reduce cooperative overhead. A state-aware algorithm is used to obtain update messages to achieve network-wide routing convergence.
It effectively reduced energy consumption, decreased data acquisition latency, improved network energy efficiency and data availability, and optimized the selection of cross-domain connection nodes.
Smart Images

Figure CN117202329B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of wireless communication technology, and specifically to an energy-efficient collaborative buffering and forwarding method for low-Earth orbit satellite networks centered on information. Background Technology
[0002] With the development of wireless communication technology, data collection via low-Earth orbit (LEO) satellite wireless networks (such as maritime data and earthquake emergency communications) has rapidly progressed. Furthermore, due to the continuous increase in the number of connected mobile devices, content-centric services will account for over 85% of internet traffic, and this proportion will continue to rise. The limitations of traditional IP architectures based on end-to-end communication models are becoming increasingly apparent, such as low content distribution efficiency and poor mobility. Therefore, Information-Centric Networks (ICNs) have emerged. Establishing an Information-Centric LEO Satellite Network (ICSN), leveraging its features such as content naming, network caching, and name routing, allows users to obtain data from multiple content sources (providers or caching nodes) in environments with highly dynamic satellite topologies and limited network resources, achieving content sharing and reducing latency. However, LEO satellite nodes are limited by processing, memory, and energy resources, making energy efficiency a major challenge for LEO satellite networks.
[0003] In Information-Centric Networks (ICNs), data packets are sent directly via message names and can be temporarily stored in the network to fulfill subsequent requests. Therefore, existing work has extensively studied improving the energy efficiency of low-Earth orbit satellite networks (LEOs) from various aspects, including hardware optimization and caching techniques. However, the impact of ICSN's unique receiver-driven communication on energy efficiency is often overlooked, potentially leading to additional energy consumption, especially due to long wait times for sudden requests. After satellite nodes perform data collection tasks and generate data, they must remain online to respond to subsequent requests. This requires both the data-generating satellite node and satellite nodes along the request path to remain active to ensure data access, resulting in significant energy waste. In ICSN scenarios with periodic data generation and occasional data requests, such as weather forecast observation data and emergency communication data, the frequency of data requests is usually lower than the frequency of data generation. In these cases, nodes can periodically plan data collection based on their own circumstances (e.g., performing a sensing task every 100ms, while occasional data requests may occur every tens of seconds or even once a day). These non-periodic or even infrequent request behaviors can cause nodes to remain in a waiting state for extended periods, increasing energy consumption. A common method to conserve energy is to put some nodes into a sleep state, but this can lead to data unavailability and additional data retrieval delays. When a request arrives in an ICSN sub-area, a service node might be in a sleep state and therefore unable to respond, or some satellite nodes on a path might be unable to send packets because they are in a sleep state. Although considerable efforts have been made to address energy conservation in ICSNs, there has been insufficient research into collaborative caching involving other nodes. Summary of the Invention
[0004] To further reduce energy consumption, network caching and sleep scheduling can be coordinated to minimize energy consumption. By cooperating with other satellite nodes, energy-constrained satellite nodes can enter a sleep state, thereby reducing total energy consumption. Therefore, this invention provides an information-centric low-Earth orbit satellite network energy-saving cooperative caching and forwarding method. It seeks a trade-off between energy consumption and data acquisition latency, determines the optimal cooperative range based on eigenvector centrality, and then makes cooperative caching decisions from the perspectives of energy consumption, freshness, and popularity to achieve comprehensive decision-making. Secondly, a domain-based topology structure based on virtual nodes is adopted to reduce cooperative overhead. After satellite nodes cache data within the regional cooperative caching range, nodes use a state-aware algorithm to obtain update messages to achieve network-wide routing convergence. Virtual nodes are mapped through an auxiliary graph algorithm and then use a shortest path algorithm to obtain the best forwarding node information, i.e., the optimal cross-domain connection node.
[0005] This invention is achieved through the following technical solution:
[0006] An energy-efficient cooperative buffering and forwarding method for information-centric low-Earth orbit satellite networks includes:
[0007] S1. Based on geographical scenarios and service requirements, the low-orbit satellite network is divided into multiple different autonomous domain networks, each of which includes multiple nodes;
[0008] S2. When a requesting node sends an interest packet about a certain request to the first receiving node closest to the requesting node, it queries whether the cache content library of the first receiving node stores a data packet that matches the interest packet; if so, it returns the data packet to the requesting node; otherwise, it jumps to S3.
[0009] S3. Check if the cached routing table of the first receiving node has a record of a node corresponding to the data packet of the interest packet. If yes, proceed to S4; otherwise, discard the interest packet and proceed to S7.
[0010] S4. Determine whether the data packet is in any node other than the first receiving node in the first autonomous system network where the first receiving node is located. If yes, define the node containing the data packet as the first data production node and jump to S5; otherwise, jump to S6.
[0011] S5. Calculate the cooperative buffer range of the first receiving node in the first autonomous system network based on energy consumption and latency factors, wake up the corresponding first cooperative node and the first data production node, and forward the interest packet to the first data production node of the first autonomous system network through the corresponding first cooperative node. The first data production node sends out the corresponding data packet and, through the first cooperative node, buffers and forwards the data packet to the requesting node in the opposite direction of the interest flow of the original interest packet according to the actual situation.
[0012] S6. Based on the virtual node information mapped by the first autonomous system network and the second autonomous system network where the node containing the data packet corresponding to the interest packet is located, establish a cross-domain connection and construct a corresponding auxiliary graph, find the optimal cross-domain connection node, and forward the interest packet to the second data production node of the second autonomous system network through the optimal cross-domain connection node. The second data production node finds the corresponding data packet by referring to the steps of S1-S5. The data packet is cached and forwarded to the request node in the opposite direction of the interest flow of the original interest packet according to the actual situation.
[0013] S7, End.
[0014] As an optimization, the node has three operating modes: active, standby, and sleep. When the node is in active mode, it acts as a data producer, performing data collection, receiving, and forwarding of interest packets and data packets, while also caching cached information from other nodes. When the node is in standby mode, it can receive interest packets and data packets. Based on caching decisions, the node switches to active mode when it needs to perform forwarding and / or caching operations. When the node is in sleep mode, it does not perform any operations and can be activated and woken up by calling an activation clock.
[0015] As an optimization, in S5, the specific process of calculating the cooperative caching range of the first data production node in the first autonomous region network based on energy consumption and latency factors is as follows:
[0016] A1. Calculate the total energy consumption of multi-hop cooperative caching for all nodes in the first autonomous system network. The specific formula is as follows:
[0017] E(H)=E SUM_C +(1-(1-γ) Η E A k;
[0018] Among them, E SUM_C Let E be the total energy consumption of all nodes in the first autonomous system network from data generation to data retrieval; γ∈[0,1), is the risk coefficient for cooperation between nodes; H is the maximum cooperative buffer range, i.e., the first data producing node forwards data packets to the second cooperative node at a distance of H hops, and the second cooperative node is the node in standby mode; A The energy consumption of the first data production node in active mode, where k is the number of hops from the requesting node to the first data production node;
[0019] A2. Calculate the total latency in the first autonomous system network using the following formula:
[0020] T(H) = T NON -Y Η T C -(1-Y Η (T) NON +T D )
[0021] =Y Η (T NON -T C )-(1-Y Η )T D
[0022] =(1-γ) Η (TK-TK'+T D )-TD ;
[0023] Among them, T NON =TK, representing the average latency when nodes in the first autonomous system network do not cooperate, T C =TK′, representing the average latency of nodes cooperating in the first autonomous system network, where T is the average latency of a node's one-hop distance. D Given a sleep schedule duration, K is the number of hops from the requesting node to the first data producer node, and K′ is the path distance from the requesting node to the second collaborating node, K′ = KH; Y H =(1-γ) H γ∈[0,1), is the risk coefficient for cooperation;
[0024] A3. Design a cooperative caching objective function H based on the total energy consumption and the total latency. * :
[0025]
[0026] in, and These are the weights of total energy consumption and total latency, respectively.
[0027] A4. Substitute the eigenvector centrality into the cooperative caching objective function and calculate the cooperative caching range.
[0028] As an optimization, the total energy consumption E of all nodes in the first autonomous system network during the time period t from data generation to data retrieval is... SUM_C Specifically, this includes energy consumption during the activity phase (E). C1 Energy consumption E during the collaboration phase C2 Energy consumption E during the waiting phase C3 Energy consumption E during the reaction stage C4 ,in,
[0029] Energy consumption during the activity phase E C1 The specific formula is:
[0030] E C1 =E A +E S (M t -1);
[0031] Among them, E A E represents the energy consumption of a node in active mode. S M represents the power consumption of the node in standby mode. t The number of nodes in the first autonomous domain network;
[0032] Energy consumption E during the collaboration phase C2 The specific formula is:
[0033] E C2 =E A Θ Η +E S (M t -Θ Η );
[0034] Where H is the maximum collaborative cache range of the data production node, and Θ Η This represents the number of all nodes in active mode within the maximum collaborative cache range.
[0035] Energy consumption E during the waiting phase C3 The specific formula is:
[0036] E C3 =(t wait -1)(E D (Θ Η -M C )+E S (M t -Θ Η +M C ));
[0037] Where, Θ Η -M C =Θ Η -(Θ Η -Θ Η-1 P represents the number of nodes in sleep mode in the first autonomous region network, and t represents the number of nodes in sleep mode. wait E represents the total duration of the second collaborating node's collaboration and waiting phases. D M represents the energy consumption of nodes in sleep mode within the first autonomous region network. C M represents the number of nodes participating in cooperation outside the second cooperative cache range, i.e., nodes not at the H hop participating in cooperation. C =P(Θ) Η -Θ Η-1 );
[0038] Energy consumption E during the reaction stage C4 The specific formula is:
[0039] E C4 =E A k'+E S (M t -Θ Η +M C -k')+E D (Θ Η -M C ).
[0040] As an optimization, the specific steps for A4 are as follows:
[0041] A4.1. Use eigenvector centrality to evaluate the risk coefficient γ, and then substitute this risk coefficient γ into the cache objective function H. * The following formula is obtained:
[0042]
[0043] Z is a constant;
[0044]
[0045] N(v) is a set of neighborhoods of node v, obtained through a routing protocol, and G is the first autonomous system network;
[0046] A4.2 Calculate the optimal collaborative cache range using the power-law iterative eigenvalue algorithm. S5.4.1 The collaborative cache objective function using eigenvector centrality is used to obtain the optimal collaborative cache range.
[0047] As an optimization, in S5, the specific process by which the first data production node sends out the corresponding data packet and, through at least one of the second cooperating nodes, caches the data packet according to the actual situation and forwards it to the requesting node is as follows:
[0048] B1. The first data production node sends a data packet to the second cooperative node within the second cooperative buffer range;
[0049] B2. The second cooperating node determines whether the data packet is already cached in its own cache content library. If so, it discards and forwards the data packet; otherwise, it jumps to B3.
[0050] B3. Determine whether to cache the data packet based on the caching decision. If yes, proceed to B4; otherwise, discard and forward the data packet.
[0051] B4. Update the second cooperating node's own cached routing table and send cache update information to the second cooperating node's neighboring nodes;
[0052] B5. When the neighbor node receives the cache update message from the second cooperating node, it queries its own cache routing table to see if there is a record of the second cooperating node's cache update message. If so, the cache update information is discarded; otherwise, the neighbor node updates its own cache routing table.
[0053] As an optimization, the caching decision specifically includes the node's remaining energy ratio, cache freshness, and cache popularity, wherein...
[0054] The remaining energy ratio B energy ∈(0,1), representing the ratio of the remaining energy in a node to the total energy of that node;
[0055] The cache freshness refers to a measure of the freshness of cached content, and the specific formula is as follows:
[0056]
[0057] Among them, |C v_ability | indicates the caching capacity of the current node, and time(c) indicates the time that the corresponding cached content is stored in the node;
[0058] The cache popularity factor takes into account the request's preference for different content. The node will make caching decisions based on the content popularity factor, and the specific formula is as follows:
[0059]
[0060] in, I represents the total number of requests received by all nodes within an autonomous system network. v The number of requests received by node v;
[0061] The cache decision P v The calculation formula is as follows:
[0062] P v =ω1B energy (v)+ω2F fresh (v)+ω3Q v ;
[0063] Among them, ω1, ω2, and ω3 are adjustment parameters for the caching probability.
[0064] As an optimization, in S6, based on the virtual node information mapped by the first autonomous system network and the second data production node containing the data packet corresponding to the interest packet, a cross-domain connection is established and a corresponding auxiliary graph is constructed. The optimal cross-domain connection node is found, the second data production node is woken up, and the interest packet is forwarded to the second data production node of the second autonomous system network through the optimal cross-domain connection node. The specific process is as follows:
[0065] S6.1 Based on the domain information of the autonomous system network, the second cooperative node in the first autonomous system network and the third cooperative node in the second autonomous system network are copied as the second sub-cooperative node and the third sub-cooperative node, respectively.
[0066] S6.2. Find the physical links between each second cooperating node, between third cooperating nodes, and between second and third cooperating nodes through the domain link routing table;
[0067] S6.3. Set up the first virtual node and the second virtual node in the first autonomous system network and the second autonomous system network, and find the second sub-cooperative node and the third sub-cooperative node that can be used as the first virtual node and the second virtual node respectively through the hash mapping method, and connect the corresponding virtual node and sub-cooperative node with virtual links to complete the construction of the auxiliary graph.
[0068] S6.4 Set the link cost and use the shortest path algorithm to obtain the routing information from the first virtual node to the second virtual node from the auxiliary graph, thereby finding the second sub-cooperative node and the third sub-cooperative node with the minimum corresponding link cost as the most optimal cross-domain connection nodes.
[0069] As an optimization, there is at least one first virtual node located in the first autonomous system network and at least one second virtual node located in the second autonomous system network.
[0070] As an optimization, the link cost includes:
[0071] Actual physical link cost:
[0072]
[0073] β+δ=1;
[0074]
[0075] in, This indicates the transmission delay between two nodes; This represents the energy consumption required to transmit data between two nodes. β and δ are adaptive factors for link weight, which can be adjusted according to the actual situation of the link.
[0076] Virtual link cost:
[0077]
[0078] α∈[0,1];
[0079]
[0080] in, Represents the centrality of a node. Represents node S m To node S n The total number of all paths Represents node S m To node S n Passing through node S i The number of paths, This represents a node's content service capability, that is, the degree to which a node is successful in acquiring specific data. and Representing node S respectively i The total number of forwarded requests and the number of responses received, where S m S is the node that initiated the request. n It is a data production node that caches data packets, S i From S m To S n The intermediate nodes of this path are called collaborating nodes.
[0081] Compared with the prior art, the present invention has the following advantages and beneficial effects:
[0082] This invention seeks a trade-off between energy consumption and data acquisition latency, determines the optimal collaborative range based on eigenvector centrality, and then makes collaborative caching decisions from the perspectives of energy consumption, freshness, and popularity to achieve a comprehensive decision. Secondly, it adopts a domain-based topology structure based on virtual nodes to reduce collaborative overhead. After satellite nodes cache data within the regional collaborative caching range, nodes use a state-aware algorithm to obtain update messages to achieve network-wide routing convergence. Virtual nodes are mapped using an auxiliary graph algorithm and then use a shortest path algorithm to obtain the best forwarding node information, i.e., the optimal cross-domain connection node. Attached Figure Description
[0083] To more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be considered as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort. In the drawings:
[0084] Figure 1 This is a flowchart illustrating the process after a node receives an interest packet in an information-centric low-Earth orbit satellite network energy-saving cooperative buffering and forwarding method as described in this invention.
[0085] Figure 2 This is a flowchart of the cache collaboration process in an information-centric low-Earth orbit satellite network energy-saving collaborative cache forwarding method described in this invention.
[0086] Figure 3 This is a schematic diagram of the node working model;
[0087] Figure 4 A diagram illustrating the energy consumption of each node in the collaborative caching process.
[0088] Figure 5 A schematic diagram of a satellite network model with virtual nodes;
[0089] Figure 6A schematic diagram of selecting an auxiliary graph for virtual nodes (the left side is the original graph, and the right side is the auxiliary graph).
[0090] Figure 7 This is an example diagram illustrating cache update scenarios.
[0091] Figure 8 This is a schematic diagram of the storage structure of the link routing table. Detailed Implementation
[0092] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail below with reference to the embodiments and accompanying drawings. The illustrative embodiments and descriptions of the present invention are only used to explain the present invention and are not intended to limit the present invention.
[0093] Example
[0094] An energy-efficient cooperative buffering and forwarding method for information-centric low-Earth orbit satellite networks includes:
[0095] S1. Based on geographical scenarios and service requirements, the low-orbit satellite network is divided into multiple different autonomous domain networks, each of which includes multiple nodes;
[0096] To reduce the scope of collaborative caching, based on the different tasks within the satellite coverage area and considering factors such as application scenarios, service types, geographical locations, and application density in the ground area, the low-Earth orbit satellite network is divided into several autonomous system networks of different sizes with fine granularity, reducing the significant overhead that may be caused by flooding cooperation. For example... Figure 5 As shown, based on geographical scenarios and service needs, the network can be divided into urban areas, desert areas, ocean areas, etc.
[0097] In this embodiment, the node has three working modes: active, standby, and sleep. When the node is in active mode, it is a data production node, performing data collection, receiving and forwarding of interest packets and data packets, and caching cached information of nodes other than the data production node. When the node is in standby mode, it can receive interest packets and data packets. Based on caching decisions, when the node needs to perform forwarding and / or caching operations, it switches to active mode. When the node is in sleep mode, it does not perform any operations and can be activated and woken up by calling an activation clock.
[0098] Considering multi-hop low-Earth orbit satellite network scenarios where satellite nodes frequently and periodically perform transmission tasks and generate collected data, while users randomly request data, typical real-world scenarios include maritime satellite networks, emergency satellite networks, and remote sensing satellite networks. For example, in the case of a satellite network used for weather data collection, the satellite periodically performs data collection tasks and generates data packets, while users only request this data when they are interested in the weather. Therefore, this invention defines three operating modes for nodes: 1) active; 2) standby; 3) sleep. The transitions between the three modes are as follows: Figure 3 As shown. In active mode, satellite nodes perform data collection tasks, receive and forward data packets, and perform additional functional operations, such as caching content from other nodes. In standby mode, nodes operate in a low-power mode, consuming less energy than in active mode. In this mode, nodes can still receive interest packets and data packets. However, they only switch to active mode to process data packets when other operations such as forwarding and caching are required. Sleep mode is only woken up after the activation clock is invoked, and the node consumes the least energy. For these three modes, the network energy consumption per unit of time is E, respectively. A E S and E D (E A >E S >E D ).
[0099] The processing flow of a node receiving an interest packet is as follows: Figure 1 As shown, after a user initiates a request (interest packet), when the interest packet is transmitted to a node, the node first queries its cached content library to see if the local cache contains content that satisfies the request. If not, it continues to query the link routing table. If the link routing table has a record for the request (interest packet), it determines whether the request originates from a content source node within the domain. If so, the request is forwarded to the content source node within the domain. Otherwise, based on the mapped virtual node information, a cross-domain connection is established, and the request is forwarded to a content source node outside the domain through the virtual node. If there is no record for the content name, the request is discarded and a request failure message is returned. As content is pushed and responded to within the network, the node updates its cached routing table and gradually updates its local cache according to the caching policy, and then sends update announcements to neighboring nodes.
[0100] Specifically, in step S2, when a requesting node sends an interest packet about a certain request to the first receiving node closest to the requesting node, it queries whether the cache content library of the first receiving node stores a data packet that matches the interest packet; if so, it returns the data packet to the requesting node; otherwise, it jumps to step S3.
[0101] S3. Check if the cached routing table of the first receiving node has a record of a node corresponding to the data packet of the interest packet. If yes, proceed to S4; otherwise, discard the interest packet and proceed to S7.
[0102] S4. Determine whether the data packet is in any node other than the first receiving node in the first autonomous system network where the first receiving node is located. If yes, define the node containing the data packet as the first data production node and jump to S5; otherwise, jump to S6.
[0103] S5. Calculate the first cooperative cache range of the first data production node in the first autonomous system network based on energy consumption and latency factors, and define the nodes participating in cooperation within and outside the first cooperative cache range as the first cooperative node and the second cooperative node, respectively. Wake up the first data production node to put the first data production node into active mode, and put the first cooperative node, or the first cooperative node and the second cooperative node into standby mode. Cache the data packets generated by the first data production node according to the actual situation and forward them to the requesting node through the first cooperative node, or the first cooperative node and the second cooperative node.
[0104] In this embodiment, in S5, the specific process by which the first cooperating node, or the first and second cooperating nodes, cache and forward the data packet to the requesting node according to the actual situation is as follows:
[0105] B1. The first cooperating node, or the first cooperating node and the second cooperating node, receive the data packet;
[0106] B2. Determine whether the data packet is already cached in its own cache content library. If so, discard and forward the data packet; otherwise, jump to B3.
[0107] B3. Determine whether to cache the data packet based on the caching decision. If yes, proceed to B4; otherwise, discard and forward the data packet.
[0108] B4. Update the cached routing table of the first cooperating node, or the first and second cooperating nodes themselves, and send cache update information to the neighboring nodes of the first cooperating node, or the first and second cooperating nodes.
[0109] B5. When the neighbor node receives a cache update message from the first cooperating node, or the first cooperating node and the second cooperating node, it queries its own cache routing table to see if there is a record of the cache update message from the first cooperating node, or the first cooperating node and the second cooperating node. If so, the cache update information is discarded; otherwise, the neighbor node updates its own cache routing table.
[0110] like Figure 2The diagram illustrates the collaborative caching process for each node. After a ground node or satellite node generates data (i.e., after the data generating node generates data packets), the corresponding collaborative caching range hop count H is first calculated based on energy consumption and latency factors. * Then, the data packet is pushed to the corresponding collaborating node within the domain. At this point, the collaborating node checks whether the data packet is already cached in its own cache content library. If it is, it discards the data packet; otherwise, it decides whether to cache it based on the caching decision. If the data packet is cached, it updates its own cache routing table and sends cache update information to neighboring nodes; otherwise, it discards the interest packet and the data packet. When a neighboring node receives the cache update message, it checks whether there is a record in its own cache routing table. If there is a record, it discards the cache update information; otherwise, it updates its cache routing table.
[0111] The energy consumption of satellite networks for collaborative caching is as follows Figure 4 As shown, the total energy consumption includes energy used for collaboration and energy used for normal data retrieval. Without loss of generality, it is assumed that the number of satellites in the region is M within the time period t. t The four stages of a node and their energy consumption are described below:
[0112] 1) Active Phase. Satellite nodes that generate data operate and perform tasks in active mode, while other satellite nodes operate in standby mode.
[0113] E C1 =E A +E S (M t -1) (1)
[0114] Among them, E A E represents the energy consumption of a node in active mode. S M represents the power consumption of the node in standby mode. t The number of nodes in the first autonomous domain network;
[0115] 2) Cooperation Phase: The satellite node that generates data sends the data packet to the cooperating node at a distance of H hops. Therefore, given a maximum cooperative buffer range of H hops, the number of active satellite nodes within this maximum cooperative buffer range is Θ. Η The number of nodes that are not H-hops (greater than H-hops) is Θ. Η -Θ Η-1 Assuming the probability of a node at a non-H hop participating in cooperation is P, then the number of nodes is M. C =P(Θ) Η -Θ Η-1 When non-H-hop nodes participate in cooperation, all nodes within the domain need to switch to active mode to forward and cache data. The energy consumption can be expressed as:
[0116] EC2 =E A Θ Η +E S (M t -Θ Η (2)
[0117] Where H is the maximum collaborative cache range of the data production node, and Θ Η This represents the number of all nodes in active mode within the maximum collaborative cache range.
[0118] 3) Waiting Phase: Nodes participating in collaborative caching are in standby mode, while other nodes within the domain are in sleep mode. The number of sleeping nodes is Θ. Η -M C =Θ Η -(Θ Η -Θ Η-1 )P
[0119] E C3 =(t wait -1)(E D (Θ Η -M C )+E S (M t -Θ Η +M C (3) where Θ Η -M C =Θ Η -(Θ Η -Θ Η-1 P represents the number of sleeping nodes in the first autonomous region network that are in sleep mode, and t represents the number of sleeping nodes in sleep mode. wait The total duration of the collaboration phase and the waiting phase for the collaborating node, E D M represents the energy consumption of a sleeping node in sleep mode within the first autonomous region network. C This represents the number of nodes participating in collaboration outside the maximum collaborative cache range, i.e., nodes that are not at the H-hop point participating in collaboration.
[0120] M C =P(Θ) Η -Θ Η-1 );
[0121] 4) Response Phase: When a request is sent, the forwarding nodes and data providing nodes on the routing path are awakened to forward the data packet. Since the cooperating nodes at non-H hops satisfy the following condition: assuming k is the number of hops from the external (requesting node) to the data producing node, the path distance from the user to the cooperating node is k' = k - H.
[0122] E C4 =E A k'+E S (Mt -Θ Η +M C -k')+E D (Θ Η -M C (4)
[0123] Since k is the number of hops, and one hop represents one node, k can represent the number of cooperating nodes.
[0124] From the above, the total energy consumption from data generation to data retrieval (the entire process from data generation to satisfying the interest packet, which is the data generation node forwarding the data packet to the requesting node) can be expressed as:
[0125]
[0126] However, since some nodes along the path may be in a dormant state, requests may not receive a response, therefore, it cannot be guaranteed that all participating collaborative nodes will be able to fulfill the request. If this happens, the user must resend the data request. Generally, the larger the scope of collaboration, the less likely collaborative nodes are to fulfill future requests. This is because of the randomness of request generation and delivery to content producers. The farther away the collaborative node is, the more likely the request may have already been received by the content producer, making it less likely that the collaborative node will receive subsequent requests. Therefore, the relationship between data availability and the distance between collaborative nodes and producer nodes is represented as Y = X(H). Without losing generality, we take Y... Η =(1-γ) Η , where γ∈[0,1) is the risk coefficient of cooperation. If the above situation occurs, the request must wait for the sleeping node to be woken up and directed to the data producer. Therefore, the total energy consumption of multi-hop cooperative caching is
[0127] E(H)=E SUM_C +(1-(1-γ) Η E A k (6)
[0128] Among them, E SUM_C Let E be the total energy consumption of all nodes in the first autonomous system network from data generation to data retrieval; γ∈[0,1), is the risk coefficient for cooperation between nodes; H is the maximum cooperative buffer range, i.e., the data producing node forwards data packets / interest packets to cooperative nodes within a distance of H hops, where the cooperative nodes are nodes in standby mode; A The energy consumption of the data producer node in active mode, where k is the number of hops from the request node to the data producer node;
[0129] Furthermore, it is assumed that the average delay in both non-collaborative and collaborative scenarios is T. NON =Tk,T C=Tk', where T is the average transmission delay over one hop distance. Given the sleep schedule time T D If the request cannot be satisfied through collaborative caching, the latency will be T. NON +T D This is because it must wait for the producer to wake up from sleep mode. Therefore, the total latency reduction can be derived as follows:
[0130]
[0131] Among them, T NON =TK, representing the average latency time when nodes in the first autonomous system network do not cooperate, T C =TK′, representing the average latency time when nodes cooperate in the first autonomous system network, where T is the average latency of one hop distance for a node. D Given a sleep schedule duration, K is the number of hops from the requesting node to the data producing node, and K′ is the path distance from the requesting node to the collaborating node, where K′ = KH; Y H =(1-γ) H γ∈[0,1), is the risk coefficient for cooperation;
[0132] The goal of collaborative caching is to reduce energy consumption and data retrieval latency by finding the optimal range; therefore, the optimal collaborative cache range can be expressed as:
[0133]
[0134] in and These are the weights for energy saving and latency reduction, respectively. For simplicity, we take...
[0135] Collaborative cache range calculation:
[0136] Given a risk coefficient γ, eigenvector centrality is introduced to derive an optimal cooperative caching range. Eigenvector centrality is a classic measure of node influence in a network, reflecting the importance of a node's neighborhood; that is, the more neighbors a node's neighbors can connect to, the higher the eigenvector centrality. If a node's neighbors have more connections to the outside world, the node is more likely to receive more requests, and therefore, it can cooperate with its neighbors with lower risk. Assuming the network topology of a satellite network is G=(V,E), the eigenvector centrality of node v∈V can be expressed as:
[0137]
[0138] Where N(v) is a set of neighbors (the set of adjacent nodes) of node v, obtained through a routing protocol; G is the first autonomous system network. If a node's region is closely connected to potential consumers, the risk of it performing cache cooperation will be lower. Therefore, eigenvector centrality can be used to estimate the risk coefficient, which can be calculated using the following formula:
[0139] Z is a constant (10)
[0140] In summary, the scope of cooperation is H. * It can be determined by the eigenvector centrality, as shown below:
[0141]
[0142] The centrality of an eigenvector is calculated or approximated using a power-law iterative eigenvalue algorithm, followed by the calculation of its range. This range can be calculated offline or inferred from a coarse calculation of the centrality, reducing the computational complexity of calculating the eigenvector centrality.
[0143] Given a collaboration scope H * At that time, each H * The hopping node decides whether to cache the collaborating node based on its own three dimensions of status: 1) remaining energy ratio; 2) cache freshness; 3) cache popularity. That is, the caching decision specifically includes the node's remaining energy ratio, cache freshness, and cache popularity.
[0144] 1) Residual energy ratio: B energy ∈(0,1), the ratio of the remaining energy to the total energy in a node.
[0145] 2) Cache Freshness: A measure of the freshness of cached content, denoted by time(c) representing the time the content has been stored in node v, |C v_ability | indicates the caching capacity of the current node. Considering the data lifecycle of the perception task, nodes that cache stale content are more likely to act as collaborating nodes to store newly generated, more valuable data.
[0146]
[0147] 3) Cache popularity: Considering the request's preference for different content, the node will make caching decisions based on the content popularity.
[0148]
[0149] in I represents the total number of requests received by all nodes within the domain. v This represents the number of requests received by node v. The collaborative caching decision is based on these three parameters, and H...* The number of nodes that jump can become collaborators is P. v ∈[0,1], P v The calculation is as follows:
[0150] P v =ω1B energy (v)+ω2F fresh (v)+ω3Q v (14)
[0151] ω1, ω2, and ω3 are the adjustment parameters for the caching probability.
[0152] S6. Based on the virtual node information mapped by the first autonomous system network and the second autonomous system network where the node containing the data packet corresponding to the interest packet is located, establish a cross-domain connection and construct a corresponding auxiliary graph, find the optimal cross-domain connection node, and forward the interest packet to the second data production node of the second autonomous system network through the optimal cross-domain connection node. The second data production node refers to the steps of S5 to find the corresponding data packet and caches the data packet according to the actual situation and forwards it to the request node.
[0153] Each region uses a unified network standard to manage and maintain domain link routing tables within the region, accelerating route convergence. Network dynamism is masked by virtual nodes; specifically, a region can have multiple virtual nodes. As satellites move, the virtual node constellation is replaced by the satellite closest to its logical location within the domain, enabling inter-domain routing information exchange. When cross-domain data transmission is required, inter-domain data forwarding is completed based on the virtual node-domain link routing table.
[0154] like Figure 5 As shown, the regional network is identified by a globally unique Geographic Area Identifier (GID). The GID consists of the region's longitude (Lo), latitude (La), and radius (R), named according to the rule (Lo-La-R), such as: urban areas (60-50-150), desert areas (90-45-100), and ocean areas (20-15-120). R is determined by factors ρ such as application demand density and geographical location, i.e., R∝Δ / ρ, Δ>0, where Δ is an adjustment parameter. Specifically, sparsely distributed terminals have smaller coverage areas, while densely distributed terminals have larger coverage areas. According to the regional naming rule, the logical location of a virtual node can be named (Lo-La-R, Vir...). i For example, a virtual node in a desert autonomous region is represented as (90-45-100,2).
[0155] In this embodiment, in S6, the specific process of establishing a cross-domain connection based on the virtual node information mapped by the first autonomous system network and the second autonomous system network where the node containing the data packet corresponding to the interest packet is located, and forwarding the interest packet to the second data production node of the second autonomous system network through the virtual node is as follows:
[0156] S6.1 Based on the domain information of the autonomous system network, the first cooperative node in the first autonomous system network and the second cooperative node in the second autonomous system network are copied as the first sub-cooperative node and the second sub-cooperative node, respectively.
[0157] S6.2. Find the physical links between each first cooperating node, between each second cooperating node, and between each first cooperating node and second cooperating node through the domain link routing table;
[0158] S6.3. Set up the first virtual node and the second virtual node in the first autonomous system network and the second autonomous system network, and find the first sub-cooperative node and the second sub-cooperative node that can be used as the first virtual node and the second virtual node respectively by hash mapping method, and connect the corresponding virtual node and sub-cooperative node with virtual links to complete the construction of the auxiliary graph.
[0159] S6.4. Set the link cost and use the shortest path algorithm to obtain the routing information from the first virtual node to the second virtual node from the auxiliary graph, thereby finding the first and second sub-cooperative nodes with the lowest corresponding link costs as the optimal cross-domain connection nodes. After setting the link cost, the link with the lowest link cost can be obtained according to the shortest path algorithm based on link state (link cost).
[0160] Because a virtual node mapping method is used, inter-domain communication is actually completed by the satellite closest to the virtual node. Therefore, inter-domain routing path selection can be transformed into routing path selection between virtual nodes. Thus, to select the optimal replacement satellite and ensure inter-domain information transmission, the original topology can be transformed into an auxiliary graph by constructing a virtual link with two virtual nodes at its ends. A design example is provided. Figure 6 As shown, Figure 6 The left side shows the original satellite network topology. Figure 6 The right side shows the auxiliary diagram after the original topology transformation.
[0161] First, based on the domain information, node N in the original network is copied into N' and N'', such as node S1 becoming S'1, S5 becoming S''5, etc. Then, S'' is copied into S''. m To S n The physical link is S' m →S' n S' m →S″ n and S″m →S″ n In addition, dashed links are added to indicate that the corresponding node is a virtual node (Vir). i and Vir j Candidates, such as Figure 6 As shown on the right, the virtual node Vir i The candidate nodes are S′1, S′3, and S″6; Vir i and Vir j It can be node S′3, etc. After all nodes and links are constructed, link costs are set in the auxiliary graph based on the node and link configurations. Finally, when the auxiliary graph is complete, the shortest path algorithm is used to obtain Vir from the auxiliary graph. i To Vir j Routing information.
[0162] In this embodiment, the link cost includes:
[0163] The cost of the solid line link is calculated as follows:
[0164]
[0165] β+δ=1; (15)
[0166]
[0167] In formula (15), This indicates the transmission delay between two nodes; This represents the energy consumption required to transmit data between two nodes. β and δ are adaptive factors for link weight, which can be adjusted according to the actual situation of the link. The physical link is S' m →S' n S' m →S″ n The solid link cost, S′ m To copy S m The obtained node, S″ m To copy S′ m The obtained node, S′ n To copy S n The obtained node, S″ n To copy S′ n The obtained node.
[0168] The dashed link reflects the importance of a node. The link cost is measured by the normalized cost of node centrality and content service capability, calculated as follows:
[0169]
[0170] α∈[0,1]; (17)
[0171]
[0172] In equation (17), This reflects the centrality of the nodes. Represents node S m To S n The total number of all paths Represents node S m To S n Passing through node S i The number of paths. It describes the node's content service capabilities, that is, the degree to which the node is successful in acquiring specific data. and Representing node S respectively i The total number of forwarded requests and the number of responses received, where Sm is the node that initiated the request, Sn is the data-producing node that caches data packets, and S... i From S m To S n The intermediate nodes on this path are the collaborating nodes. After setting the link cost, the shortest path algorithm based on link state (link cost) can be used to find the link with the minimum link cost, thereby identifying the first and third sub-collaborating nodes. These are virtual nodes Vir i →S′ i Virtual node Vir i →S″ i S′ i →Virtual Node Vir j S″ i →Virtual Node Vir j Virtual link cost, Vir i For the virtual node corresponding to node N′ in the network, Vir j The virtual node corresponding to node N″ in the network.
[0173] During packet return and cooperation, nodes cache packets based on random cooperative caching decisions. Therefore, to obtain real-time caching information within the network, adjacent nodes need to periodically exchange state information, allowing nodes to dynamically adjust routing decisions when network states change. State-aware algorithms are effective at acquiring state information between nodes in smaller networks with low update overhead. Therefore, after a node caches a packet, it floods the updated information to adjacent nodes using a state-aware algorithm, achieving rapid convergence of network packet caching information. Cache updates mainly occur under the following conditions, for example... Figure 7 As shown:
[0174] 1) Cache of nodes in the data packet return path. In this case, since the data packet is typically cached at the user (requesting node) edge, i.e., a node one hop away from the user, cache updates occur within one hop. For example, user u... i After obtaining content M1, the content is cached in the user's access node. At this time, the node will send a cache update announcement to neighboring nodes.
[0175] 2) Satellite nodes (data generation nodes) generate real-time content and push it to the cache within the LEO network. For example, after a node generates observation data M6, it calculates the optimal cooperative cache range H. * Then push the content to H * Jump node, H * The hop node decides whether to cache content based on the caching policy. After caching, it notifies neighboring nodes of the cached information, thus converging the routing information.
[0176] 3) Cache push from the terrestrial network to the LEO network. This is similar to 2), where the content server pushes content M5 to the optimal cache collaboration node and then notifies neighboring nodes of the cache update information.
[0177] After the aforementioned collaborative caching and updates are completed, nodes not participating in the collaboration can periodically execute a sleep schedule to reduce energy consumption. Nodes participating in the collaboration, after caching the content, only need to remain in a waiting state (i.e., standby state) to fulfill subsequent requests. (Scheduling phase reference) Figure 4 .
[0178] To speed up the search for real satellite nodes through virtual nodes, i.e., the mapping between virtual and real satellite nodes, the search for the real satellite node is determined by looking up the index of the virtual node, and a hash mapping method is used to allocate memory addresses. Figure 8 The data structure is displayed as a virtual node table design, where Vir m This indicates the location number of the virtual node; the memory addresses in Entry1 and Entry2 are respectively Vir. n and Vir i For ease of representation, virtual nodes (Lo-La-R, Vir) will be used. i Replace with Vir i The index i in the array is used. This storage method generates an array to store the memory address of each virtual node and uses hash operations to reduce search complexity.
[0179] First, the virtual node index is input into a hash function. To restrict the output hash value h to a range of array addresses, h is subjected to a ⊕ (exclusive OR) operation with respect to the array size. This operation, in most cases, achieves a one-to-one mapping between each virtual node index and an array entry. However, hash collisions may occur, where identical h's will map to the same array item. In such cases, the network Vir... i and Vir m Having the same h', then as Figure 8 The link routing table shown is stored as follows: Viri and Virm. Virm is located at the next position in the link routing table, and so on. The table data mainly records the actual replacement of satellite nodes by virtual nodes, as well as their start and end times, and stores the time items using a binary tree.
[0180] S7, End.
[0181] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. 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 low-Earth orbit satellite network energy-saving collaborative buffering and forwarding method centered on information, characterized in that, include: S1. Based on geographical scenarios and service requirements, the low-orbit satellite network is divided into multiple different autonomous domain networks, each of which includes multiple nodes; S2. When a requesting node sends an interest packet about a certain request to the first receiving node closest to the requesting node, it queries whether the cache content library of the first receiving node stores a data packet that matches the interest packet; if so, it returns the data packet to the requesting node; otherwise, it jumps to S3. S3. Check if the cached routing table of the first receiving node has a record of a node corresponding to the data packet of the interest packet. If yes, proceed to S4; otherwise, discard the interest packet and proceed to S7. S4. Determine whether the data packet is in any node other than the first receiving node in the first autonomous system network where the first receiving node is located. If yes, define the node containing the data packet as the first data production node and jump to S5; otherwise, jump to S6. S5. Calculate the first cooperative cache range of the first data production node in the first autonomous system network based on energy consumption and latency factors, and define the nodes participating in cooperation within and outside the first cooperative cache range as the first cooperative node and the second cooperative node, respectively. Wake up the first data production node to put it into active mode, and put the first cooperative node, or the first cooperative node and the second cooperative node, into standby mode. When caching and / or forwarding operations need to be performed, switch to active mode, and cache and forward the data packets generated by the first data production node to the requesting node according to the actual situation through the first cooperative node, or the first cooperative node and the second cooperative node. In S5, the specific process of calculating the cooperative caching range of the first data production node in the first autonomous region network based on energy consumption and latency factors is as follows: A1. Calculate the total energy consumption of multi-hop cooperative caching for all nodes in the first autonomous system network. The specific formula is as follows: ; in, The total energy consumption of all nodes in the first autonomous domain network from data generation to data retrieval; H represents the risk coefficient for cooperation between nodes, and H is the maximum cooperative buffer range. That is, the first data production node forwards the data packet to the first cooperative node that is H hops away, which means the first data production node forwards the data packet to the first cooperative node within the first cooperative buffer range. The first cooperative node is the node in standby mode. The energy consumption of the first data production node in active mode, where k is the number of hops from the requesting node to the first data production node; A2. Calculate the total latency in the first autonomous system network using the following formula: ; in, , representing the average latency when nodes in the first autonomous system network do not cooperate. Let represent the average latency when nodes cooperate in the first autonomous system network, and T be the average latency for a node's one-hop distance. Given a sleep schedule duration, K is the number of hops from the requesting node to the first data producing node. This is the path distance from the requesting node to the first collaborating node. ; , This represents the risk factor for cooperation. A3. Design a cooperative caching objective function based on the total energy consumption and the total latency. : ; in, and These are the weights of total energy consumption and total latency, respectively. A4. Substitute the eigenvector centrality into the cooperative caching objective function and calculate the cooperative caching range; The specific steps for A4 are as follows: A4.1 Using eigenvector centrality to assess risk coefficient Then the risk coefficient Substitute into the cache target function The following formula is obtained: ; ; ; It is a node The set of adjacent nodes is obtained through a routing protocol; G is the first autonomous system network; A4.2 Calculate the optimal collaborative cache range by using the power-law iterative eigenvalue algorithm to calculate the collaborative cache objective function using eigenvector centrality in A4.1; S6. Based on the virtual node information mapped by the first autonomous system network and the second autonomous system network where the node containing the data packet corresponding to the interest packet is located, establish a cross-domain connection and construct a corresponding auxiliary graph, find the optimal cross-domain connection node, and forward the interest packet to the second data production node of the second autonomous system network through the optimal cross-domain connection node. The second data production node refers to the steps of S1-S5 to find the corresponding data packet and caches the data packet according to the actual situation and forwards it to the request node. S7, End.
2. The energy-saving collaborative buffering and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 1, is characterized in that... The node has three operating modes: active, standby, and sleep. When the node is in active mode, it acts as a data producer, performing data collection, receiving, and forwarding of interest packets and data packets, while also caching cached information from other nodes. When the node is in standby mode, it can receive interest packets and data packets. Based on caching decisions, the node switches to active mode when it needs to perform forwarding and / or caching operations. When the node is in sleep mode, it does not perform any operations and can be activated and woken up by calling an activation clock.
3. The energy-saving collaborative buffering and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 1, is characterized in that... The total energy consumption of all nodes in the first autonomous system network during the time period t from data generation to data retrieval. Specifically, this includes energy consumption during the activity phase. Energy consumption during the collaboration phase Energy consumption during the waiting phase Energy consumption during the reaction stage ,in, Energy consumption during the activity phase The specific formula is: ; in, This represents the energy consumption of a node in active mode. This represents the energy consumption of the node in standby mode. The number of nodes in the first autonomous domain network; Energy consumption during the collaboration phase The specific formula is: ; Where H represents the maximum collaborative cache range of the data production node, i.e., the first collaborative cache range. This represents the number of all nodes in active mode within the maximum collaborative cache range. Energy consumption during the waiting phase The specific formula is: ; in, , representing the number of nodes in sleep mode in the first autonomous region network. This represents the total duration of the collaboration phase and the waiting phase for both the first and second collaborating nodes. This represents the energy consumption of nodes in sleep mode within the first autonomous region network. This refers to the number of participating nodes outside the first collaborative cache range, i.e., non- The number of nodes participating in the collaboration at the jump point, i.e., the number of second collaborating nodes. ; Energy consumption during the reaction stage The specific formula is: 。 4. The energy-saving collaborative buffering and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 1, is characterized in that... In S5, the specific process by which the first cooperating node, or the first and second cooperating nodes, cache and forward the data packet to the requesting node according to the actual situation is as follows: B1. The first cooperating node, or the first cooperating node and the second cooperating node, receive the data packet; B2. Determine whether the data packet is already cached in its own cache content library. If so, discard and forward the data packet; otherwise, jump to B3. B3. Determine whether to cache the data packet based on the caching decision. If yes, proceed to B4; otherwise, discard and forward the data packet. B4. Update the cached routing table of the first cooperating node, or the first and second cooperating nodes themselves, and send cache update information to the neighboring nodes of the first cooperating node, or the first and second cooperating nodes. B5. When the neighbor node receives a cache update message from the first cooperating node, or the first cooperating node and the second cooperating node, it queries its own cache routing table to see if there is a record of the cache update message from the first cooperating node, or the first cooperating node and the second cooperating node. If so, the cache update information is discarded; otherwise, the neighbor node updates its own cache routing table.
5. The energy-saving cooperative caching and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 4, is characterized in that... The caching decision specifically includes the node's remaining energy ratio, cache freshness, and cache popularity, wherein... The remaining energy ratio , which represents the ratio of the remaining energy in a node to the total energy of that node; The cache freshness refers to a measure of the freshness of cached content, and the specific formula is as follows: ; in, This indicates the caching capacity of the current node, and time(c) represents the time the corresponding cached content has been stored in the node. The cache popularity is determined by considering the interest package's preference for different content. The node will make caching decisions based on the content popularity, using the following formula: ; in, This represents the total number of requests received by all nodes within the autonomous system network. This represents the number of requests received by node v. The caching decision The calculation formula is as follows: ; in, It is a parameter for adjusting the likelihood of caching.
6. The energy-saving cooperative caching and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 1, is characterized in that... In S6, based on the virtual node information mapped by the first autonomous system network and the second data production node containing the data packet corresponding to the interest packet, a cross-domain connection is established and a corresponding auxiliary graph is constructed. The optimal cross-domain connection node is found, the second data production node is woken up, and the interest packet is forwarded to the second data production node of the second autonomous system network through the optimal cross-domain connection node. The specific process is as follows: S6.1 Based on the domain information of the autonomous system network, the first cooperative node in the first autonomous system network and the third cooperative node in the second autonomous system network are copied as the first sub-cooperative node and the third sub-cooperative node, respectively. S6.
2. Find the physical links between each first cooperating node, between each third cooperating node, and between each first cooperating node and a third cooperating node through the domain link routing table; S6.
3. Set up the first virtual node and the second virtual node in the first autonomous system network and the second autonomous system network, and find the first sub-cooperative node and the third sub-cooperative node that can be used as the first virtual node and the second virtual node respectively by hash mapping. Connect the corresponding virtual node and sub-cooperative node with virtual links to complete the construction of the auxiliary graph. S6.4 Set the link cost and use the shortest path algorithm to obtain the routing information from the first virtual node to the second virtual node from the auxiliary graph, thereby finding the first sub-cooperative node and the third sub-cooperative node with the minimum corresponding link cost as the optimal cross-domain connection nodes.
7. The energy-saving cooperative caching and forwarding method for low-Earth orbit satellite networks centered on information, as described in claim 6, is characterized in that... There is at least one first virtual node located in the first autonomous system network, and at least one second virtual node located in the second autonomous system network.
8. A low-Earth orbit satellite network energy-saving cooperative caching and forwarding method centered on information, as described in claim 6, is characterized in that... The link cost includes: Actual physical link cost: ; ; ;in, This indicates the transmission delay between two nodes; This represents the energy consumption required to transmit data between two nodes. and This is a link weight adaptive factor that can be adjusted according to the actual situation of the link. The physical links are respectively , The cost of the solid line link, For copying The obtained nodes, For copying The obtained nodes, For copying The obtained nodes, For copying The obtained node; Virtual link cost: ; ; ; in, Virtual nodes Virtual nodes , , The cost of virtual links, Represents a node centrality, Represents a node To the node The total number of all paths Represents a node To the node Passing through the node The number of paths, Represents a node Content service capability, that is, the degree to which a node is successful in acquiring specific data. and Representing nodes respectively The total number of forwarded requests and the number of responses received, where Sm is the node that initiated the request, Sn is the data-producing node that caches data packets, and Si is the intermediate node (cooperating node) on the path from Sm to Sn. for The copied node, for The copied node.
Citation Information
Patent Citations
Spatial information network cache decision-making method
CN114759970A
Medical data caching method based on named data network
CN116455978A
