A SFC deployment and migration method based on VNF dependent components

Abstract

This invention discloses a method for SFC deployment and migration based on VNF-dependent components, belonging to the field of communication technology. This method jointly optimizes SFC deployment and migration strategies that consider VNF-dependent component migration to address mobile user service provision issues. The method is based on service request information submitted by mobile terminal users to the SDN controller. The SDN controller analyzes the user's demand type and resource requirements based on the service request information, including bandwidth, computing, storage, and latency requirements. Simultaneously, it combines network topology, edge server resource status, and network operation conditions, employing deep reinforcement learning techniques to optimize SFC deployment and migration. This invention comprehensively considers the influence between VNF-dependent component migration, SFC deployment, and SFC migration decisions, efficiently providing mobile services.

Description

Technical Field

[0001] The technical field of this invention is advanced networking and communication technology, specifically relating to a method for deploying and migrating SFC based on VNF dependent components. Background Technology

[0002] With the proliferation of mobile devices and applications, the delivery of mobile services has attracted widespread attention from academia and industry. Mobile service delivery involves not only resolving service offloading issues but also service migration to maintain Quality of Service (QoS). Some studies have proposed service migration strategies to mitigate server failures, alleviate overload, or achieve cost-effectiveness, aiming to maintain optimal service delivery performance. However, these methods are not well-suited to the mobile service delivery scenario. Addressing the migration challenges posed by user mobility is crucial for ensuring continuous service delivery and optimizing overall performance.

[0003] Some studies have considered the migration issues caused by user mobility, but none have taken into account the significant performance impact of service state migration due to user mobility. Later studies considered service state migration to mitigate the impact of user mobility. Researchers have extensively explored the complexities of SFC deployment and migration, including SFC deployment and migration under conditions of random user movement and SFC deployment and migration assuming known user movement trajectories. However, accurately predicting the trajectories of mobile users remains a significant challenge in most cases.

[0004] While previous research has provided effective solutions for mobile services in various environments (such as MEC and cloud / fog computing) and considered user mobility, these strategies are difficult to apply to new scenarios due to the issues of SFC deployment and migration involving VNF-dependent component migration. Zhang et al. addressed the migration problem of VNF-dependent components during SFC deployment; however, their research focused on the static deployment of SFCs. When considering user mobility, the SFC itself must be migrated simultaneously. This invention proposes an innovative method for online deployment of mobility-aware SFCs in dense MEC networks integrated with 3GPP cellular networks, and proposes an SFC migration control strategy that redeploys these chains while considering the migration of VNF-dependent components. Summary of the Invention

[0005] Purpose of the invention: In order to optimize end-to-end communication latency and resource usage costs for mobile users, reduce interruption latency, improve service reception rate, efficiently manage limited resources, and provide high-quality mobile services, this invention provides an SFC deployment and migration method based on VNF dependent components.

[0006] Technical Solution: A method for deploying and migrating SFC based on VNF-dependent components. This method is based on service request information submitted by mobile terminal users to the SDN controller. The SDN controller analyzes the user's demand type and resource requirements based on the service request information, including bandwidth, computing, storage, and latency requirements. Simultaneously, it combines network topology, edge server resource status, and network operation conditions, employing deep reinforcement learning technology to optimize the deployment and migration of SFC. The method includes the following steps:

[0007] (1) Construct a communication architecture that includes mobile users, edge servers and SDN controllers;

[0008] In this communication architecture, mobile users generate dynamic service requirements, including computing tasks, data storage, and communication services. Edge servers are deployed close to the users to provide computing and communication services. The SDN controller monitors the mobile users' service requirements and the resource status of the edge servers in real time and dynamically adjusts resource allocation and scheduling strategies based on network topology information.

[0009] (2) Establish user request model, communication model and computing model to describe the system's operating mechanism and performance indicators;

[0010] The user request model includes SFC deployment requests and migration requests. Deployment requests describe the user's requirements for bandwidth resources, computing resources, memory resources, and storage resources, while migration requests are used to transfer the currently deployed SFC to the target edge server when the user moves.

[0011] The communication model defines a bandwidth allocation strategy for edge servers and calculates transmission rate and data transmission time based on Shannon's formula and channel interference index, thereby achieving a reasonable allocation of communication resources.

[0012] The computational model includes end-to-end communication latency, resource usage cost, and interruption latency during SFC migration, which is used to evaluate service performance and provide theoretical support for system resource optimization and scheduling.

[0013] (3) Based on the network topology information, user request model, communication model and computing model of the communication system, formulate a resource scheduling optimization problem and clarify the optimization objective;

[0014] The optimization problems include maximizing the service access rate of user requests, minimizing end-to-end communication latency and SFC migration latency, and maximizing resource utilization. During the SFC migration process, the migration path of VNFs and the resource usage cost of dependent components are optimized to reduce service interruption time caused by migration.

[0015] (4) Consider the deployment and migration of SFC based on deep reinforcement learning for VNF dependent components;

[0016] The SDN controller collects user requests, edge server resource status, and network topology information in real time, and uses deep reinforcement learning algorithms to dynamically make decisions on the following:

[0017] User access strategy: Intelligently select the optimal edge server for access based on the user's geographical location and business needs;

[0018] SFC Deployment and Migration Strategy: Dynamically adjust the SFC deployment and migration path by comprehensively considering resource utilization, communication latency, and migration latency;

[0019] VNF Dependency Component Migration Strategy: For VNF dependencies in SFC, prioritize migrating critical VNF components to ensure the integrity and continuity of the service chain.

[0020] Furthermore, the user request model, communication model, and calculation model described in step (2) are as follows:

[0021] (21) User Request Model

[0022] User request r is represented as SFCr. It is a quintuple of SFCr, where b r Defined as bandwidth requirement, Indicates communication latency requirements, TTL r It is the lifecycle of SFCr, and SFC is described as a directed graph G. r =(F r E r ),in Let f represent the set of VNFs in SFCr, where any VNF f rh All require CPU resources and memory resources E r This represents the virtual link of SFCr, and the arrival time of SFCr is expressed as... m represents the number of time slots, and Δ represents the length of a time slot;

[0023] For SFC deployment, use decision variables Indicates whether to pass through edge server node n i Access to the network Used to indicate whether VNF f rh Placed on edge server node n i superior, It is a binary variable used to represent a logical link. Is it mapped to link e? ij The above is mathematically represented as:

[0024]

[0025] For SFC migration, when the VNF is deployed to edge server node n i Or migrate to edge server node n i When n i If the component resources required to support VNF ​​operation are lacking, the VNF dependent components must be migrated from other nodes to the edge server node n. i To ensure the proper functioning of the VNF;

[0026] (22)Communication model

[0027] edge service node n i The bandwidth resources are evenly allocated to SFCr to ensure fair distribution, and the mathematical expression is as follows:

[0028]

[0029] Where, N ni Indicates from edge service node n i The SFC set accessed This represents the server node n. i bandwidth resources on Indicates r j From server node n i Accessing the network, if from n i 1 indicates network access; otherwise, 0 indicates accessibility.

[0030] Use parameters This indicates that SFCr accesses edge server node n i The channel interference experienced is caused by channel contention from other SFCs simultaneously accessing the same node. Indicates SFCr j Transmission capability Indicates channel gain, affected by SFCr j With edge service node n i The mathematical expression for the distance between them and the power gain effect caused by small-scale fading in the channel is:

[0031]

[0032] Based on this, the \ in the formula represents the influence of SFCr and the service node n. i The transmission rate between them is obtained by the following formula:

[0033]

[0034] Where, ρ 2 It refers to the signal-to-noise ratio and data transmission delay. This refers to the data to be transmitted from SFCr to n. i The required time is expressed as:

[0035]

[0036] (23) Computational Model

[0037] SFCr's end-to-end communication delay D r Defined as the sum of data transmission delay, VNF processing delay, and link propagation delay, as shown below:

[0038]

[0039] The deployment cost of SFC is defined as follows:

[0040]

[0041] in, It is the migration component ψ q The path used It is a binary variable representing a path. On It is mapped to link e ij superior, n i The unit storage cost.

[0042] Furthermore, this method minimizes downtime and enhances success rate through pre-copy migration. Downtime is defined as the duration of the final stage of the migration process, during which the service is temporarily interrupted to complete final data synchronization. Indicates the start time of the interruption period. The expressions for the two, representing the end time of the interruption period, are as follows:

[0043]

[0044] in, This represents the transition time in the q-th iteration. This is the actual number of iterations. Indicates the VNF to be migrated f rh memory page size, Indicates VNF f rh The migration rate, therefore, the interrupt time is defined as follows:

[0045]

[0046] M in the formula r This represents the set of VNFs that need to be migrated in SFCr.

[0047] Furthermore, the optimization problem and optimization objective described in step (3) are as follows:

[0048] Constraint 1: Each f rh A VNF instance can only be successfully deployed on one server node at most. VNF instances are indivisible and are represented as follows:

[0049]

[0050] Constraint 2: The communication quality between users connected to the network through the same node is mutually influential, and each physical node can accommodate a maximum of n. max To maintain service quality for all users, the mathematical representation is:

[0051]

[0052] Constraint 3: On any server node n i Above, the total CPU and memory resource requirements for SFC deployment and migration must not exceed the maximum CPU and memory resource capacity, mathematically expressed as:

[0053]

[0054] Constraint 4: The used storage resources will not exceed the maximum storage resource capacity, mathematically expressed as:

[0055]

[0056] Among them, Ψ ni It is node n i A collection of components on;

[0057] Constraint 5: Any link e ij The total bandwidth consumption must be less than the maximum bandwidth capacity, as follows:

[0058]

[0059] The first item represents the link bandwidth used for SFC deployment, and the second item represents the bandwidth used for SFC migration.

[0060] Constraint 6: The end-to-end delay of any SFCr must satisfy the following constraint:

[0061]

[0062] Based on the constraints defined above, the JSDM-VDSM problem includes SFC deployment and migration strategies, and VNF dependent component migration strategy Z. The goal of this method is to maximize QoS by improving service reception rate, reducing resource usage costs, and minimizing communication latency and interruption latency. The mathematical definition of the optimization objective is as follows:

[0063]

[0064] Among them, Z r Z(t) is a decision variable used to determine whether the SFC request r is accepted; if accepted, Z... r (t) is set to 1, otherwise it is set to 0; w1, w2, w3 and w4 in the formula are the weights of different indicators.

[0065] Furthermore, step (4) also includes:

[0066] The network architecture identifies SFCs that need to be orchestrated, including SFCs that need to be deployed and SFCs that are planned to be migrated, and views SFC migration as a joint optimization of VNF redeployment and SFC rerouting.

[0067] Obtain the status of the environment and the SFC currently to be orchestrated, including the available CPU, memory, storage, bandwidth and component resources of the physical nodes; the CPU, memory and component resource requirements of the VNFs to be deployed; the number of VNFs that have not yet been placed in the deployed SFC; the partial communication latency of the current SFC; and the lifespan of the SFC.

[0068] The obtained state s ι Input to Actor old Network, obtain VNF deployment decisions a ι Subsequently, iterative interactions with the environment were conducted to collect a set of experimental data [s] ι ,a ι ,r ι ,s ι+1 ] and store it in memory;

[0069] Calculate the advantage function based on experimental data. Used to update the Critic network; uses a truncated objective function. The Actor network is updated, and after multiple iterations, the parameters of the Actor network are passed to the Actor. old The network iterates continuously until it converges.

[0070] The solution process for the JSDM-VDSM problem includes obtaining the SFC request r and determining the user access node. and SFC deployment or migration strategies Based on strategy and Determine the migration strategy for VNF dependent components Before learning the service delivery strategy, the SDN controller collects global information and releases SFCs that have reached the end of their lifecycle, and then identifies the SFCs that need to be deployed or migrated.

[0071] The SFC request for deployment is an SFC request that arrives randomly within the current time slot t, while the SFC request for migration is an SFC request that meets the migration criteria within the same time slot. If there are no requests to process, the SDN controller will move to the next time slot and continue to acquire requests until r. batch Not empty; then, r batch The SFC requests in the process will be deployed or migrated sequentially.

[0072] For deployment and migration requests, the user area associated with the SFC request is first determined. Then, the SDN controller evaluates whether the base station in the current area has sufficient capacity. If there is sufficient capacity, access is made through that base station. If there is insufficient capacity, the SDN controller plans to connect through neighboring areas. First, neighboring areas are identified, and then the access capacity of these areas is calculated. Next, the number of accessible nodes with the maximum access capacity is determined, and the area that can accommodate the maximum capacity is selected. Furthermore, an access node is selected from the set of max_num_nodes with a probability of 1-∈ and an access node is selected from the set of n_max_num_nodes with a probability of ∈.

[0073] After selecting the access point, determine the deployment or migration strategy for the SFC and the migration strategy for VNF dependent components. First, identify the first VNF ​​of the SFCr, then obtain the current state, and proceed according to the strategy π. θold (a ι ,s ι ), determine VNF f rh The placement position of f rh Can be placed on node a ι Up, but a ι Currently does not have f rh To run the required component resources, the necessary f must be provided. rh The dependent components are migrated to a ι Then place the next VNF ​​and record the updated state s. ι+1 ;

[0074] The above process continues until the SFC is successfully deployed or migrated; rewards are calculated based on the deployment results: if deployment is successful, a reward for successful deployment is calculated; if migration is successful, a reward for successful migration is calculated; if SFC deployment or migration fails, the system will revert to the last successfully achieved state.ι-u+1 .

[0075] Beneficial Effects: The method described in this invention comprehensively considers the mobility of end users and the service requests of mobile end users. First, it constructs an SFC deployment and migration network architecture that considers the migration of VNF-dependent components. Then, based on this architecture, it proposes a joint SFC deployment and migration method for VNF-dependent components, including: obtaining SFC request information, determining access nodes, and designing SFC deployment and migration decisions and VNF-dependent component migration strategies. Finally, experimental evaluations were conducted, and the results show that, because this invention comprehensively considers the influence between VNF-dependent component migration, SFC deployment, and SFC migration decisions, it can provide efficient and high-performance mobile services. Attached Figure Description

[0076] Figure 1 This is a flowchart illustrating the method described in this invention;

[0077] Figure 2 This invention provides an example diagram of mobile services under the MEC system architecture in this embodiment.

[0078] Figure 3 This is a network architecture diagram for SFC deployment and migration considering the migration of VNF dependent components in an embodiment of the present invention;

[0079] Figure 4 shows the algorithm convergence diagram in the embodiment of the present invention, wherein: Figure 4(a) is the convergence diagram comparing the average reception rate of SFC, and Figure 4(b) is the convergence diagram comparing the policy gradient loss.

[0080] Figure 5 is a schematic diagram comparing the algorithm performance under different numbers of service function chains and different migration conditions in the embodiments of the present invention. In the figure: Figure 5(a) is the average reception rate under speed=high,|F|=3, Figure 5(b) is the average end-to-end communication latency under speed=high,|F|=3, Figure 5(c) is the average number of migrations of VNF dependent components under speed=high,|F|=3, Figure 5(d) is the average migration cost of VNF dependent components under speed=high,|F|=3, Figure 5(e) is the average placement cost of VNF dependent components under speed=high,|F|=3, Figure 5(f) is the average interruption latency of SFC under speed=high,|F|=3, Figure 5(g) is the total number of migrations of VNF dependent components under speed=high,|F|=3, Figure 5(h) is the total migration cost of VNF dependent components under speed=high,|F|=3, and Figure 5(i) is the total placement cost of VNF dependent components under speed=high,|F|=3.

[0081] Figure 6 is a schematic diagram comparing the algorithm performance under different migration conditions and different user movement speeds in the embodiments of the present invention, wherein: Figure 6(a) is Γ d =240, |F|=3, average reception rate, Figure 6(b) is Γ d Average end-to-end communication delay at |F|=3 and |F=240;

[0082] Figure 7 is a schematic diagram comparing the algorithm performance under different user movement speeds and different business function chain lengths in the embodiments of the present invention, wherein: Figure 7(a) is when dis = 2, Γ d =240 average reception rate, Figure 7(b) is the average reception rate at dis=2,Γ d =240 Average end-to-end communication delay. Detailed Implementation

[0083] To illustrate the technical solutions disclosed in this invention in detail, the invention will be further described below with reference to the accompanying drawings and embodiments.

[0084] This invention provides a method for deploying and migrating SFC based on VNF-dependent components, which maximizes QoS by improving service reception rate, reducing resource usage costs, and minimizing communication latency and interruption latency.

[0085] While new smart mobile devices, such as wearables, smart vehicles, robots, and drones, have significantly improved our quality of life by offering greater convenience and efficiency, they have also led to an increase in computationally intensive tasks that demand high computing power and performance, particularly in low-latency and continuous communication. Given the limited resources of mobile devices, such as computing power and battery life, processing these tasks locally is often impractical. With the exponential growth in the number of mobile devices and the increasing data traffic generated by computationally intensive tasks, future congestion of core networks may be exacerbated. To address these challenges, offloading mobile services to edge servers closer to mobile devices is a promising solution. This approach not only improves quality of service by providing additional resources and reducing communication latency but also helps alleviate network congestion.

[0086] By integrating Network Function Virtualization (NFV) technology, mobile services are represented by ordered Virtual Network Functions (VNFs), which constitute Service Function Chains (SFCs). Deploying these SFCs helps alleviate the problem of resource-constrained edge nodes struggling to accommodate large-scale, resource-intensive services, such as high-definition video streaming, real-time gaming, and complex data analytics, thereby improving the success rate of mobile service offloading. However, the high mobility and uneven distribution of mobile devices lead to significant spatiotemporal network load imbalances. Furthermore, all mobile devices compete for limited edge resources. Therefore, meeting Quality of Service (QoS) requirements through SFC deployment is a challenging task that urgently needs to be addressed. Seamlessly migrating SFCs is crucial to maintaining service continuity as mobile devices move, which further complicates service delivery.

[0087] In previous research, VNFs have been implemented as packages, with functional instantiation achieved by placing these VNF packages on servers. However, in large-scale heterogeneous edge networks, the diversity of virtualization platforms and the frequent iterations of VNFs require decoupling of VNF packages. Decoupling of VNF packages is crucial for flexible and efficient VNF ​​placement, ensuring rapid adaptation to evolving business needs. Deploying or migrating SFCs requires deploying VNFs to servers with the necessary component resources (e.g., Ubuntu, VEKET, IPFilter, and Hash Switch) – collectively referred to as VNF dependencies – to support VNF ​​operation. In previous SFC deployments or migrations, assuming each server node could support all VNFs was unrealistic, especially considering the limited resources of edge server nodes.

[0088] When the node hosting a VNF lacks the necessary components to support its operation, the VNF's dependent components can be migrated to the corresponding server node so that the VNF can function properly. The deployment and migration of SFCs is inherently a complex process. Including the migration of VNF dependent components, while more accurately reflecting the real-world scenario, further exacerbates the challenges of mobile service delivery.

[0089] When considering VNF dependent components, reuse and migration strategies must be evaluated during SFC deployment to ensure optimal VNF operation. Reusing component resources can reduce storage costs but may increase communication latency. On the other hand, migrating components incurs migration and storage costs but may reduce communication latency. During SFC migration, it is crucial to consider the VNF to be migrated and the target migration node. VNFs can be migrated to nodes that support or do not support VNF ​​operation. If a node does not support VNF ​​operation, the VNF dependent components must also be migrated to that node. Therefore, how to efficiently manage limited resources while providing high-quality mobility services is an urgent problem that needs to be solved.

[0090] This invention provides a method for SFC deployment and migration based on VNF-dependent components. The method involves mobile users sending service request information to the SDN controller. The SDN controller senses the user request, tracks the mobile user's location, coordinates the deployment and migration of SFCs (including VNF-dependent component migration), and manages network resource allocation. It considers user mobility, the need for high computing power and low latency, resource constraints at edge nodes, and the challenges posed by VNF-dependent component migration. Due to the complexity of the problem, we propose a JSDM-VDSM architecture based on deep reinforcement learning, laying the foundation for JSDM-VDSM algorithm design. This algorithm includes a mechanism for determining the SFC to be migrated, user access policies, SFC deployment and migration policies, and VNF-dependent component migration policies. Compared with traditional methods, the proposed mechanism provides a more flexible, efficient, and adaptable solution because it can intelligently adapt to dynamic and complex network environments. Simulation experiments evaluate the performance of JSDM-VDSM. The results show that the proposed algorithm exhibits superior performance compared to benchmark algorithms, particularly in terms of reception rate and end-to-end latency.

[0091] The implementation process of the technical solution provided by this invention is described in detail below.

[0092] The method described in this invention enables SFC deployment and migration that considers the migration of VNF-dependent components. It mainly comprises three components: mobile terminal users, a 3GPP-based MEC network, and an SDN controller. Mobile terminal users submit service requests and move over time; the MEC network server provides communication and computing resources to mobile terminal users; and the SDN controller manages terminal users and network resources.

[0093] The main implementation process of the method described in this invention is as follows: Figure 1 As shown, based on the above technical solution, further detailed description is provided in the embodiments, specifically including the following steps:

[0094] (1) Establish a communication system architecture

[0095] This embodiment constructs a communication system architecture consisting of mobile users, edge servers, and an SDN controller. Mobile users are primarily responsible for dynamically generating service requests, edge servers provide computing and communication services to mobile users, and the SDN controller plays a coordinating and management role in the entire system, responsible for real-time allocation of user service requests and edge server resources. This architecture, through the collaborative work of mobile users, edge servers, and the SDN controller, ensures that the network can efficiently handle complex resource scheduling and dynamic service requests.

[0096] (2) Establishment of user request model, communication model and computing model

[0097] This embodiment establishes a user request model, a communication model, and a computation model, providing a comprehensive description of the communication system. The user request model defines SFC deployment and migration requests, covering the requirements for bandwidth, computing, memory, and storage resources, while also including lifecycle and communication latency requirements. The communication model details the bandwidth allocation strategy for edge servers, quantifying communication channel interference, data transmission rate, and transmission time. The computation model, starting from end-to-end communication latency, resource usage costs, and SFC migration interruption latency, provides a theoretical basis for system performance analysis and optimization.

[0098] (3) Formulating the optimization problem and determining the optimization objective

[0099] This embodiment defines an optimization problem based on the system architecture and related model information, and clarifies the overall optimization objectives. The optimization problem mainly revolves around the user request model, communication model, computing model, and edge server topology information, aiming to address extreme user demands and edge resource constraints. The optimization objectives not only need to meet users' requirements for mobility, high computing power, and low latency, but also need to maximize the utilization of limited edge node resources and effectively address the complexity and challenges of migrating VNF-dependent components, ensuring a dual improvement in network performance and resource efficiency.

[0100] (4) SFC deployment and migration method based on VNF dependency component migration

[0101] This embodiment proposes a method for SFC deployment and migration based on VNF-dependent components, and constructs a network architecture based on this method. By introducing deep reinforcement learning technology, this method intelligently decides on the SFC migration mechanism, user access policy, and SFC deployment and migration strategy, while optimizing the migration of VNF-dependent components. Compared with traditional methods, this method can not only better adapt to dynamic and complex network environments, but also significantly reduce migration latency, improve the flexibility and efficiency of resource allocation, and provide important support for the efficient management of future network communications.

[0102] The implementation process is described in detail below.

[0103] 1) Construction of communication system architecture

[0104] This architecture comprises three distinct layers: the Mobile User (MU) layer, the MEC layer, and the Software-Defined Networking (SDN) control layer. Specifically, it consists of mobile users, edge servers, and an SDN controller to meet the communication and computing needs of dynamic network environments. In this architecture, mobile users interact with the system by generating service requests (such as computing tasks or data transmission requests). Edge servers, deployed close to users, provide low-latency, high-performance services by allocating computing, storage, and bandwidth resources. The SDN controller, as the core unit of global resource scheduling, monitors user service requests, network topology information, and edge server resource status in real time, and intelligently optimizes resource allocation and scheduling strategies. In the communication process, mobile user requests are first parsed by the SDN controller. Based on the user's request type and the current network status, the controller schedules the request to the optimal edge server, thereby achieving efficient utilization of computing and communication resources. By simulating different user needs and network scenarios, this architecture validates its flexibility in resource management and its rapid response capability to user needs in dynamic environments.

[0105] Specifically, the runtime is divided into discrete time slots. The time interval is Δ.

[0106] MU (Mobile User): Let MU = {u1, u2, ..., u} r ,…,u |MU| Let} represent the set of mobile users. Each mobile user generates a computationally intensive SFC request within time T. Mobile user u r The position in time slot t is determined by This means that in each time slot, the communication system must handle both newly arriving requests and ongoing user requests, stemming from the random arrival and movement of users. Therefore, for these newly arriving requests, corresponding SFCs need to be deployed, while for existing users, migration may be necessary to ensure continuous service.

[0107] Edge server: set This is a collection of edge servers. Each edge server is located at the base station and is equipped with specific computing, memory, storage, and bandwidth resources, each used for... and This indicates that these resources enable the server to provide communication and computing services to mobile users. The location of the edge server is fixed, with a height of 0. The location of the edge server is represented by... It means that, among them is n i The network horizontal coordinates, These are its network vertical coordinates. Defining edge server coordinates helps provide accurate geographic location information, which is crucial for optimizing service selection for mobile users. Due to the limited resources of edge servers and the constantly changing locations of mobile users, SFC's VNFs may be hosted on different edge servers, and each server may serve different mobile users in different time slots.

[0108] Controller: The SDN controller plays a crucial role in sensing user requests, tracking the location of mobile users, coordinating the deployment and migration of SFCs (including the migration of VNF-dependent components), and managing network resource allocation. For example, ... Figure 2 As shown, within time slot t, the SDN controller receives new service requests from users 1, 2, and 3, each accessing the network through their respective base stations. The SDN controller is responsible for coordinating the deployment of SFCs to edge servers to fulfill user requests. VNF1 and VNF2 of SFC1 are located in MEC1, VNF1 of SFC2 is located in MEC2, and VNF2 is located in MEC3. For SFC3, VNF2 is located in MEC1, and VNF4 is located in MEC2. However, due to the lack of necessary component resources in MEC1 to support the operation of VNF1, software1 needs to be migrated from MEC2 to MEC1 to ensure the proper functioning of VNF1. In the next time slot t+1, users 1, 2, and 3 have moved to new locations. User 1 moves more slowly, continuing to access services through the previous base station and edge server. User 2 moves at a moderate pace, modifying their access method using a rerouting strategy. Meanwhile, user 3 moves more quickly, ensuring service quality and continuity by strategically migrating VNFs and switching to a new access point. VNF2 in SFC3 is migrated from MEC1 to MEC3, while VNF4 in SFC3 is migrated from MEC2 to MEC3. However, since MEC3 lacks the necessary components to support VNF4 execution, software4 needs to be migrated from MEC2 to MEC3. During SFC deployment and migration, the SDN controller manages the allocation and release of network resources to adapt to changing needs.

[0109] 2) Establishment of user request model, communication model and computing model

[0110] To meet the needs of the communication system, a user request model, a communication model, and a computation model were established to describe the operating mechanisms and performance indicators of each part of the system.

[0111] User Request Model: User requests are divided into two categories: SFC deployment requests and migration requests. SFC deployment requests include requirements for bandwidth, computing, memory, and storage resources, and define the request lifecycle and communication latency requirements. Migration requests are mainly for user mobility scenarios, requiring the migration of the currently deployed SFC to a new edge server while ensuring service continuity and low latency during the migration process.

[0112] Communication Model: The communication model establishes methods for calculating interference, transmission rate, and data transmission time of the communication channel by defining the bandwidth allocation strategy for edge servers. Specifically, the communication model estimates channel capacity based on Shannon's law and calculates the transmission rate and communication latency for each request by combining interference and bandwidth resource status.

[0113] Computational Model: By analyzing the allocation and usage costs of computing resources on edge servers, the computational model defines a method for calculating end-to-end communication latency. Simultaneously, the model describes the interruption latency of the business function chain during migration, used to evaluate the impact of migration strategies on service performance. These models collectively provide a clear theoretical foundation for the design of optimization algorithms.

[0114] Specifically, a user request is represented as an SFC request. To more clearly illustrate the mathematical processing, this invention represents any user request as SFCr, where SFC requests represent all user requests, and SFCr is a specific one among the SFC requests. It is a quintuple of SFCr, where b r Defined as bandwidth requirement, TTL indicates communication latency requirements. r =l·Δ is the lifetime of SFCr. SFC is described as a directed graph G. r =(F r E r ),in This represents the set of VNFs in SFCr. Each VNF f rh CPU resources required and memory resources E r This represents the virtual link of SFCr. The arrival time of SFCr is expressed as...

[0115] Γ d Γ represents the set of SFC deployment requests. m This represents a set of migration requests. In this invention, "base station," "edge server," and "physical node" are synonyms and can be used interchangeably. For SFC deployment, it is necessary to determine which base station the user accesses the network through, the deployment location of VNFs, the migration of VNF dependent components, and the mapping path of virtual links, represented as follows:

[0116] Decision variables in (1) Determine whether to go through the edge server n i Connect to the network.

[0117] (2) Decide whether to include VNF f rh Placed in n i superior.

[0118] (3) It is a binary variable, representing Is it mapped to link e? ij superior.

[0119]

[0120]

[0121] For SFC migration, it is necessary to determine the appropriate timing of the migration, confirm whether a switch of access point is required, identify the VNFs that need to be migrated, and determine whether VNF dependent components need to be migrated. r This represents the set of VNFs waiting to be migrated. It records the source node where each VNF to be migrated is located. Memory filth rate and the size of the memory pages occupied This is also crucial. Furthermore, careful selection of the target node to which the VNF will be migrated is necessary. Ψ={ψ1,ψ2,...,ψ q ,...,ψ |Ψ|} is a collection of components that support VNF ​​operation. Represents node n i The set of VNF dependent components on node n. When a VNF is deployed to node n... i Or migrate to node n i When n i If the component resources required to support VNF ​​operation are lacking, the VNF dependent components must be migrated from other nodes to n. i This is to ensure the proper functioning of the VNF. Represents the VNF dependent component ψ q Storage resources used. Binary variables. In (4), component ψ is represented. q Does it exist at node n? i Up. We use Represented as f rh The collection of all migrated components.

[0122]

[0123] Edge server n i The bandwidth resources are evenly allocated to SFCr(MUr) to ensure fair allocation, as expressed in the following expression:

[0124]

[0125] in, Indicates from n i The set of MUs accessed. Parameters defined in formula (6). This indicates that MUr accesses the edge service node n. i The channel interference experienced is caused by channel contention resulting from other MUs simultaneously accessing the same node. Indicates MUR j Transmission capacity Indicates channel gain, subject to MUr j With node n i The distance between them and the power gain effect caused by small-scale fading in the channel,

[0126]

[0127] Based on this, SFCr and service node n i The transmission rate between them can be obtained using the following formula:

[0128]

[0129] Where, ρ 2 It is the signal-to-noise ratio. The transmission delay expressed in formula (8) refers to the transmission delay of the data to be transmitted from SFCr to n. i The time required.

[0130]

[0131] The end-to-end communication delay of SFCr is defined as the sum of data transmission delay, VNF processing delay, and link propagation delay, as shown below:

[0132]

[0133] During SFC deployment, certain costs are incurred due to the use of CPU and memory resources, as well as the migration of VNF dependent components. Since CPU and memory usage costs are inherent to the deployment process, this study primarily focuses on the costs associated with the migration of VNF dependent components. Therefore, the deployment cost of SFC is defined as follows:

[0134]

[0135] in, It is the migration component ψq The path used It is a binary variable representing a path. on It is mapped to link e ij superior, Represents n i The unit storage cost.

[0136] This invention employs a pre-copy migration technique, characterized by flexible migration features, minimized downtime, and enhanced success rate. In the pre-copy migration technique, downtime is defined as the duration of the final stage of the migration process, during which the service is temporarily interrupted to complete the final data synchronization. Indicates the start time of the interruption period. Indicates the end time of the interruption period. These are detailed in (11) and (12) respectively:

[0137]

[0138] Among them, (11) and (12) This represents the transition time in the q-th iteration. It is the actual number of iterations, while Indicates VNF f rh The migration rate. Therefore, the definition of interrupt time is as shown in (13):

[0139]

[0140] 3) Formulating the optimization problem and setting the optimization goal

[0141] Based on the user request model, communication model, and computation model, the optimization problem aims to address the following challenge: how to rationally allocate the limited resources of edge servers while simultaneously meeting users' demands for low latency and high computing power, given dynamically changing user requests. Optimization objectives include:

[0142] Improve service reception rate: Maximize the number of user requests received to ensure the system's service capacity under high load scenarios.

[0143] Minimize communication and migration latency: Optimize SFC deployment and migration strategies, reduce end-to-end communication latency, and minimize the impact of SFC migration on user experience.

[0144] Improve resource utilization: While ensuring service quality, make full use of the computing, bandwidth and storage resources of edge servers to avoid resource waste.

[0145] Addressing the challenges of VNF dependency migration: For the dependencies between VNFs in the business function chain and the dependencies between VNFs and components, reduce the migration complexity and performance overhead caused by dependencies by optimizing migration paths and strategies.

[0146] Specifically, this step analyzes the constraints that arise during the deployment and migration of SFCs, particularly those related to the migration of VNF dependent components, and then formulates this issue. Constraint (14) ensures that each f rh Only one server node can be successfully deployed at most, which means that VNF ​​instances are indivisible:

[0147]

[0148] The communication quality between users connected to the network through the same node is mutually influential. It is proposed that each physical node can accommodate a maximum of n... max Individual users, to maintain service quality for all users:

[0149]

[0150] Constraints (16) and (17) ensure that on any server node n i Above, the total CPU and memory resource requirements for SFC deployment and migration must not exceed the maximum CPU and memory resource capacity:

[0151]

[0152] Constraint (18) ensures that the used storage resources do not exceed the maximum storage resource capacity:

[0153]

[0154] in, It is node n i A collection of components on [the platform].

[0155] Any link e ij The total bandwidth consumption must be less than the maximum bandwidth capacity. This condition is as follows:

[0156]

[0157] The first term represents the link bandwidth used for SFC deployment, and the second term represents the bandwidth used for SFC migration. To ensure the quality of service delivered, latency constraint (20) ensures that the end-to-end latency of any SFCr must be met, with the following specific restrictions:

[0158]

[0159] Defining JSDM-VDSM issues: SFC deployment strategy and SFC migration strategy (i.e. and ), and VNF dependency component migration strategy This will significantly affect the quality of the services provided. Under constraints (14)-(20) and regarding... and Under this decision-making framework, the objective of this invention is to maximize QoS by improving service reception rate, reducing resource usage costs, and minimizing communication latency and interruption latency. The optimization objective is defined as follows:

[0160]

[0161] Among them, Z r Z(t) is a decision variable used to determine whether the SFC request r is accepted. If accepted, Z... r (t) is set to 1 otherwise to 0. w1, w2, w3, and w4 are the weights of different indicators.

[0162] 4) SFC Deployment and Migration Method Based on VNF Dependency Component Migration

[0163] This step is also a proposed SFC deployment and migration method based on VNF dependent components, and a deep reinforcement learning-driven intelligent network architecture is constructed.

[0164] A network architecture for joint VNF ​​dependency component migration (JSDM-VDSM) is constructed, including a mobile user access module, an edge server resource management module, and an SDN control module. This architecture supports dynamic user access, resource scheduling, and service migration functions, and models the dependencies between different VNFs in the service function chain.

[0165] Deep Reinforcement Learning Algorithm Design: A deep reinforcement learning technology is employed to design an intelligent decision-making module, including an SFC migration mechanism, user access strategy, SFC deployment and migration strategy, and VNF dependent component migration strategy. By defining a reward function, the algorithm is encouraged to maximize resource utilization, reduce migration latency, and ensure the integrity of the business function chain during the migration process.

[0166] Strategy Optimization and Training: In a simulated environment, user requests and network resource states are generated as training samples. A deep reinforcement learning model is used to continuously optimize resource allocation strategies. Emphasis is placed on optimizing VNF migration paths, prioritizing the migration of VNF components with critical dependencies to ensure service quality is not affected during the migration process.

[0167] The specific implementation process includes the following steps:

[0168] First, the network architecture of JSDM-VDSM is introduced. Then, JSDM-VDSMA is proposed to provide an SFC deployment and migration strategy with VNF dependent component migration policies.

[0169] Figure 3 The network architecture of JSDM-VDSM is demonstrated, which mainly consists of two key steps: environment and training network. Due to the random arrival and movement of users, JSDM-VDSM employs an online learning method to dynamically determine A, X, Y, and Z in a dynamic network environment.

[0170] Initially, the network architecture identifies SFCs that require orchestration, including SFCs that need to be deployed and SFCs that are planned to be migrated. SFC deployment involves placing VNFs on physical nodes and mapping virtual links between VNFs to corresponding physical links. The migration process moves VNFs from their original physical nodes to new physical nodes while simultaneously rerouting SFCs to improve quality of service. Therefore, SFC migration can be viewed as a joint optimization of VNF redeployment and SFC rerouting.

[0171] Next, the system acquires the environment and the status of the SFC currently to be orchestrated. This status includes the available CPU, memory, storage, bandwidth, and component resources of the physical nodes; the CPU, memory, and component resource requirements of the VNFs to be deployed; the number of VNFs not yet placed in the deployed SFC; the partial communication latency of the current SFC; and the TTL (Time To Live) of the SFC.

[0172] The obtained state s ι Input to Actor old Network, obtain VNF deployment decisions a ι Subsequently, iterative interactions with the environment were conducted to collect a set of experimental data [s] ι ,a ι ,r ι ,s ι+1 The data is stored in memory. The advantage function is calculated based on the experimental data. This is used to update the Critic network. Next, the objective function is truncated. Update the Actor network. After several iterations, the parameters of the Actor network are passed to the Actor... old The network iterates until it converges.

[0173] JSDM-VDSMA includes algorithms for obtaining SFC requests and determining user access nodes. Methods, and SFC deployment or migration strategies The method, based on strategy and VNF Dependency Component Migration Strategy Before learning the service delivery strategy, the SDN controller first gathers global information and releases SFCs that have reached the end of their lifecycle. Then, it identifies SFCs that need to be deployed or migrated. SFCs requesting deployment are SFC requests that arrive randomly within the current time slot t, while SFCs requesting migration are SFC requests that meet the migration criteria within the same time slot. If there are no requests to process, the controller moves to the next time slot and continues to acquire requests until r... batch Not empty. Then, r batch The SFC requests in the process will be deployed or migrated sequentially.

[0174] For deployment and migration requests, selecting access nodes for network communication is crucial. This selection plays a key role in determining access latency, which is affected by the node's remaining capacity and the communication distance between the node and the user. Initially, the user area associated with the SFC request is determined. Then, the controller assesses whether the base station in the current area has sufficient capacity. If sufficient capacity is available, access is granted through that base station. If insufficient capacity is available, the controller plans to connect through neighboring areas. First, neighboring areas are identified, and then their access capacities are calculated. Next, the number of accessible nodes with the maximum access capacity is determined, and the area capable of accommodating the maximum capacity is selected. Then, an access node is selected from the set of `max_num_node` with probability 1-∈ and from the set of `n_max_num_node` with probability ∈.

[0175] After selecting the access point, the deployment or migration strategy for the SFC, as well as the migration strategy for VNF dependent components, are determined. This process first identifies the first VNF ​​of the SFCr and then obtains its current state. Based on the policy π... θold (a ι ,s ι ), determine f rh The placement position. If f rh It can be placed on node a ι Up, but a ι Currently does not have f rh To run the required components, the required f must be... rh The dependent components are migrated to a ι Then place the next VNF ​​and record the updated state s. ι+1 This process continues until the SFC is successfully deployed or migrated. Rewards are calculated based on the deployment result: if deployment is successful, a reward is calculated; if migration is successful, a reward is calculated. If SFC deployment or migration fails, the system will revert to the last successfully achieved state. ι-u+1 .

[0176] In this embodiment, simulation experiments were conducted to verify the actual effect of the present invention. To better illustrate the effect of the present invention, a rerouting-based algorithm was used for comparison. Figure 4 shows the convergence behavior of JSDM-VDSMA and JSDR-VDSMA, highlighting the changing trends of the receiver rate and policy gradient loss. As shown in Figure 4(a), the average receiver rate of SFC shows a continuous upward trend with the increase of learning steps, eventually stabilizing at a stable value. Meanwhile, the policy gradient loss in Figure 4(b) gradually decreases with the increase of learning steps, eventually converging to a stable state. This indicates that the learning process of both algorithms is effective and stable. Experimental results show that compared with traditional methods, the JSDM-VDSM method has significant advantages in communication latency, service continuity, and adaptability to dynamic environments. Its effects mainly include the following points:

[0177] (1) Performance analysis as the number of SFC requests changes

[0178] Figure 5 illustrates the performance evaluation of the JSDM-VDSMA and JSDR-VDSMA algorithms under different migration conditions (i.e., dis, where dis represents the SFC migration threshold) and different numbers of SFC requests. In this evaluation, the user's movement speed is set to a relatively fast scenario (e.g., a car, with a speed range of [135 km / h, 150 km / h]), and the SFC length is fixed at 3. As shown in Figure 5(a), the reception rate decreases with the increase in the number of SFC requests. This decrease is because the growth rate of the number of received SFCs is slower than the growth rate of the number of SFC requests. However, the total number of received SFCs continues to increase. When the shortest hop distance between the area where the SFC was last deployed and the current area is greater than or equal to the migration threshold dis, and the SFC is within its lifecycle, the SFC will be migrated. Furthermore, Figure 5(a) also shows that changes in the migration threshold dis lead to performance differences between the JSDM-VDSMA and JSDR-VDSMA algorithms.

[0179] For JSDM-VDSMA, the SFC (Site Found) reception rate decreases as the migration threshold `dis` increases. This is because it assumes all users generating SFC requests move randomly from the center of the MEC (Multi-access Edge Computing) network. As `dis` increases, SFCs remain in their original access and deployment locations for longer periods, causing congestion at nodes near the MEC center. When a new SFC request arrives at the MEC center, there may not be sufficient resources to support its deployment, leading to a decrease in reception rate. Conversely, when `dis` is smaller, SFC migrations are more frequent. In this case, previously deployed SFCs migrate to servers farther from the MEC center, freeing up resources for new requests and thus improving the SFC reception rate. For JSDR-VDSMA, the SFC reception rate also decreases as the migration condition `dis` increases. In MEC networks, user QoS is interdependent, so the maximum number of users each base station can accommodate is limited. As `dis` increases, access resources near the MEC center are rapidly consumed. When a new SFC request arrives at the MEC center, insufficient resources are available to support its deployment, resulting in a decrease in SFC reception rate. An interesting observation is that, as shown in Figure 5(a), JSDM-VDSMA consistently outperforms JSDR-VDSMA in SFC reception rate under all migration conditions. This superior performance is attributed to JSDM-VDSMA's ability to not only switch access nodes but also migrate VNFs when the migration threshold is reached, thus ensuring optimal SFC performance. In contrast, JSDR-VDSMA only switches access nodes without migrating VNFs, resulting in relatively lower performance.

[0180] Figure 5(b) illustrates the end-to-end communication latency, showing that latency increases with the number of SFC requests. This is because the increased number of received SFCs leads to increased access latency, processing latency, and link latency. Furthermore, Figure 5(b) also shows that the latency of both the JSDM-VDSM and JSDR-VDSM algorithms increases with the migration condition dis. This is because a higher dis reduces the frequency of SFC migrations, meaning that when users move to more distant locations, they still communicate with the original access node and previously deployed SFCs, resulting in higher access latency. Therefore, an increase in dis leads to an increase in overall latency.

[0181] Under different migration conditions, JSDM-VDSMA consistently outperforms JSDR-VDSMA in terms of latency. This is because JSDM-VDSMA not only switches access nodes during SFC migration but also migrates VNFs, providing more high-quality resources for the newly arrived SFC that needs to be deployed, thereby reducing overall latency and showing better performance compared to JSDR-VDSMA.

[0182] Figures 5(c), (d), (e), and (f) show the average number of migrations, average migration cost, average placement cost of VNF dependent components, and average interruption latency of SFC, respectively. The formula for interruption latency is defined in (13). As shown in Figure 5(c), under different migration conditions, the average number of migrations of VNF dependent components in the JSDM-VDSM and JSDR-VDSM algorithms is comparable, approximately half the length of the SFC. Figure 5(d) shows that under different migration conditions, the average migration cost of VNF dependent components in the JSDM-VDSM and JSDR-VDSM algorithms is also similar. The migration cost of VNF dependent components is defined as the shortest path for migrating VNF dependent components. In Figure 5(e), it is evident that under different migration conditions, JSDM-VDSMA consistently outperforms JSDR-VDSMA in terms of VNF dependent component placement cost. Finally, Figure 5(f) shows that the interruption latency of JSDM-VDSMA remains unchanged under different migration conditions. This is because, by definition, the calculation of interruption latency is independent of migration conditions (i.e., ...). and ).

[0183] Figures 5(g), (h), and (i) illustrate the total number of migrations, total migration cost, and total placement cost of VNF dependent components, respectively. These results clearly demonstrate that the JSDM-VDSM algorithm consistently outperforms the JSDR-VDSM algorithm under different migration conditions. Furthermore, the total number of migrations, total migration cost, and total placement cost of VNF dependent components increase significantly with the increase in the number of SFC requests.

[0184] (2) Performance analysis at different moving speeds

[0185] Figure 6 illustrates the performance evaluation of the JSDM-VDSM and JSDR-VDSM algorithms under different migration conditions and user mobility speeds. In this evaluation, the number of SFCs is fixed at 240, and the length of each SFC is 3. Figure 6(a) shows that under various migration conditions, JSDM-VDSMA consistently outperforms JSDR-VDSMA in terms of SFC reception rate.

[0186] For JSDM-VDSMA, when dis is set to 1 or 2, the SFC (Signal Found) reception rate increases with increasing user mobility speed. This is because higher speeds increase the likelihood of meeting the migration condition (dis = 1 or 2), thus increasing the chance of SFC migration. SFC migration frees up resources that can be allocated to newly arriving SFC requests, thereby improving the reception rate. Conversely, when dis = 3, the reception rate remains almost constant. This is because even at higher mobility speeds, it is difficult for an SFC to meet the dis = 3 condition within its finite lifetime, therefore the probability of SFC migration is low at all mobility speeds, and the reception rate remains stable. Similarly, for JSDR-VDSMA, when dis = 1, the reception rate increases with increasing user mobility speed. However, when dis = 2 or 3, the reception rate remains almost constant for similar reasons to JSDM-VDSMA.

[0187] Figure 6(b) illustrates the end-to-end latency, revealing that JSDM-VDSMA consistently outperforms JSDR-VDSMA in reducing latency under various constraints. For both JSDM-VDSM and JSDR-VDSMA algorithms, when dis=1, latency decreases with increasing user movement speed. This is because higher speeds make it easier to satisfy the migration condition dis=1, allowing SFC migration. This allows VNFs to migrate to more suitable nodes and obtain appropriate access nodes, thus reducing latency. However, when dis=3, latency increases with increasing speed. This is because even under high-speed conditions, it is still difficult to satisfy the threshold dis=3. Therefore, users continue to connect through their original access nodes despite greater distances, leading to increased access latency and a rise in end-to-end latency. When dis=2, the latency of both algorithms is relatively high under medium-speed conditions. This is because the probability of satisfying the migration condition dis=2 is lower compared to high-speed movement. Furthermore, the distance between the user's current location and the access node is larger compared to low-speed movement, resulting in higher access latency. Therefore, in this medium-speed scenario, end-to-end latency increases significantly.

[0188] (3) Performance analysis as SFC length changes

[0189] Different services require different VNF ​​functionalities, therefore the length of the SFC (Service Controller) will vary depending on the service requirements. For example, some services may need more complex VNFs, while others may only need simpler VNFs. An effective algorithm should be adaptable to multiple services, ensuring flexibility in VNF placement and SFC configuration to meet the specific needs of various services. Figure 7 shows the performance evaluation of the JSDM-VDSM and JSDR-VDSM algorithms under different user mobility speeds and different SFC lengths. In this evaluation, the migration condition dis is fixed at 2, while the number of SFCs remains at 240.

[0190] Figure 7(a) shows that the reception rate of both algorithms decreases as the SFC length increases. This is because longer SFCs consume more resources, thus reducing the number of SFCs that can be received. Furthermore, it can be observed that JSDM-VDSMA achieves a higher reception rate at higher migration speeds for different SFC lengths. This is because higher speeds increase the likelihood of meeting the migration condition (dis=2), allowing the release of occupied resources through migration. These released resources can be used to deploy newly arriving SFC requests. For JSDR-VDSMA, the reception rate is relatively high at higher speeds when the SFC length is 2 or 3. However, when the SFC length increases to 4, the reception rate decreases at high speeds. This is because even if the migration condition (dis=2) is met, JSDR-VDSMA only switches access points without migrating VNFs. Longer SFCs consume more resources, resulting in a reduced number of SFCs received. Overall, JSDM-VDSMA consistently outperforms JSDR-VDSMA in terms of reception rate.

[0191] Figure 7(b) shows that the end-to-end latency of both algorithms increases with the increase of the SFC length. This is because a longer SFC leads to more VNF processing latency and link latency. When the SFC length is 2, the two algorithms perform similarly in terms of latency. For SFC lengths of 3 or 4, JSDM-VDSMA significantly outperforms JSDR-VDSMA in reducing latency under high-speed conditions. This is because at higher speeds, the migration condition (dis=2) is more easily met, and the migration-based method of JSDM-VDSMA reduces latency more effectively than the rerouting method of JSDR-VDSMA.

[0192] In summary, the experimental results show that, compared with rerouting-based algorithms, the method described in this invention consistently maintains superior performance in terms of average service function link rate, average end-to-end latency, and average resource usage cost, especially when users are moving at high speeds.

Claims

1. A method for deploying and migrating SFC based on VNF-dependent components, wherein the method is based on mobile terminal users submitting service request information to the SDN controller, the SDN controller analyzing the user's demand type and resource requirements based on the service request information, including bandwidth resources, computing resources, storage resources, and latency requirements, and simultaneously combining network topology, edge server resource status, and network operation status, using deep reinforcement learning technology to optimize the deployment and migration of SFC, characterized in that: Includes the following steps: (1) Construct a communication architecture that includes mobile users, edge servers and SDN controllers; In this communication architecture, mobile users generate dynamic service requirements, including computing tasks, data storage, and communication services. Edge servers are deployed close to the users to provide computing and communication services. The SDN controller monitors the mobile users' service requirements and the resource status of the edge servers in real time and dynamically adjusts resource allocation and scheduling strategies based on network topology information. (2) Establish user request model, communication model and computing model to describe the operation mechanism and performance indicators of the communication system; The user request model includes SFC deployment requests and migration requests. Deployment requests describe the user's requirements for bandwidth resources, computing resources, memory resources, and storage resources, while migration requests are used to transfer the currently deployed SFC to the target edge server when the user moves. The communication model defines a bandwidth allocation strategy for edge servers and calculates transmission rate and data transmission time based on Shannon's formula and channel interference index, thereby achieving a reasonable allocation of communication resources. The computational model includes end-to-end communication latency, resource usage cost, and interruption latency during SFC migration, which is used to evaluate service performance and provide theoretical support for communication system resource optimization and scheduling. (3) Based on the network topology information, user request model, communication model and computing model of the communication system, formulate a resource scheduling optimization problem and clarify the optimization objective; The optimization problems include maximizing the service access rate of user requests, minimizing end-to-end communication latency and SFC migration costs, and maximizing resource utilization. During the SFC migration process, optimize the migration path of VNF and the resource cost of dependent components to reduce service interruption time caused by migration. (4) Consider the deployment and migration of SFC based on deep reinforcement learning for VNF dependent components; The SDN controller collects user requests, edge server resource status, and network topology information in real time, and uses deep reinforcement learning algorithms to dynamically make decisions on the following: User access strategy: Intelligently select the optimal edge server for access based on the user's geographical location and business needs; SFC Deployment and Migration Strategy: Dynamically adjust the SFC deployment and migration path by comprehensively considering resource utilization, communication latency, and migration latency; VNF Dependency Component Migration Strategy: For VNF dependencies in SFC, prioritize migrating critical VNF components to ensure the integrity and continuity of the service chain.

2. The SFC deployment and migration method based on VNF dependent components according to claim 1, characterized in that, The user request model, communication model, and calculation model mentioned in step (2) are as follows: (21) User Request Model User request r is represented as SFC r. It is a quintuple of SFC r, where b r Defined as bandwidth requirement, Indicates communication latency requirements, TTL r This refers to the lifecycle of an SFC (Self-Functional Graph), which is described as a directed graph G. r =(F r E r ),in Let r represent the set of VNFs in SFC, and let f be any VNF. rh All require CPU resources and memory resources E r This represents the virtual link of SFC r, and the arrival time of SFC r is expressed as... m represents the number of time slots, and Δ represents the length of a time slot; For SFC deployment, use decision variables Indicates whether to pass through edge server node n i Access to the network Used to indicate whether VNFf rh Placed on edge server node n i superior, It is a binary variable used to represent a logical link. Is it mapped to link e? ij The above is mathematically represented as: For SFC migration, when the VNF is deployed to edge server node n i Or migrate to edge server node n i When n i If the component resources required to support VNF ​​operation are lacking, the VNF dependent components must be migrated from other nodes to the edge server node n. i This is to ensure the proper functioning of the VNF; (22)Communication model edge server node n i The bandwidth resources are evenly allocated to SFC r to ensure fair allocation, and its mathematical expression is as follows: in, Indicates from edge server node n i The SFC set accessed Indicates server node n i bandwidth resources on Indicates r j From server node n i Accessing the network, if from n i 1 indicates network access; otherwise, 0 indicates accessibility. Use parameters This indicates that SFC r accesses edge server node n. i The channel interference experienced is caused by channel contention from other SFCs simultaneously accessing the same node. SFC r j Transmission capability Indicates channel gain, affected by SFC r j With edge server node n i The mathematical expression for the distance between them and the power gain effect caused by small-scale fading in the channel is: Based on this, SFC r and server node n i The transmission rate between them is obtained using the following formula: Where, ρ 2 It refers to the signal-to-noise ratio and data transmission delay. This refers to the transmission of the data to be transmitted from SFC r to the service node n. i The required time is expressed as: (23) Computational Model SFC r's end-to-end communication delay D r Defined as the sum of data transmission delay, VNF processing delay, and link propagation delay, as shown below: The deployment cost of SFC is defined as follows: in, It is the migration component ψ q The path used It is a binary variable representing a path. on It is mapped to link e ij superior, Represents n i unit storage cost, M r This represents the set of VNFs that need to be migrated in SFC r.

3. The SFC deployment and migration method based on VNF dependent components according to claim 2, characterized in that, This method minimizes downtime and enhances success rate through pre-copy migration. Downtime is defined as the duration of the final stage of the migration process, during which the service is temporarily interrupted to complete final data synchronization. Indicates the start time of the interruption period. The expressions for the two, representing the end time of the interruption period, are as follows: in, This represents the transition time in the q-th iteration. This is the actual number of iterations. Indicates the VNF to be migrated f rh memory page size, VNFf rh The migration rate, therefore, the interrupt time is defined as follows: M in the formula r This represents the set of VNFs that need to be migrated in SFC r.

4. The SFC deployment and migration method based on VNF dependent components according to claim 1, characterized in that: The optimization problem and optimization objective described in step (3) are as follows: Constraint 1: Each f rh A VNF instance can only be successfully deployed on one server node at most. VNF instances are indivisible and are represented as follows: Constraint 2: The communication quality between users connected to the network through the same node is mutually influential, and each physical node can accommodate a maximum of n. max To maintain service quality for all users, the mathematical representation is: Constraint 3: On any server node n i Above, the total CPU and memory resource requirements for SFC deployment and migration must not exceed the maximum CPU and memory resource capacity, mathematically expressed as: Constraint 4: The used storage resources will not exceed the maximum storage resource capacity, mathematically expressed as: in, It is node n i A collection of components on; Constraint 5: Any link e ij The total bandwidth consumption must be less than the maximum bandwidth capacity, which can be expressed mathematically as: The first item represents the link bandwidth used for SFC deployment, and the second item represents the bandwidth used for SFC migration. Constraint 6: The end-to-end delay of any SFC r must satisfy the following constraint: Based on the constraints defined above, the JSDM-VDSM problem includes SFC deployment strategy, SFC migration strategy, and VNF dependent component migration strategy. The goal of this method is to maximize QoS by improving service reception rate, reducing resource usage costs, and minimizing communication latency and interruption latency. The mathematical definition of the optimization objective is as follows: Among them, Z r Z(t) is a decision variable used to determine whether the SFC request r is accepted; if accepted, Z... r (t) is set to 1, otherwise it is set to 0; w1, w2, w3 and w4 in the formula are the weights of different indicators.

5. The SFC deployment and migration method based on VNF dependent components according to claim 1, characterized in that: Step (4) also includes: The network architecture identifies SFCs that need to be orchestrated, including SFCs that need to be deployed and SFCs that are planned to be migrated, and views SFC migration as a joint optimization of VNF redeployment and SFC rerouting. Obtain the status of the environment and the SFC currently to be orchestrated, including the available CPU, memory, storage, bandwidth and component resources of the physical nodes; the CPU, memory and component resource requirements of the VNFs to be deployed; the number of VNFs that have not yet been placed in the deployed SFC; the partial communication latency of the current SFC; and the lifespan of the SFC. The obtained state s ι Input to Actor old Network, obtain VNF deployment decisions a ι Subsequently, iterative interactions with the environment were conducted to collect a set of experimental data [s] ι ,a ι ,r ι ,s ι+1 ] and store it in memory; Calculate the advantage function based on experimental data. Used to update the Critic network; uses a truncated objective function. The Actor network is updated, and after multiple iterations, the parameters of the Actor network are passed to the Actor. old The network iterates continuously until it converges. The solution process for the JSDM-VDSM problem includes obtaining the SFC request r and determining the user access node. and SFC deployment or migration strategies Based on strategy and Obtain VNF dependency component migration strategy Before learning the service delivery strategy, the SDN controller collects global information and releases SFCs that have reached the end of their lifecycle, and then identifies the SFCs that need to be deployed or migrated. The SFC request for deployment is an SFC request that arrives randomly within the current time slot t, while the SFC request for migration is an SFC request that meets the migration criteria within the same time slot. If there are no requests to process, the SDN controller will move to the next time slot and continue to acquire requests until r. batch Not empty; then, r batch The SFC requests in the process will be deployed or migrated sequentially. For deployment and migration requests, the user area associated with the SFC request is first determined. Then, the SDN controller evaluates whether the base station in the current area has sufficient capacity. If there is sufficient capacity, access is made through that base station. If there is insufficient capacity, the SDN controller plans to connect through neighboring areas. First, neighboring areas are identified, and then the access capacity of these areas is calculated. Next, the accessible node max_num_node with the maximum access capacity is determined, and the area that can accommodate the maximum capacity is selected. Then, an access node is selected from max_num_node with probability 1-∈ and an access node is selected from the set n_max_num_node with probability ∈, where n_max_num_node is the set of accessible nodes. After selecting the access point, determine the deployment or migration strategy for the SFC and the migration strategy for VNF dependent components. First, identify the first VNF ​​of the SFCr, then obtain the current state, and proceed according to the strategy π. θold (a ι ,s ι ), determine VNFf rh The placement position of f rh Can be placed on node a ι Up, but a ι Currently does not have f rh To run the required components, the required f must be... rh The dependent components are migrated to a ι Then place the next VNF ​​and record the updated state s. ι+1 ; The above process continues until the SFC is successfully deployed or migrated; rewards are calculated based on the deployment results: if deployment is successful, a reward for successful deployment is calculated; if migration is successful, a reward for successful migration is calculated; if SFC deployment or migration fails, the communication system will revert to the last successfully achieved state. ι-u+1 .

Images