A cross-layer service protection switchover method and device

Abstract

A cross-layer service protection switching method and device, the method comprising: performing resource mapping on optical layer resources and electrical layer resources of a cross-layer service to construct an optical-electrical collaborative resource pool; if a fault type of the cross-layer service is an optical layer fault, determining a protection switching strategy according to a resource state of the optical-electrical collaborative resource pool through a pre-established reinforcement learning intelligent decision algorithm, and performing a corresponding protection switching action; the protection switching strategy comprises an optical layer rerouting strategy, an electrical layer protection switching strategy, and an optical-electrical collaborative protection switching strategy. The method realizes unified scheduling and sharing of optical layer and electrical layer resources, and improves network resource utilization. The pre-established reinforcement learning intelligent decision algorithm is used to determine the protection switching strategy, so that the strategy decision comprehensively considers the resource availability of the optical layer and the electrical layer, can select the optimal strategy according to the resource state of the optical layer and the electrical layer, and effectively improves the success rate of cross-layer service fault recovery.

Description

Technical Field

[0001] This application relates to the field of computer network communication technology, specifically to a cross-layer service protection switching method and apparatus. Background Technology

[0002] With the rapid development of new business scenarios such as artificial intelligence, cloud computing, 5G, and the Industrial Internet, modern OTN (Optical Transport Network) faces severe challenges in carrying dynamic, differentiated, and highly reliable services. Traditional networks adopt a layered and independent service creation architecture: the optical layer relies on hardware-based fixed switching (such as channel protection) to achieve millisecond-level rapid recovery from service failures, but with low resource utilization; the electrical layer provides flexible recovery through dynamic rerouting, but with slow response speeds and uncertain success rates and recovery times. To combine the advantages of both, the industry initially used generalized multi-protocol label switching to achieve limited information exchange and introduced path calculation units for cross-layer routing. In recent years, software-defined network architectures have further obtained the entire network topology through a centralized controller, attempting to coordinate multi-layer protection actions in the management plane or set timers to avoid switching conflicts. However, existing technologies have not yet broken through the layered control paradigm, resulting in high control signaling overhead, weak real-time decision-making, and the isolation of optical and electrical layer resources in planning and scheduling, making global optimization difficult. Currently, while collaborative solutions based on centralized controllers and software-defined networks offer logically unified management, in actual deployments, their decision-making loop relies on periodic information collection and configuration distribution, resulting in significant "execution gaps" that cannot meet the millisecond-level real-time requirements of optical layer protection. Simultaneously, multi-layer resources (wavelength, time slots, tags) remain under separate management, lacking a unified resource abstraction and dynamic programming interface, leading to rigid protection strategies and inflexible resource utilization.

[0003] In traditional OTN networks, both the optical and electrical layers operate on a PC (Permanent Connection) basis. With the addition of a control plane, these layers are often deployed independently, hindering the implementation of hybrid optical-electrical collaborative networking. Cross-layer services only support a single control plane, and optical-electrical layer protection resources are reserved independently. When resources in one layer are scarce, resources in the other layer remain idle, leading to waste. Protection failover methods are simplistic, only monitoring the operational status of a single layer during failover, failing to provide overall coordination and resulting in suboptimal service performance after failover. Especially in scenarios with multiple or complex failures, existing mechanisms struggle to achieve rapid rerouting through optical-electrical collaboration, potentially exceeding the service SLA (Service Level Agreement) requirements. Summary of the Invention

[0004] This application provides a cross-layer service protection switching method and apparatus, which can solve the technical problems in existing optical transport networks caused by the independence and lack of coordination between optical and electrical layer resources, resulting in low resource utilization, single protection switching method, and low recovery success rate in complex fault scenarios.

[0005] To achieve the above objectives, in a first aspect, this application provides a cross-layer service protection switching method, the method comprising: By mapping the optical and electrical layer resources of cross-layer services, a photoelectric collaborative resource pool is constructed.

[0006] If the fault type of the cross-layer service is an optical layer fault, a protection switching strategy is determined by a pre-established reinforcement learning intelligent decision-making algorithm based on the resource status of the optoelectronic collaborative resource pool, and the corresponding protection switching action is executed; the protection switching strategy includes an optical layer rerouting strategy, an electrical layer protection switching strategy, and an optoelectronic collaborative protection switching strategy.

[0007] Furthermore, in one embodiment, the step of mapping the optical layer resources and electrical layer resources of cross-layer services to construct an optoelectronic collaborative resource pool includes: The optical layer wavelength resources are mapped to the optical multiplexing segment layer topology, and the electrical layer time slot resources are mapped to the optical channel layer layer topology.

[0008] The mapped optical layer resources are abstracted into multiple optical layer resource objects, and the mapped electrical layer resources are abstracted into multiple electrical layer resource objects.

[0009] A resource association model for optical layer resource objects and electric layer resource objects is established, and the optical-electric collaborative resource pool is composed of optical layer resource objects, electric layer resource objects and the resource association model of the entire network.

[0010] Furthermore, in one embodiment, if the fault type of the cross-layer service is an electrical layer fault, an electrical layer protection switching action is performed.

[0011] Furthermore, in one embodiment, based on the resource status of the optoelectronic collaborative resource pool, a protection switching strategy is determined through a pre-established reinforcement learning intelligent decision-making algorithm, including: The state space is initialized based on the optical layer wavelength resource status, electrical layer time slot resource status, and link status of the entire network topology of the optoelectronic collaborative resource pool. Features are extracted from the state space according to the reinforcement learning intelligent decision-making algorithm, and protection switching strategy is output by combining cross-layer service configuration information and experience replay mechanism.

[0012] Furthermore, in one embodiment, the step of combining cross-layer service configuration information and experience replay mechanism to output protection switching strategy includes: Based on the configuration information, if a pre-configured optical layer path exists, an optical layer rerouting strategy is output; otherwise, the resource availability of the optical layer and electrical layer is calculated respectively, and the historical success rates of the optical layer rerouting strategy and the electrical layer protection switching strategy are obtained through the experience replay mechanism.

[0013] If the resource availability and historical success rate of the optical layer exceed their respective thresholds, and the resource availability and historical success rate of the electrical layer also exceed their respective thresholds, then the strategy with the higher historical success rate will be adopted.

[0014] If only the optical layer meets the threshold condition, then the strategy of the optical layer is adopted; if only the electrical layer meets the threshold condition, then the strategy of the electrical layer is adopted.

[0015] If neither the optical layer nor the electrical layer meets the threshold conditions, an optoelectronic collaborative protection switching strategy is output.

[0016] Furthermore, in one embodiment, performing the corresponding protection switching action includes: For the optical layer rerouting strategy, if an optical layer pre-fabricated path exists, cross-layer services will be switched from the faulty path to the optical layer pre-fabricated path; if no optical layer pre-fabricated path exists, the optical transport network control plane will be triggered to re-find the path and perform dynamic rerouting.

[0017] For the electric layer protection switching strategy, the electric layer protection switching action is executed.

[0018] For the optoelectronic coordinated protection switching strategy, the protection switching actions corresponding to the optical layer rerouting strategy and the electrical layer protection switching strategy are executed respectively.

[0019] Furthermore, in one embodiment, if the dynamic rerouting fails to successfully restore the cross-layer service, the electrical layer protection switching action is performed.

[0020] Furthermore, in one embodiment, the electrode layer protection switching operation includes: In the case of pre-prepared electrical paths in the cross-layer services, pre-prepared path switching is performed first. If pre-prepared path switching fails, the optical transport network control plane is triggered to re-find the route and perform dynamic rerouting.

[0021] For port-level failures, the transmission path of the cross-layer service is switched from the working port to the protection port.

[0022] Furthermore, in one embodiment, the reinforcement learning intelligent decision-making algorithm further includes a reward function, which includes a positive reward for successful cross-layer service fault recovery and a negative reward for unsuccessful cross-layer service fault recovery or insufficient link resources.

[0023] Secondly, this application provides a cross-layer service protection switching device, the device comprising: The resource pool construction module is used to map optical and electrical layer resources of cross-layer services to build an optoelectronic collaborative resource pool.

[0024] The protection switching module is used to determine the protection switching strategy based on the resource status of the optoelectronic collaborative resource pool and a pre-established reinforcement learning intelligent decision-making algorithm if the fault type of the cross-layer service is an optical layer fault, and to execute the corresponding protection switching action. The protection switching strategy includes an optical layer rerouting strategy, an electrical layer protection switching strategy, and an optoelectronic collaborative protection switching strategy.

[0025] The beneficial effects of the technical solutions provided in this application include: This application breaks the traditional model of independent management of optical and electrical layer resources by mapping optical and electrical layer resources for cross-layer services and constructing an optoelectronic collaborative resource pool. This achieves unified scheduling and sharing of optical and electrical layer resources, effectively avoiding idle waste caused by independent resource reservation and improving the overall network resource utilization. If the fault type of the cross-layer service is an optical layer fault, a pre-established reinforcement learning intelligent decision-making algorithm is used to determine the protection switching strategy based on the resource status of the optoelectronic collaborative resource pool, and the corresponding protection switching action is executed. This allows the strategy decision to comprehensively consider the resource availability of both the optical and electrical layers, thus ensuring that the protection switching method is no longer limited to a single optical or electrical layer recovery method. It can select the optimal strategy based on the resource status of both the optical and electrical layers. Furthermore, in complex fault scenarios such as multi-point faults, it can dynamically avoid risks and select recovery paths with high success rates, thereby effectively improving the success rate of cross-layer service fault recovery and ensuring high service reliability. Attached Figure Description

[0026] Figure 1 This is a flowchart of the cross-layer service protection switching method according to an embodiment of this application.

[0027] Figure 2 This is a flowchart illustrating the establishment of optoelectronic resource mapping and cross-layer service connections in an embodiment of this application.

[0028] Figure 3 This is a flowchart of the reinforcement learning intelligent decision-making algorithm in an embodiment of this application.

[0029] Figure 4 This is a flowchart illustrating the detailed steps of cross-layer service protection switching in an embodiment of this application.

[0030] Figure 5 This is a block diagram of a cross-layer service protection switching device according to an embodiment of this application.

[0031] Figure 6 This is a detailed architectural diagram of the cross-layer service protection switching device according to an embodiment of this application. Detailed Implementation

[0032] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0033] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0034] Firstly, embodiments of this application provide a method for cross-layer service protection switching.

[0035] In one embodiment, see Figure 1 As shown, the above-mentioned cross-layer service protection switching method includes: S1. Map the optical and electrical layer resources of cross-layer services to build an optoelectronic collaborative resource pool.

[0036] S2. If the fault type of the cross-layer service is an optical layer fault, the protection switching strategy is determined by a pre-established reinforcement learning intelligent decision-making algorithm based on the resource status of the optoelectronic collaborative resource pool, and the corresponding protection switching action is executed. The protection switching strategies include optical layer rerouting strategy, electrical layer protection switching strategy and optoelectronic collaborative protection switching strategy.

[0037] In this embodiment, the optical transport network management plane first receives the service creation parameters selected by the user. Then, it maps the optical layer wavelength resources and electrical layer time slot resources of the cross-layer service to a unified map and constructs an optoelectronic collaborative resource pool. Based on this, the management plane creates the cross-layer service according to the available resource status in the resource pool. This cross-layer service includes optical layer intelligent services, electrical layer intelligent services, optical layer traditional services, and electrical layer traditional services. Parameter control allows the creation of either PC (Permanent Connection) or SPC (Soft Permanent Connection) services. PC services only support 1+1 protection, while SPC services support dynamic rerouting, rerouting 1+1, and permanent 1+1 protection types. To avoid users creating optical layer and electrical layer services in layers, a task management approach is adopted. Users only need to specify the type and source / destination information of the client layer service to be created to create optical layer, electrical layer, and client layer services with one click, binding them into an end-to-end cross-layer service.

[0038] In this embodiment, by uniformly mapping the optical layer wavelength resources and electrical layer time slot resources of cross-layer services and constructing an optoelectronic collaborative resource pool, the inherent architecture of independent management and mutual isolation of optical and electrical layer resources in traditional optical transport networks is broken. This allows the management plane to grasp the occupancy status, available capacity, and correlation of optical and electrical layer resources in real time from a global perspective. This enables on-demand dynamic allocation and sharing of resources during cross-layer service creation and fault switching, effectively avoiding resource idleness and waste caused by independent layer reservation. Furthermore, for optical layer fault scenarios, instead of simply triggering fixed hardware switching, a pre-established reinforcement learning intelligent decision-making algorithm is invoked based on the real-time and complete resource status information provided by the collaborative resource pool. This algorithm comprehensively considers the idle status of optical layer wavelengths, the fragmentation degree of electrical layer time slots, the overall network topology link health status, and successful patterns learned through experience replay from historical switching records. It dynamically outputs multiple strategies such as optical layer rerouting, electrical layer protection switching, or optoelectronic collaborative protection switching, and executes corresponding switching actions according to the strategies.

[0039] Therefore, this method can achieve millisecond-level fast switching using optical layer hardware when a pre-built path exists, and can flexibly invoke electrical layer rerouting or optoelectronic coordination mechanisms for recovery when a pre-built path is missing or the fault is complex. It achieves the complementary advantages of the "fast" optical layer and the "accurate" electrical layer, significantly improving the success rate of cross-layer service recovery and the efficiency of global resource utilization in complex scenarios such as multi-point faults and shared risk link groups. At the same time, it reduces invalid switching and network oscillations through intelligent decision-making, ensuring the satisfaction of service level agreements.

[0040] Furthermore, in one embodiment, in step S1 above, the optical layer resources and electrical layer resources of cross-layer services are mapped to construct an optoelectronic collaborative resource pool. The specific steps are as follows: The optical layer wavelength resources are mapped to the optical multiplexing segment layer topology, and the electrical layer time slot resources are mapped to the optical channel layer layer topology.

[0041] The mapped optical layer resources are abstracted into multiple optical layer resource objects, and the mapped electrical layer resources are abstracted into multiple electrical layer resource objects.

[0042] Establish a resource association model for optical layer resource objects and electric layer resource objects, and form an optoelectronic collaborative resource pool consisting of optical layer resource objects, electric layer resource objects, and resource association models of the entire network.

[0043] The aforementioned optical layer resource objects include optical layer resource types (such as fiber links, wavelengths, optical cross-connect resources, etc.), capacity, occupancy status, nodes, and link attributes. The aforementioned electrical layer resource objects include electrical layer resource types (such as electrical layer cross-connect matrices, electrical layer processing units, electrical layer ports, time slots, etc.), occupancy status, nodes, and link attributes. Mapping strategies include direct optical layer mapping, direct electrical layer mapping, and optoelectronic collaborative mapping. Direct optical layer mapping is suitable for large-granularity services, such as 100G and above, directly establishing end-to-end services at the optical layer. Direct electrical layer mapping is suitable for small-granularity services, achieving service aggregation and scheduling through electrical layer time slot switching. Optoelectronic collaborative mapping uses intelligent algorithms to jointly optimize optical and electrical layer resources, dynamically selecting based on service requirements.

[0044] In this embodiment, optical layer wavelength resources and electrical layer time slot resources are mapped, and optoelectronic resources are uniformly incorporated into the optoelectronic collaborative resource pool for management. For small-granularity services, direct electrical layer mapping is preferred, while for large-granularity services, direct optical layer mapping is preferred. For services that require dynamic bandwidth adjustment, direct optical layer mapping and direct electrical layer mapping are dynamically switched according to the current network resource utilization.

[0045] The specific steps for mapping optical layer wavelength resources to the hierarchical topology of the optical multiplexing section are as follows: The management plane of the optical transport network collects wavelength resource information from optical layer network elements through interfaces or internal protocols. For each optical fiber link, each wavelength it carries is treated as an independent resource item and registered to the link corresponding to the OMS (Optical Multiplex Section) layer topology. The data structure of each wavelength resource item includes at least: resource type, capacity, occupancy status, belonging node, and belonging OMS link ID (Identifier).

[0046] The specific steps for mapping electrical layer time slot resources to optical channel layer hierarchical topology are as follows: The management plane of the optical transport network collects ODUk (Optical Data Unit - level k) time slot resource information from electrical layer network elements. For each electrical layer processing unit or cross-connect matrix, the time slots it supports are registered as independent resource items to the corresponding node or link in the OCh (Optical Channel) layer topology. The data structure of each time slot resource item includes at least: resource type, time slot granularity, occupancy status, belonging node, and the service identifier it carries.

[0047] The resource association model described above maintains a resource association table. Each record in this table contains a unique identifier (such as a wavelength ID) for an optical layer resource object and unique identifiers (such as a list of time slot IDs) for one or more associated electrical layer resource objects. When the management or control plane of the optical transport network performs cross-layer service creation or protection switching, it can quickly locate the status of electrical layer resources bound to optical layer resources by querying this resource association model, or find available optical layer bearer resources based on electrical layer resource requirements, thereby achieving coordinated scheduling and status synchronization of optoelectronic resources.

[0048] Furthermore, a specific implementation example of an optoelectronic resource mapping and cross-layer business connection establishment process is provided, see [link to relevant documentation]. Figure 2 As shown, the detailed steps are as follows: A1. Associate the optical layer wavelength resources with the OMS hierarchical topology.

[0049] A2. Associate the electrical layer wavelength resources with the OCh layered topology.

[0050] A3. Abstraction and modeling of optoelectronic collaborative resources.

[0051] A4. Mapping strategy selection: Determine the mapping method. If it is direct mapping of the optical layer, proceed to step A5; if it is direct mapping of the electrical layer, proceed to step A8; if it is optoelectronic co-mapping, proceed to step A11.

[0052] A5. Wavelength Resource Status Check. In this step, intelligent services can only use wavelength resources reserved in the optical transport network control plane, while traditional services can only use wavelength resources reserved in the optical transport network management plane. If a wavelength resource is already occupied by an intelligent service, its status will show as "Control Plane Occupied"; if a wavelength resource is already occupied by a traditional service, its status will show as "Management Plane Occupied".

[0053] A6. Wavelength Resource Allocation. The default state of wavelength resources is reserved for the control plane. If traditional services need to be created, wavelength resources need to be allocated to the management plane.

[0054] A7. Establish end-to-end OCh service.

[0055] A8. Time Slot Resource Status Check. In this step, intelligent services can only use time slot resources reserved in the optical transport network control plane, while traditional services can only use time slot resources reserved in the optical transport network management plane. If a time slot resource is already occupied by an intelligent service, its status will show as "Control Plane Occupied"; if a time slot resource is already occupied by a traditional service, its status will show as "Management Plane Occupied".

[0056] A9. Time Slot Resource Allocation. The default state of time slot resources is reserved for the control plane. If traditional services need to be created, the time slot resources need to be allocated to the management plane.

[0057] A10. Establish end-to-end ODUk services.

[0058] A11. Optoelectronic Hybrid Resource Status Check. This step mainly checks whether the optical layer resources and electrical layer resources are available. Cross-layer services can only be created if both the optical and electrical layers are available simultaneously; otherwise, layered creation is the only option.

[0059] A12. Coordinated Allocation of Optoelectronic Resources. If a cross-layer service is created, the resource status needs to be assigned to the cross-layer resource reservation status.

[0060] A13. Establish end-to-end cross-layer services. When creating end-to-end cross-layer services, if there are reusable optical layer services, the existing optical layer services should be reused first.

[0061] A14. Topology updates and resource synchronization allow users to monitor the resource usage of the entire network in real time.

[0062] It is understandable that the order of steps A1 and A2 above can be interchanged, or they can be performed simultaneously.

[0063] Furthermore, in one embodiment, step S2 above further includes: if the fault type of the cross-layer service is an electrical layer fault, performing an electrical layer protection switching action.

[0064] Furthermore, in one embodiment, in step S2 above, if the fault type of the cross-layer service is an optical layer fault, the protection switching strategy is determined based on the resource status of the optoelectronic collaborative resource pool through a pre-established reinforcement learning intelligent decision-making algorithm. The specific steps are as follows: The state space is initialized based on the optical layer wavelength resource status, electrical layer time slot resource status, and link status of the entire network topology in the optoelectronic collaborative resource pool. Features are extracted from the state space using a reinforcement learning intelligent decision-making algorithm, and protection switching strategies are output by combining cross-layer service configuration information and experience replay mechanism.

[0065] The link status can be obtained through the link management protocol or the optical monitoring channel.

[0066] In this embodiment, the reinforcement learning intelligent decision-making algorithm includes a state space and an action space. The state space includes the optical layer wavelength resource state, the electrical layer time slot resource state, and the link state, which serve as the inputs to the reinforcement learning intelligent decision-making algorithm. The action space includes optical layer rerouting actions, electrical layer protection switching actions, and optoelectronic coordinated protection switching actions, which serve as the outputs of the reinforcement learning intelligent decision-making algorithm.

[0067] In this embodiment, the state space is initialized based on the optical layer wavelength resource status, electrical layer time slot resource status, and the link status of the entire network topology in the optoelectronic collaborative resource pool. A reinforcement learning intelligent decision-making algorithm is used to extract features from the state space. Simultaneously, the configuration information of cross-layer services and an experience replay mechanism are combined to output protection switching strategies. This allows the decision-making process at the moment of failure to fully perceive the real-time resource status of the optical and electrical layers, link health, and historical successful switching experiences. This avoids decision-making biases caused by relying solely on local information or fixed rules, effectively improving the accuracy and adaptability of the protection switching strategy. It provides a reliable basis for subsequent execution of optical layer rerouting, electrical layer protection switching, or optoelectronic collaborative switching. Furthermore, the experience replay mechanism continuously optimizes decision quality, improving the recovery success rate of cross-layer services in complex fault scenarios.

[0068] Furthermore, in one embodiment, the reinforcement learning intelligent decision-making algorithm further includes a reward function, which includes a positive reward for successful cross-layer service fault recovery and a negative reward for unsuccessful cross-layer service fault recovery or insufficient link resources.

[0069] In this embodiment, the learning mechanism of the reinforcement learning intelligent decision-making algorithm adopts a policy network, a value network, and an experience replay mechanism. The policy network is used to map the state space to protection actions, the value network is used to evaluate the long-term value of protection switching actions, and the experience replay mechanism is used to learn from protection switching records to improve the success rate of protection switching.

[0070] See Figure 3 As shown, Figure 3 The flowchart for the reinforcement learning intelligent decision-making algorithm is as follows: B1. Collect the optical layer wavelength resources, electrical layer time slot resources, and link status of the entire network, initialize the state space, and use it as input for the reinforcement learning intelligent decision-making algorithm.

[0071] B2. The reinforcement learning intelligent decision-making algorithm extracts features from the data in the state space (optical layer wavelength resources, electrical layer time slot resources, and link state) to obtain feature values.

[0072] B3. Input the extracted feature values ​​into the policy network of the reinforcement learning intelligent decision-making algorithm, and the policy network will map the feature values ​​to protection actions.

[0073] B4. Output protection switching strategy decision and map it to the action space.

[0074] B5. Based on the protection switching strategy decision, determine the specific protection switching strategy in the action space.

[0075] B6. Perform protection switching actions in the actual photoelectric collaborative network.

[0076] B7. After performing the above protection switching action, observe the new state of the optoelectronic cooperative network and proceed to B8; at the same time, feed back the observed new state to step B2 so that it can re-enter the strategy network for the next round of iteration.

[0077] B8. Based on the execution result of the protection switching action, calculate the reward function and feed back the reward value.

[0078] B9. Store the reward value and the corresponding action status into the experience replay pool for experience replay.

[0079] B10. Based on the results of the experience replay, update the policy gradient and feed the updated parameters back to the policy network in step B3, and input them into the value network in step B11.

[0080] B11. The long-term value of protection switching actions is assessed by a value network.

[0081] B12. Perform Q (Quality) evaluation to assess the operational quality and efficiency from the state space to the policy network and finally the action space.

[0082] Furthermore, in one embodiment, the above-mentioned combination of cross-layer service configuration information and experience replay mechanism to output protection switching strategy, the specific steps of which are as follows: Based on the configuration information, if a pre-configured optical layer path exists, the optical layer rerouting strategy is output; otherwise, the resource availability of the optical layer and electrical layer is calculated separately, and the historical success rates of the optical layer rerouting strategy and the electrical layer protection switching strategy are obtained through the experience replay mechanism.

[0083] In this embodiment, by reading the service configuration information of cross-layer services (including protection type configuration and prefabricated path information), if the service configuration information contains an allocated optical layer protection channel identifier, it is determined that an optical layer prefabricated path exists.

[0084] If the resource availability and historical success rate of the optical layer exceed their respective thresholds, and the resource availability and historical success rate of the electrical layer also exceed their respective thresholds, then the strategy with the higher historical success rate will be adopted.

[0085] If only the optical layer meets the threshold condition, then the strategy of the optical layer is adopted; if only the electrical layer meets the threshold condition, then the strategy of the electrical layer is adopted.

[0086] If neither the optical layer nor the electrical layer meets the threshold conditions, an optoelectronic collaborative protection switching strategy is output.

[0087] In this embodiment, based on the configuration information, if a pre-prepared optical layer path exists, an optical layer rerouting strategy is output. If no pre-prepared optical layer path exists, the availability of optical layer resources and electrical layer resources are calculated according to the state space. Historical success rates of the optical layer rerouting strategy and electrical layer protection switching strategy are obtained from the experience replay mechanism. When a pre-prepared optical layer path exists, rapid hardware switching is prioritized to ensure timely service recovery. When no pre-prepared path exists, the most reliable protection strategy is dynamically selected based on a joint evaluation of real-time resource availability and historical success experience: when both the optical and electrical layers meet the threshold, the strategy with the higher success rate is selected; if only one layer meets the threshold, the strategy for that layer is directly adopted; if neither layer meets the threshold, optoelectronic collaborative protection switching is triggered. This hierarchical decision-making and intelligent selection mechanism effectively avoids the risk of failure caused by blind switching under resource constraints or low historical switching success rates, while balancing recovery speed and success rate, significantly improving the reliability of protection switching and resource utilization efficiency for cross-layer services in complex fault scenarios.

[0088] Furthermore, in one embodiment, a specific implementation of a cross-layer service protection switching process is given, see [link to relevant documentation]. Figure 4 As shown, Figure 4 A flowchart outlining the detailed steps for cross-layer service protection failover is provided below: C1. Initialize the entire network topology, optical layer resources, and electrical layer resources, and create cross-layer services.

[0089] C2. Collect fault information and locate the fault.

[0090] C3. Determine the fault type based on fault location. If the fault occurs at the user-side port or electrical layer processing unit, it is determined to be an electrical layer fault, and proceed to step C4. If the fault occurs at the line-side optical fiber or optical layer wavelength channel, it is determined to be an optical layer fault, and proceed to step C5.

[0091] C4. Execute the electric layer protection switching strategy and proceed to step C11.

[0092] C5. Determine if a light layer prefabrication path exists. If yes, proceed to step C6; otherwise, proceed to step C7.

[0093] C6. Execute the optical layer rerouting policy and proceed to step C11.

[0094] C7. Determine the strategy to be executed based on the resource availability and historical success rate of the optical and electrical layers.

[0095] If the resource availability and historical success rate of the optical layer exceed their respective thresholds, and the resource availability and historical success rate of the electrical layer also exceed their respective thresholds, proceed to step C8. For example, if the resource availability of the optical layer exceeds a preset first threshold and the historical success rate of the optical layer rerouting strategy is greater than a preset second threshold, and the resource availability of the electrical layer exceeds a preset third threshold and the historical success rate of the electrical layer protection switching strategy is greater than a preset fourth threshold, proceed to step C8.

[0096] The four thresholds mentioned above are independent of each other and can be dynamically adjusted. The first threshold is set according to the network-wide redundancy strategy of optical layer wavelength resources and service SLA requirements. In this embodiment, it can be set to 70%. The second threshold is set according to the statistical quantile (such as median or average success rate) of the historical success rate of the optical layer rerouting strategy in the experience replay mechanism. In this embodiment, it can be set to 85%. The third threshold is set according to the fragmentation degree of electrical layer time slot resources and service granularity matching requirements. In this embodiment, it can be set to 75%. The fourth threshold is set according to the statistical quantile of the historical success rate of the electrical layer protection switching strategy in the experience replay mechanism. In this embodiment, it can be set to 80%.

[0097] If only the optical layer meets the threshold condition, proceed to step C9. For example, if only the resource availability of the optical layer exceeds the first threshold and the historical success rate of the optical layer rerouting strategy is greater than the second threshold, proceed to step C9.

[0098] If only the electrical layer meets the threshold condition, proceed to step C4. For example, if only the resource availability of the electrical layer exceeds the third threshold mentioned above and the historical success rate of the electrical layer protection switching strategy is greater than the fourth threshold mentioned above, proceed to step C4.

[0099] If neither the optical layer nor the electrical layer meets the threshold conditions, proceed to step C10. For example, if the optical layer does not meet the condition that its resource availability exceeds the first threshold and the historical success rate of its optical layer rerouting strategy is greater than the second threshold, and the electrical layer does not meet the condition that its resource availability exceeds the third threshold and the historical success rate of its optical layer rerouting strategy is greater than the fourth threshold, proceed to step C10.

[0100] C8. Execute the strategy with the higher historical success rate and proceed to step C11. For example, if the historical success rate of the optical layer rerouting strategy is higher than that of the electrical layer protection switching strategy, execute the optical layer rerouting strategy; if the historical success rate of the electrical layer protection switching strategy is higher than that of the optical layer rerouting strategy, execute the electrical layer protection switching strategy, and then proceed to step C11.

[0101] C9. Execute the optical layer rerouting policy and proceed to step C11.

[0102] C10. Execute the photoelectric collaborative protection switching strategy and proceed to step C11.

[0103] C11. Synchronize the status of resources across the entire network.

[0104] C12. Record the cross-layer service protection switching process, including whether the switching was successful, the cross-layer service paths before and after the switching, and the reason for the switching. Store the records in the database related to the optical transport network management plane. Limit the number of records and the storage time to avoid excessive data due to network instability, which would lead to low query and addition efficiency.

[0105] Furthermore, in one embodiment, in step S2 above, the corresponding protection switching action is performed, specifically as follows: For the optical layer rerouting strategy, if an optical layer pre-fabricated path exists, cross-layer services will be switched from the faulty path to the optical layer pre-fabricated path; if no optical layer pre-fabricated path exists, the optical transport network control plane will be triggered to re-pathfind and perform dynamic rerouting.

[0106] For the electric layer protection switching strategy, the electric layer protection switching action is executed.

[0107] For the optoelectronic coordinated protection switching strategy, the protection switching actions corresponding to the optical layer rerouting strategy and the electrical layer protection switching strategy are executed respectively.

[0108] In this embodiment, corresponding protection switching actions are executed for different strategies: if an optical layer rerouting strategy is adopted and a pre-built optical layer path exists, the cross-layer service is directly switched from the faulty path to the pre-built path, utilizing the rapid switching capability of the optical layer hardware to achieve millisecond-level service recovery; if no pre-built path exists, dynamic rerouting in the control plane is triggered, which takes slightly longer but still completes the recovery; if an electrical layer protection switching strategy is adopted, the switching is executed according to the electrical layer mechanism, making full use of the flexible scheduling capability of the electrical layer; if an optoelectronic collaborative protection switching strategy is adopted, the corresponding actions of the optical layer and electrical layer are executed simultaneously to achieve dual-layer collaborative recovery. Through the above differentiated execution mechanism, the most suitable switching action can be accurately matched under different fault scenarios and resource conditions, ensuring rapid recovery when a pre-built path exists, and providing a backup guarantee through dynamic rerouting or collaborative switching when a path is missing, thereby comprehensively improving the real-time performance, flexibility, and reliability of cross-layer service protection switching.

[0109] Furthermore, in one embodiment, if the aforementioned dynamic rerouting fails to successfully restore the cross-layer service, then an electrical layer protection switching action is performed.

[0110] In this embodiment, when optical layer dynamic rerouting fails to restore cross-layer services, an electrical layer protection switching action is immediately triggered, and the electrical layer takes over the service recovery process. This mechanism provides an effective fallback guarantee for optical layer recovery failures, avoiding the problem of prolonged service interruption or switching failure due to single-layer rerouting failures. Through relay protection between the optical and electrical layers, the flexible rerouting and port switching capabilities of the electrical layer are fully utilized. When optical layer recovery is unsuccessful, the system promptly switches to the electrical layer switching path, thereby significantly improving the final success rate of cross-layer service protection switching and the robustness of the optical transport network system under complex fault scenarios, ensuring that services can be restored as quickly and reliably as possible.

[0111] Furthermore, in one embodiment, the above-mentioned operation of performing electrical layer protection switching includes the following specific steps: For cross-layer services where there are pre-prepared electrical paths, pre-prepared path switching is performed first. If pre-prepared path switching fails, the optical transport network control plane is triggered to re-pathfind and perform dynamic rerouting.

[0112] For port-level failures, the transmission path for cross-layer services is switched from the working port to the protection port.

[0113] In this embodiment, for electrical layer protection switching actions, if a pre-prepared electrical layer path exists in the cross-layer service, the pre-prepared path switching is executed first, utilizing the pre-set channel to achieve rapid service recovery. If the pre-prepared path switching fails, the control plane is triggered to re-pathfind and execute dynamic rerouting, providing secondary protection through flexible path calculation. For port-level faults, the transmission path is directly switched from the faulty working port to the backup protection port, achieving rapid local switching. Through the above-mentioned hierarchical processing mechanism, the speed of the pre-prepared electrical layer path and the flexibility of dynamic rerouting are fully utilized, while providing precise port switching for port-level faults, effectively improving the success rate and response speed of electrical layer protection switching, and enhancing the recovery capability of cross-layer services under various electrical layer fault scenarios.

[0114] Secondly, embodiments of this application also provide a cross-layer service protection switching device.

[0115] In one embodiment, see Figure 5 As shown, the aforementioned cross-layer service protection switching device includes a resource pool construction module and a protection switching module, specifically: The resource pool construction module is used to map optical and electrical layer resources of cross-layer services to build an optoelectronic collaborative resource pool.

[0116] The protection switching module is used to determine the protection switching strategy based on the resource status of the optoelectronic collaborative resource pool and a pre-established reinforcement learning intelligent decision-making algorithm if the fault type of the cross-layer service is an optical layer fault, and then execute the corresponding protection switching action. The protection switching strategies include optical layer rerouting strategy, electrical layer protection switching strategy, and optoelectronic collaborative protection switching strategy.

[0117] Furthermore, a specific embodiment of a cross-layer service protection switching device architecture is given, see [link to documentation]. Figure 6 As shown, Figure 6 This is a detailed architectural diagram of a cross-layer service protection switching device, including a cross-layer service management module, a resource mapping management module, a network-wide topology management module, and an intelligent algorithm module. Specifically: The cross-layer business management module is used to manage cross-layer businesses, specifically including functions such as creating, switching, and modifying protection types for cross-layer businesses.

[0118] The resource mapping management module is used to map optical layer resources and electrical layer resources, abstract and model optoelectronic collaborative resources, and is also responsible for resource allocation and synchronization of resource occupancy status after business creation.

[0119] The network topology management model is used to manage the hierarchical topology of the entire network, such as OMS and OCh hierarchical topology, as well as to receive alarms, performance and other information of the links, and to synchronize link status and resource utilization in real time.

[0120] The intelligent algorithm module is used to perform backtracking and learning by combining the topology and resource status of the entire network with the service failover records. It records paths with high failover success rates and performs cross-layer service protection failover by combining the resource availability and fault type of the entire network.

[0121] To ensure accurate communication between modules, a message bus is introduced. Each communication uses a Session ID plus a password to guarantee security and reliability. To avoid unordered communication between modules, a message queue is used to process messages from each module.

[0122] This application achieves unified scheduling and dynamic allocation of optical layer wavelengths and electrical layer time slots by incorporating optical layer resources and electrical layer resources into a unified resource pool. It supports an on-demand reservation mechanism, effectively eliminating resource idleness caused by independent layered reservations and substantially improving the overall network resource utilization. Unified scheduling of optical layer wavelengths and electrical layer time slots enables dynamic allocation of optoelectronic resources based on actual service bandwidth requirements and real-time network load. This avoids resource idleness and waste caused by independent layered reservations in traditional solutions, allowing for the support of more cross-layer services under the same physical resource conditions and significantly reducing operating costs.

[0123] Based on a comprehensive assessment of fault type and resource status, a multi-dimensional decision-making algorithm is constructed. This algorithm can accurately determine the optimal switching action when a fault occurs, effectively avoiding network oscillations caused by ineffective switching and significantly improving the accuracy of protection switching decisions. This multi-dimensional decision-making mechanism not only considers the physical location of the fault (user side or line side) and the fault type (port fault, link fault, signal degradation, etc.), but also integrates the service level agreement requirements of the service, the current availability of resources at the optical and electrical layers, and historical successful switching experience. This avoids blind switching caused by incomplete information or fixed rules in traditional solutions. Simultaneously, through the synergy of a global resource view and intelligent algorithms, it can dynamically identify and bypass shared risk link groups. Even in extreme scenarios such as multi-point faults and complex risks, it can still select the path with the highest recovery success rate, thereby greatly enhancing the survivability of cross-layer services and the compliance capability with service level agreements.

[0124] By combining global resource views and intelligent algorithms for optimization, shared risky links can be dynamically avoided and the optimal recovery path can be selected when a fault occurs, significantly improving the success rate of service recovery under adverse conditions. When multiple faults occur simultaneously in the network or there is a risk of shared risky link groups, traditional solutions often fail to switchover or cause secondary service interruptions due to their inability to coordinate the overall situation. However, this application, based on a real-time updated optoelectronic collaborative resource pool and historical experience playback, can identify potential risky paths in advance and actively exclude these high-risk paths during decision-making, prioritizing recovery channels with proven success rates. This significantly improves the success rate of service recovery in complex scenarios such as multi-point faults and shared risky link groups.

[0125] It should be noted that the sequence numbers of the embodiments in this application are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0126] The terms "comprising" and "having," and any variations thereof, in the specification, claims, and accompanying drawings of this application are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to such process, method, product, or apparatus. The terms "first," "second," and "third," etc., are used to distinguish different objects, etc., and do not indicate a sequence, nor do they limit "first," "second," and "third" to different types.

[0127] In the description of the embodiments of this application, terms such as "exemplary," "for example," or "for instance" are used to indicate examples, illustrations, or explanations. Any embodiment or design described as "exemplary," "for example," or "for instance" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of terms such as "exemplary," "for example," or "for instance" is intended to present the relevant concepts in a concrete manner.

[0128] In the description of the embodiments of this application, unless otherwise stated, " / " means "or". For example, A / B can mean A or B. The "and / or" in the text is merely a description of the relationship between related objects, indicating that there can be three relationships. For example, A and / or B can mean: A exists alone, A and B exist simultaneously, and B exists alone. In addition, in the description of the embodiments of this application, "multiple" means two or more.

[0129] In some processes described in the embodiments of this application, multiple operations or steps are included in a specific order. However, it should be understood that these operations or steps may not be executed in the order they appear in the embodiments of this application, or they may be executed in parallel. The sequence number of the operation is only used to distinguish different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed sequentially or in parallel, and these operations or steps may be combined.

[0130] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) as described above, and includes several instructions to cause a terminal device to execute the methods described in the various embodiments of this application.

[0131] The above are merely preferred embodiments of this application and do not limit the patent scope of this application. Any equivalent structural or procedural transformations made using the content of this application's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of this application.

Claims

1. A method for cross-layer service protection switching, characterized in that, The method includes: Map the optical and electrical layer resources of cross-layer services to construct an optoelectronic collaborative resource pool; If the fault type of the cross-layer service is an optical layer fault, a protection switching strategy is determined by a pre-established reinforcement learning intelligent decision-making algorithm based on the resource status of the optoelectronic collaborative resource pool, and the corresponding protection switching action is executed; the protection switching strategy includes an optical layer rerouting strategy, an electrical layer protection switching strategy, and an optoelectronic collaborative protection switching strategy.

2. The cross-layer service protection switching method as described in claim 1, characterized in that, The process of mapping optical and electrical layer resources for cross-layer services to construct an optoelectronic collaborative resource pool includes: The optical layer wavelength resources are mapped to the optical multiplexing segment layer topology, and the electrical layer time slot resources are mapped to the optical channel layer layer topology. The mapped optical layer resources are abstracted into multiple optical layer resource objects, and the mapped electrical layer resources are abstracted into multiple electrical layer resource objects. A resource association model for optical layer resource objects and electric layer resource objects is established, and the optical-electric collaborative resource pool is composed of the optical layer resource objects, electric layer resource objects and the resource association model of the entire network.

3. The cross-layer service protection switching method as described in claim 1, characterized in that, If the fault type of the cross-layer service is an electrical layer fault, an electrical layer protection switching action is performed.

4. The cross-layer service protection switching method as described in claim 1, characterized in that, Based on the resource status of the optoelectronic collaborative resource pool, a protection switching strategy is determined through a pre-established reinforcement learning intelligent decision-making algorithm, including: The state space is initialized based on the optical layer wavelength resource status, electrical layer time slot resource status, and link status of the entire network topology of the optoelectronic collaborative resource pool. Features are extracted from the state space according to the reinforcement learning intelligent decision-making algorithm, and protection switching strategy is output by combining cross-layer service configuration information and experience replay mechanism.

5. The cross-layer service protection switching method as described in claim 4, characterized in that, The protection failover strategy, which combines cross-layer service configuration information and experience replay mechanism, includes: Based on the configuration information, if a pre-configured optical layer path exists, an optical layer rerouting strategy is output; otherwise, the resource availability of the optical layer and electrical layer is calculated respectively, and the historical success rates of the optical layer rerouting strategy and the electrical layer protection switching strategy are obtained through the experience replay mechanism. If the resource availability and historical success rate of the optical layer exceed their respective thresholds, and the resource availability and historical success rate of the electrical layer also exceed their respective thresholds, then the strategy with the higher historical success rate will be adopted. If only the optical layer meets the threshold condition, then the strategy of the optical layer is adopted; if only the electrical layer meets the threshold condition, then the strategy of the electrical layer is adopted. If neither the optical layer nor the electrical layer meets the threshold conditions, an optoelectronic collaborative protection switching strategy is output.

6. The cross-layer service protection switching method as described in claim 1, characterized in that, The execution of the corresponding protection switching action includes: For the optical layer rerouting strategy, if an optical layer pre-fabricated path exists, cross-layer services will be switched from the faulty path to the optical layer pre-fabricated path; if no optical layer pre-fabricated path exists, the optical transport network control plane will be triggered to re-find the route and perform dynamic rerouting. For the electric layer protection switching strategy, execute the electric layer protection switching action; For the optoelectronic coordinated protection switching strategy, the protection switching actions corresponding to the optical layer rerouting strategy and the electrical layer protection switching strategy are executed respectively.

7. The cross-layer service protection switching method as described in claim 6, characterized in that, If the dynamic rerouting fails to restore the cross-layer service, then the electrical layer protection switching action is performed.

8. The cross-layer service protection switching method as described in claim 3 or 6, characterized in that, The electric layer protection switching action includes: In the case of electrical layer pre-prepared paths in the cross-layer services, pre-prepared path switching is performed first. If pre-prepared path switching fails, the optical transport network control plane is triggered to re-find the route and perform dynamic rerouting. For port-level failures, the transmission path of the cross-layer service is switched from the working port to the protection port.

9. The cross-layer service protection switching method as described in claim 4, characterized in that, The reinforcement learning intelligent decision-making algorithm also includes a reward function, which includes a positive reward for successful cross-layer business fault recovery and a negative reward for unsuccessful cross-layer business fault recovery or insufficient link resources.

10. A cross-layer service protection switching device, characterized in that, The device includes: The resource pool construction module is used to map optical layer resources and electrical layer resources of cross-layer services to build an optoelectronic collaborative resource pool. The protection switching module is used to determine the protection switching strategy based on the resource status of the optoelectronic collaborative resource pool and a pre-established reinforcement learning intelligent decision-making algorithm if the fault type of the cross-layer service is an optical layer fault, and to execute the corresponding protection switching action. The protection switching strategy includes an optical layer rerouting strategy, an electrical layer protection switching strategy, and an optoelectronic collaborative protection switching strategy.

Images