A method for optimizing the deployment of service function chains in a space-ground integrated network

Abstract

This invention discloses a service function chain (SFC) deployment optimization method in a satellite-terrestrial converged network. The optimization method includes: constructing an optimization problem with constraints such as QoS requirements, latency, computation, and communication resources of service requests, aiming to maximize total service profit; solving the optimization problem by applying VNF node placement and inter-VNF routing strategies to each service function chain request, thereby obtaining an optimized deployment scheme for the satellite-terrestrial converged network service function chain. This invention decouples user-requested services into several VNFs, which are arranged in a specified order to form a service function chain. By creating multiple SFC requests and sharing the underlying network to support different applications, this method achieves heterogeneous management and collaboration between satellite and terrestrial networks, effectively improving the problem of not being able to obtain optimal node mapping and link routing decisions in heterogeneous and complex network scenarios.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a method for optimizing the deployment of service function chains in a space-ground converged network. Background Technology

[0002] With the development of internet technology, new applications are emerging in many fields, placing more stringent demands on network infrastructure, including coverage, seamless access, latency sensitivity, and computing and storage resources. While terrestrial networks are the primary communication infrastructure, they suffer from coverage gaps, with data leaks occurring in disaster relief sites, remote mountainous areas, or marine environments. Fortunately, Low Earth Orbit (LEO) satellite networks offer lower latency and global coverage, providing data storage, computing, and communication services to remote users not covered by terrestrial networks. However, due to the physical limitations (size and energy) of satellite nodes, LEO satellites are less capable than terrestrial networks in terms of computing and storage capabilities. Nevertheless, satellite nodes offer wider coverage, as fewer satellite nodes are needed to reach the target node. This characteristic reduces transmission hops and significantly saves bandwidth resources, while terrestrial networks are better suited for handling computationally intensive tasks. Based on the different characteristics of satellites and ground nodes, combining the two enables heterogeneous satellite-ground networks to collaborate and fully leverage their complementary advantages to meet the different Quality of Service (QoS) requirements of new applications. Therefore, Satellite Ground Integrated Network (SGIN) is regarded as one of the most effective architectures for future wireless communication networks.

[0003] However, SGIN involves both satellite and terrestrial networks, which operate independently and are built for specific purposes and missions. They are heterogeneous and lack a collaborative mechanism, making it very difficult to efficiently coordinate satellite-terrestrial network resources. Summary of the Invention

[0004] To address the lack of collaboration mechanisms between existing satellite and terrestrial networks, which hinders efficient coordination of satellite-terrestrial integrated network resources, this invention proposes a service function chain deployment optimization method in satellite-terrestrial integrated networks. This invention, based on a terrestrial network, introduces a satellite network to form a heterogeneous satellite-terrestrial integrated network. Furthermore, by dynamically calling corresponding node resources to place Virtual Functions (VNFs), the total service profit can be maximized while meeting the constraints of QoS requirements, latency, computation, and communication resources for service requests.

[0005] This invention is achieved through the following technical solution:

[0006] A method for optimizing the deployment of service function chains in a space-ground converged network, the method comprising:

[0007] Construct an optimization problem with constraints such as QoS requirements of service requests, latency, computation and communication resources, and the goal of maximizing total service profit;

[0008] The optimization problem is solved by applying VNF node placement and inter-VNF routing strategies to each service function chain request, resulting in an optimized deployment scheme for the satellite-ground converged network service function chain.

[0009] This invention decouples user-requested services into several Virtual Function Networks (VNFs). These VNFs are arranged in a specified order and form a Service Function Chain (SFC). By creating multiple SFC requests and sharing the underlying network to support different applications, this invention enables heterogeneous management and collaboration between satellite and terrestrial networks, effectively improving the problem of not being able to obtain optimal node mapping and link routing decisions in heterogeneous and complex network scenarios.

[0010] As a preferred embodiment, the optimization problem of the present invention is expressed as:

[0011]

[0012]

[0013]

[0014]

[0015]

[0016]

[0017] n∈1,2,…,N,m∈1,2,…,M,(n,m)∈1,2,…,E,q∈1,2,…,Q

[0018] The space-ground integrated network is represented as an undirected graph G(N,E), where N and E represent the set of nodes and the set of links, respectively; G = G s ∪G g G s G represents satellite networks. g Represents a terrestrial network; C n B represents the computational resource capacity of node n∈N. (n,m) The bandwidth resource capacity of link (n,m)∈E is represented by D; (n,m) D represents the single-step transmission delay; n Indicates processing latency; service function chain requests are represented as an undirected graph Q(N). q E q ,Fq ), N q and E q F represents the set of virtual nodes and the set of virtual links that make up the service function chain request, respectively. q =(f1,...,f F ) indicates the component type of the service function chain request q, s q and d q These represent the start and end points of a service function chain request q, respectively. Represents virtual node n q ∈N q The required computing resources are: Represents virtual link (n q ,m q )∈E q Required bandwidth resources; D q Indicates the delay requirement; x nq ∈[0,1] represents node n q Was the mapping successful? Represents link (n) q ,m q Whether the mapping is successful on (n,m)∈E; z n,f ∈[0,1] indicates whether node n has a VNF of type f, and the maintenance and startup resources required for this VNF are denoted as C. f ;a q ∈[0,1] indicates whether the service function chain request q was successfully served; This represents the computational resource cost consumed at node n∈N; Represents link (n) q ,m q )∈E q The cost of communication resources consumed; U represents the total service profit, R q α represents the benefit derived from the service function chain request q. s and α g These represent the different weights for satellite nodes and ground nodes, respectively.

[0019] In a preferred embodiment, the first constraint of the present invention is a node computing resource constraint: for any physical node, its total computing resource consumption must not exceed its computing resource capacity; the second constraint is a link bandwidth resource constraint: for any physical link, its total bandwidth resource consumption must not exceed its bandwidth resource capacity; the third constraint is a VNF placement constraint: for any successfully served service function chain request q, each VNF contained therein can only be embedded in one physical node; the fourth constraint is a service function chain request latency constraint: service function chain requests must be accepted within the latency requirement; and the fifth constraint is a node traffic conservation constraint: traffic conservation is a necessary condition for successfully constructing a routing path, and for any physical node, the inflow traffic must equal the outflow traffic.

[0020] In a preferred embodiment, the computational resource cost consumed on the node of the present invention is expressed as:

[0021]

[0022] in, The maintenance and startup costs associated with a VNF of type f are represented by the following: The cost incurred by running a VNF of type f.

[0023] In a preferred embodiment, the communication resource cost consumed on the link of the present invention is expressed as:

[0024]

[0025] As a preferred embodiment, the specific process for placing VNF nodes according to the present invention is as follows:

[0026] For the top T% of service requests, first apply the shortest path algorithm to the satellite-ground converged network to find K routable shortest paths. If a shortest path exists, select the one with the highest weight. q If the path is (n, m), the nodes traversed by that path are used as candidate nodes for the current node. If a candidate node is not empty, traverse the VNF required for the service request and select the node with the weight. q (n) The largest node is placed in turn. If the current node does not have enough resources when placing a node, the node is removed from the node candidate domain, and the node with the largest weight is selected again. q (n) Continue placing the largest node until all VNFs have been placed;

[0027] For service requests after T%, first use the shortest path algorithm on the ground network to find K routable shortest paths. If a shortest path exists, select the one with the weight. qFor the path with the highest value (n, m), the adjacent nodes of the nodes traversed by this path are used as the current node candidate domains; if the node candidate domains are not empty, the VNFs required for the service request are traversed, and the weight is selected. q (n) The largest node is placed in turn. If the current node does not have enough resources when placing a node, the node is removed from the node candidate domain, and the weight is selected again. q (n) Continue placing the largest node until all VNFs have been placed; if the service is still unsuccessful when the node candidate field is empty, the node candidate field will be reacquired on the satellite-ground fusion network and the above placement process will be repeated.

[0028] As a preferred embodiment, the inter-VNF routing strategy of the present invention is specifically as follows:

[0029] For each physical node where a VNF has been placed, the shortest path algorithm is applied according to the weight. q (n,m) Select the current optimal path and compare the remaining bandwidth resources and required bandwidth resources of the traversed links. If they are satisfied, select the link to complete the routing between VNF nodes. If they are not satisfied, delete the current link and select the shortest path again. Repeat the above process until the routing is completed. If there is no available path, it is determined that the service request cannot be received.

[0030] If the placement of VNF nodes and routing between VNF nodes have been completed, then check whether the latency has exceeded the latency limit. If it has not exceeded the limit, then the current request is considered to have been successfully received.

[0031] As a preferred embodiment, the weight of the present invention q (n) is represented as:

[0032]

[0033] in, This represents the amount of CPU resources remaining in node n. χ represents the centrality of node n, where χ st (n) represents the number of shortest paths from the starting point to the ending point that pass through node n, x st (n) represents the number of shortest paths from the starting point to the ending point; Denotes the aggregation degree of node n, where A n R represents the number of nodes reachable from node n. n This represents the sum of distances from node n to all reachable nodes; This represents the degree of sharing of node n.

[0034] As a preferred embodiment, the weight of the present invention q (n,m) is represented as:

[0035]

[0036] in, and These are the remaining CPU resources of node n and the remaining bandwidth resources of link (n,m), respectively. It represents the amount of CPU resources remaining in node m. It is the degree of sharing of node n; ρ is the degree of node m; C ρ B To calculate the weights and bandwidth weights.

[0037] As a preferred embodiment, the space-ground integrated network of the present invention includes a terrestrial network and a satellite network;

[0038] The terrestrial network and the satellite network are connected via a satellite-to-ground link;

[0039] The ground network includes several server ground nodes, and the several server ground nodes are connected by ground links.

[0040] The satellite network includes several low-Earth orbit (LEO) satellite nodes, which are connected by star links.

[0041] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0042] 1. This invention maximizes the total service revenue while meeting the constraints of QoS requirements, latency, computation, and communication resources, effectively improving the problem of not being able to obtain optimal node mapping and link routing decisions in heterogeneous and complex network scenarios.

[0043] 2. This invention addresses the scarcity of satellite node resources in space-ground converged networks by introducing VNF instance sharing. It considers reusing the same virtual machine for VNFs of the same type to reduce configuration resource costs and designs a node placement algorithm to maximize the utilization of resource-scarce satellite nodes. Simultaneously, to address the issue of excessive node load caused by VNF reuse, a low-complexity inter-node routing algorithm is designed to obtain the optimal link routing strategy for the system. Combining these two points, the final result is an improvement in the service function chain request reception rate in the space-ground converged network, better ensuring that the space-ground converged network can leverage its advantages of wide-area coverage, high efficiency, and low cost.

[0044] 3. This invention can significantly improve the reception rate of service function chain without ensuring normal reception of service function chain requests, without generating additional computing and communication resource consumption, i.e. without increasing costs, and has strong application value and development potential. Attached Figure Description

[0045] The accompanying drawings, which are included to provide a further understanding of embodiments of the invention and form part of this application, do not constitute a limitation thereof. In the drawings:

[0046] Figure 1 This is a schematic diagram of the hierarchical satellite-ground converged network architecture according to an embodiment of the present invention. Solid lines represent deployed links; dotted lines represent deployable links; and dashed lines represent pre-deployed links.

[0047] Figure 2 This is an example of VNF instance sharing in an embodiment of the present invention.

[0048] Figure 3 This is a graph showing the relationship between the number of service function chains and the service function link yield in an embodiment of the present invention.

[0049] Figure 4 This is a diagram showing the relationship between the number of service function chains and total resource usage in an embodiment of the present invention.

[0050] Figure 5 This is a diagram showing the relationship between the number of service function chains and total profit in an embodiment of the present invention. Detailed Implementation

[0051] In the following, the terms “comprising” or “may include” as used in various embodiments of the invention indicate the presence of an inventive function, operation, or element, and do not limit the addition of one or more functions, operations, or elements. Furthermore, as used in various embodiments of the invention, the terms “comprising,” “having,” and their cognates are intended only to indicate a specific feature, number, step, operation, element, component, or combination of the foregoing, and should not be construed as primarily excluding the presence of one or more other features, numbers, steps, operations, elements, components, or combinations of the foregoing, or adding one or more combinations of the foregoing.

[0052] In various embodiments of the invention, the expression "or" or "at least one of A and / or B" includes any combination or all combinations of the words listed simultaneously. For example, the expression "A or B" or "at least one of A and / or B" may include A, may include B, or may include both A and B.

[0053] The expressions used in the various embodiments of the present invention (such as "first," "second," etc.) may modify various constituent elements in the various embodiments, but do not limit the corresponding constituent elements. For example, the above expressions do not limit the order and / or importance of the elements. The above expressions are only used for the purpose of distinguishing one element from other elements. For example, a first user device and a second user device refer to different user devices, although both are user devices. For example, a first element may be referred to as a second element without departing from the scope of the various embodiments of the present invention, and similarly, a second element may also be referred to as a first element.

[0054] It should be noted that if a description is made of "connecting" one component to another, then the first component can be directly connected to the second component, and a third component can be "connected" between the first and second components. Conversely, when a component is "directly connected" to another component, it can be understood that there is no third component between the first and second components.

[0055] The terminology used in the various embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to limit the various embodiments of the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Unless otherwise defined, all terms used herein (including technical and scientific terms) have the same meaning as commonly understood by one of ordinary skill in the art to which the various embodiments of the invention pertain. The terms (such as those defined in a generally used dictionary) are to be interpreted as having the same meaning as in the context of the relevant technical field and are not to be interpreted as having an idealized or overly formal meaning, unless clearly defined in the various embodiments of the invention.

[0056] 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.

[0057] Example 1:

[0058] This embodiment proposes an optimized deployment method for service function chains in a space-ground integrated network. Based on a terrestrial network, this embodiment introduces a satellite network to form a heterogeneous space-ground integrated network, specifically as follows: Figure 1As shown, the space-ground integrated network architecture includes a terrestrial network, a satellite network, and space-ground links connecting the satellite network and the terrestrial network. The terrestrial network includes several server ground nodes connected by ground links. The satellite network includes several low-Earth orbit satellite nodes connected by satellite links. By dynamically calling the corresponding node resources to place VNFs, the number of successfully served service requests is maximized under the constraints of QoS requirements, latency, computing, and communication resources of service requests.

[0059] This embodiment also proposes a service function chain deployment optimization method in a space-ground integrated network. Its basic working principle is as follows: Figure 1 As shown, the satellite network consists of four satellite nodes, namely N8-N in the diagram. 11 It has a large coverage area, connecting all ground nodes via wireless links. Ground nodes are connected via wired or wireless links. The end-user's task is represented by a service function chain consisting of multiple specific VNFs. Based on NFV technology, these VNFs can be flexibly placed on air and ground nodes. Assuming the service function chain requires 3 VNFs, let's call them VNFs. A -VNF C -VNF E Traffic starts from node N1, flows into node N7, then to node N3, and finally to node N4. The three strategies are represented by solid lines, dotted lines, and dashed lines, respectively. The solid line represents the basic strategy of SFC planning, where VNFs are placed on three terrestrial network nodes, and traffic undergoes three-hop transmission. In contrast, traffic using the dotted line strategy only undergoes two hops, reducing communication resource consumption. The dashed line shows a scenario where terrestrial network computing resources are overloaded. In this case, some VNFs in a new service function chain request are placed on the satellite network, successfully receiving the request with only two hops and significantly reducing hop latency.

[0060] In this embodiment, the space-ground fusion network is represented as an undirected graph G(N,E), where N and E represent the set of nodes and the set of links, respectively. G = G s ∪G g Among them, G s It's a satellite network, G g It is a ground network. For a node n∈N, the computational resource capacity is C. n For a link (n,m)∈E, the bandwidth resource capacity is B. (n,m) Similarly, in this embodiment, the single-step transmission delay is represented as D. (n,m) The processing delay is represented by D. n .

[0061] In this embodiment, a set of service function chain requests is considered, represented as an undirected graph Q(N). q Eq ,F q ), N q and E q F represents the set of virtual nodes and the set of virtual links that make up the service function chain request, respectively. q =(f1,...,f F ) indicates the component type of the service function chain request q, s q and d q These represent the start and end points of a service function chain request q, respectively. For virtual node n... q ∈N q In other words, the required computing resources are For virtual links (n) q ,m q )∈E q In other words, the required bandwidth resources are Meanwhile, in this embodiment, the latency requirement is represented as D. q A service request will only be successfully served if the required bandwidth is allocated to it.

[0062] This example uses Boolean variables. Represents virtual node n q Whether the placement was successful can be determined using a boolean variable. Represents link (n) q ,m q Whether the placement was successful on (n,m)∈E is determined using the Boolean variable z. n,f ∈[0,1] indicates whether node n has a VNF of type f, and the maintenance and startup resources required for this VNF are denoted as C. f Using Boolean variable a q ∈[0,1] indicates whether the service function chain request q was successfully served.

[0063] Under constraints of QoS requirements, latency, computation, and communication resources, the goal is to maximize the number of service requests that can be successfully served to achieve efficient resource utilization. Simultaneously, the cost of using communication and computation resources at air and ground nodes should be minimized. Therefore, the optimization method proposed in this embodiment is based on the difference between the total revenue and total cost of SFC, establishing the following optimization problem:

[0064]

[0065]

[0066]

[0067]

[0068]

[0069]

[0070] n∈1,2,…,N,m∈1,2,…,M,(n,m)∈1,2,…,E,q∈1,2,…,Q

[0071] The first constraint is the node computing resource constraint: for any physical node, its total computing resource consumption must not exceed its computing resource capacity. The second constraint is the link bandwidth resource constraint: for any physical link, its total bandwidth resource consumption must not exceed its bandwidth resource capacity. The third constraint is the VNF placement constraint: for any successfully served service function chain request q, each VNF contained therein can only be embedded in one physical node. The fourth constraint is the service function chain request latency constraint: service function chain requests must be accepted within the latency requirements. The fifth constraint is the node traffic conservation constraint: traffic conservation is a necessary condition for successfully constructing a routing path; for any physical node, the inflow traffic must equal the outflow traffic.

[0072] The goal of this embodiment is to maximize the profit of SFC requests and serve as many SFC requests as possible. In this embodiment, the cost of computing resources consumed on any physical node is denoted as:

[0073]

[0074] in, The maintenance and startup costs associated with a VNF of type f are represented by the following: The cost incurred by running a VNF of type f.

[0075] Similarly, in this embodiment, the cost of communication resources consumed on a physical link is denoted as:

[0076]

[0077] In fact, service requests will be blocked due to two factors: 1) insufficient physical resources; 2) end-to-end latency exceeding the deadline. From a network operations perspective, the goal is to maximize the number of service requests that can be successfully served to achieve the highest profit; at the same time, the cost of using communication and computing resources in air and ground nodes should be minimized. Therefore, this embodiment defines the total service profit as the difference between the total revenue and total cost of the SFC, and the total service profit function is expressed as:

[0078]

[0079] Among them, R q It is the benefit brought by a single service request, α s and αg These represent different weights for satellite nodes and physical nodes, respectively, to distinguish the costs of airborne nodes and physical nodes.

[0080] At the start of each time period, assume a batch of service requests arrive simultaneously, but with different service deadlines. In reality, these requests may arrive at any time during the last reconfiguration period. These arriving requests can be accumulated and processed as a batch, with a latency cap allocated to each service based on latency requirements and arrival times. Furthermore, due to the quasi-stationary nature of satellite networks, the SGIN topology is fixed within each period and can change in different reconfiguration periods. Service processes are not interrupted by reconfiguration, meaning that if a service cannot be completed in the current period, it will continue to provide service in the next reconfiguration period. However, the remaining available resources, including the computing resources of physical nodes and the bandwidth resources of physical links, must be updated. Therefore, considering the current network state comprised of the network topology and remaining available resources, a general one-time optimization problem is considered.

[0081] The above optimization problem is the SFC deployment problem in a space-ground converged network considering VNF instance sharing. It is a difficult NP-hard problem to solve, which can be decoupled into two sub-problems under the SGIN stage: VNF node placement and inter-VNF routing. By decomposing the coupled problem into a two-step optimization problem, VNF instance sharing is introduced when considering the VNF node placement problem. The contradiction between VNF instance sharing and load balancing is resolved by using a node placement factor. This maximizes sharing to improve physical network utilization while ensuring load balancing of nodes and links. In the study of the inter-VNF routing sub-problem, the link routing problem in heterogeneous networks is solved, and latency is guaranteed to accept as many latency-sensitive tasks as possible. Thus, an optimized design scheme for the deployment of service function chains in a space-ground converged network based on VNF instance sharing and network function virtualization is obtained.

[0082] Based on this, when a batch of Service Function Chain (SFC, which decouples the user's requested service into several VNFs, which are arranged in a specified order) requests arrive, the system determines the current network status and records the amount of computing and communication resources required for the service function chain request, as well as the corresponding latency limit; then, it performs VNF node placement and inter-VNF routing strategies for each request.

[0083] The specific process for placing VNF nodes is as follows:

[0084] For the top T% of service requests, the shortest path algorithm is first applied to the satellite-ground converged network to find K routable shortest paths. If a shortest path exists (i.e., K is greater than or equal to 1), then the path with the correct weight is selected. q If the shortest path with the highest weight (n, m) is selected, the nodes traversed by this shortest path with the highest weight are used as the current node candidate domains. If the node candidate domains are not empty, the VNFs required for the service request are traversed, and nodes with weights of the highest weights are selected sequentially. q (n) The largest node is placed in turn. If the current node does not have enough resources when placing a node, the node is removed from the node candidate domain, and the weight is selected again. q (n) The largest node continues to be placed until all VNFs have been placed. For service requests after T%, the shortest path algorithm is first applied to the terrestrial network to find K routable shortest paths. If a shortest path exists, the one with the highest weight is selected. q The path with the highest weight (n, m) is used as the candidate node. The adjacent nodes of the nodes traversed by this shortest path are then selected as the current node candidate domain. Since these nodes are adjacent to the shortest path nodes, their distances are also relatively short. If the candidate node domain is not empty, the VNFs required for the service request are traversed, and nodes with weights of varying values ​​are selected sequentially. q (n) The largest node is placed in turn. If the current node does not have enough resources when placing a node, the node is removed from the node candidate domain, and the weight is selected again. q (n) The largest node continues to be placed until all VNFs have been placed. If service is still unsuccessful when the node candidate domain is empty, a new node candidate domain will be obtained on the satellite-ground fusion network, and the above steps will be repeated. The value of T ranges from 0 to 100, and is set according to the specific application scenario, preferably 50.

[0085] The specific process of routing between VNF nodes is as follows: For each physical node already placed in the VNF, the shortest path algorithm is applied, based on the weight. q (n,m) Select the current optimal path and compare the remaining bandwidth resources of the traversed links with the required bandwidth resources. If they are satisfied, select the link to complete the routing between VNF nodes. If not, delete the current link, reselect the shortest path, and repeat the above steps until routing is complete. If no path is available, determine that the service request cannot be received. If VNF node placement and routing between VNF nodes are both completed, then determine whether the latency has exceeded the latency limit D. q If the time limit is not exceeded, the current request is considered to have been successfully received.

[0086] in,

[0087]

[0088]

[0089] in, and These are the remaining CPU resources of node n and the remaining bandwidth resources of link (n,m), respectively. It represents the amount of CPU resources remaining in node m. χ is the centrality of node n, where χ st (n) is the number of shortest paths from the starting point to the ending point that pass through node n. st (n) The number of shortest paths from the starting point to the ending point; It is the aggregation degree of node n, where A n R is the number of nodes reachable from node n (ignoring n). n It is the sum of the distances from node n to all reachable nodes; It is the degree of sharing of node n. This represents the degree of sharing of node m; It is the degree of node n. ρ is the degree of node m; C ρ B To calculate the weights and bandwidth weights.

[0090] Figure 2 This represents an example of shared Virtual Network Functions instances (VNFs) within this server node. A and VNF B There are two of each, so there are two VNFs. A Sharing a virtual machine, VNF B Same as above. By sharing the basic resources for starting and maintaining VNF instances among SFCs, the basic management resources for starting and maintaining VNF instances on physical nodes are reduced, thereby saving the deployment cost of the corresponding VNF type and increasing resource utilization.

[0091] Example 2:

[0092] This embodiment uses the scheme proposed in Embodiment 1 above, and compares it with Scheme 1, Scheme 2, and Scheme 3 for simulation verification, and obtains... Figures 3-5 The simulation results are shown.

[0093] in, Figure 3This indicates that even with significant resource consumption, the proposed solution in this embodiment maintains an approximately 85% service request acceptance rate when 350-370 service requests arrive, and still maintains a close to 85% acceptance rate when 370-375 service requests arrive. In contrast, Solution 1 (JSAC-SP-FF) only achieves an approximately 80% acceptance rate when 350-360 service requests arrive, and this drops to nearly 77.5% when 360-375 service requests arrive. Solution 2 (TPDS-KSP-FF) shows a linear decline in service request acceptance rate as the number of service requests gradually increases, falling below 80% when 375 service requests arrive. Solution 3 (Greedy-SP-FF) experiences a service request acceptance rate below 70% even with significant resource consumption. Based on these comparisons, it can be demonstrated that the proposed solution in this embodiment far surpasses the compared solutions in terms of resource utilization, exhibiting extremely strong performance.

[0094] Figure 4 This indicates that even with significant resource consumption, the optimization method proposed in this embodiment maintains a high service request reception rate while exhibiting a low growth trend in total resource usage, remaining at a relatively low level. In contrast, while JSAC-SP-FF has a lower service request reception rate than the proposed solution, its total resource usage increases with the number of service requests. In TPDS-KSP-FF, the service request reception rate decreases linearly with the increase in the number of service requests, yet its total resource usage remains high. In Greedy-SP-FF, despite significant resource consumption, the service request reception rate drops below 70%, resulting in a relatively low total resource usage. Based on these comparisons, it can be demonstrated that the proposed solution far surpasses the comparative solutions in terms of resource utilization.

[0095] Figure 5 This indicates that as the number of service requests gradually increases, the total profit obtained by the first scheme, JSAC-SP-FF, is far lower than that of the scheme proposed in this embodiment; the revenue from the service requests received by the second scheme, TPDS-KSP-FF, has become consistent with its cost, and therefore its total profit has stabilized and no longer increases; the third scheme, Greedy-SP-FF, is consistent with the second scheme.

[0096] Despite significant resource consumption, the total profit of the proposed solution in this embodiment still shows an upward trend, and its total profit is far higher than that of comparative solutions one, two, and three. In summary, the solution proposed in this embodiment, under the constraints of basic QoS requirements, latency, and computational and communication resources, can guarantee good load balancing, significantly improve the service function link rate of the space-ground converged network, and achieve a high total service profit, thus possessing great application value.

[0097] 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 method for optimizing the deployment of service function chains in a space-ground converged network, characterized in that, The optimization method includes: Construct an optimization problem with constraints such as QoS requirements of service requests, latency, computation and communication resources, and the goal of maximizing total service profit; By applying VNF node placement and inter-VNF routing strategies to each service function chain request, the optimization problem is solved, resulting in an optimized deployment scheme for the space-ground converged network service function chain. The optimization problem is expressed as: ； In this context, the space-ground fusion network is represented as an undirected graph. , and Represent the set of nodes and the set of links, respectively; , Indicates satellite network, Represents a ground network; Represents a node The computing resource capacity is; Indicates link The bandwidth resource capacity is; Indicates single-step transmission delay; Indicates processing latency; service function chain requests are represented as an undirected graph. , and These represent the set of virtual nodes and the set of virtual links that make up the service function chain request, respectively. Indicates a service function chain request Composition type, and Each represents a service function chain request. The starting point and the ending point; Represents virtual nodes The required computing resources are: Indicates virtual link Required bandwidth resources; Indicates the delay requirement; Represents a node Was the mapping successful? Indicates link Whether the mapping was successful superior; Represents a node Does it have a type of The VNF type is defined, and the maintenance and startup resources required by that VNF ​​type are denoted as... ; Indicates a service function chain request Whether the service was successfully provided; Represents a node The cost of computing resources consumed; Indicates link The cost of communication resources consumed above; Indicates total service profit. Indicates a service function chain request The benefits brought about and These represent the different weights for satellite nodes and ground nodes, respectively. The specific process for placing the VNF node is as follows: For the top T% of service requests, first apply the shortest path algorithm to the satellite-ground fusion network to find K routable shortest paths. If a shortest path exists, select the one with the highest weight. The highest path is used, and the nodes traversed by that path are used as candidate nodes for the current node; if the candidate node is not empty, the VNFs required for the service request are traversed, and the node with the highest weight is selected. The largest node is placed in turn. If the current node does not have sufficient resources when placing a node, the node is removed from the node candidate domain, and the node with the highest weight is selected again. The largest node continues to be placed until all VNFs have been placed. For service requests after T%, first use the shortest path algorithm on the ground network to find K routable shortest paths. If a shortest path exists, select the one with the highest weight. The highest path is used, and the adjacent nodes of the nodes traversed by this path are used as the current node candidate domains; if the node candidate domains are not empty, the VNFs required for the service request are traversed, and the node with the highest weight is selected. The largest node is placed in turn. If the current node does not have sufficient resources when placing a node, the node is removed from the node candidate domain, and the weight is selected again. The largest node continues to be placed until all VNFs have been placed. If service is still unsuccessful when the node candidate domain is empty, a new node candidate domain will be obtained on the satellite-ground fusion network, and the above placement process will be repeated.

2. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, The first constraint is the node computing resource constraint: for any physical node, its total computing resource consumption must not exceed its computing resource capacity. The second constraint is the link bandwidth resource constraint: for any physical link, its total bandwidth resource consumption must not exceed its bandwidth resource capacity. The third constraint is the VNF placement constraint: for any successfully served service function chain request… Each of its VNFs can only be embedded in one physical node; the fourth constraint is the service function chain request latency constraint, which requires that service function chain requests be accepted within the latency requirement; the fifth constraint is the node traffic conservation constraint, which is a necessary condition for successfully building a routing path, meaning that for any physical node, the inflow traffic must equal the outflow traffic.

3. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, The cost of computing resources consumed on a node is expressed as follows: ； in, Representative type is The maintenance and startup costs associated with VNFs Representative type is The cost incurred by running VNF.

4. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, The cost of communication resources consumed on the link is expressed as: 。 5. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, The specific routing strategy between VNFs is as follows: For each physical node where a VNF has been placed, the shortest path algorithm is applied according to the weight. Select the current optimal path and compare the remaining bandwidth resources of the traversed links with the required bandwidth resources. If they are satisfied, select the link to complete the routing between VNF nodes. If they are not satisfied, delete the current link and select the shortest path again. Repeat the above process until the routing is completed. If there is no available path, it is determined that the service request cannot be received. If the placement of VNF nodes and routing between VNF nodes have been completed, then check whether the latency has exceeded the latency limit. If it has not exceeded the limit, then the current request is considered to have been successfully received.

6. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, weight Expressed as: ； in, Represents a node Remaining CPU resources; Represents a node The centrality of , where This indicates passing through nodes from the starting point to the ending point. The number of shortest paths, This represents the number of shortest paths from the starting point to the ending point; Represents a node The degree of aggregation, where Indicates from node Number of nodes reached Indicates from node The sum of distances to all reachable nodes; Represents a node The degree of sharing.

7. The method for optimizing the deployment of service function chains in a space-ground converged network according to claim 1, characterized in that, Weight Represented as: ； in, and These are nodes Remaining CPU resources and links The remaining bandwidth resources; It is a node Remaining CPU resources; It is a node The degree of sharing; It is a node The degree of the node; , To calculate the weights and bandwidth weights.

8. A method for optimizing the deployment of service function chains in a space-ground converged network according to any one of claims 1-7, characterized in that, The space-ground integrated network includes a terrestrial network and a satellite network; The terrestrial network and the satellite network are connected via a satellite-to-ground link; The ground network includes several server ground nodes, and the several server ground nodes are connected by ground links. The satellite network includes several low-Earth orbit (LEO) satellite nodes, which are connected by star links.

Images