An end-to-end path computation method and system based on a domain-aggregate quantum key management layer architecture

By adopting a domain-aggregated quantum key management layer architecture in the quantum key distribution network, distributed path computation and key management are realized, which solves the problem of high single-point failure risk of the single KM centralized scheme and has the ability of redundancy backup and fault replacement.

CN123122828APending Publication Date: 2026-11-17ELECTRIC POWER RES INST CHINA SOUTHERN POWER GRID CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611502375.7
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-09-24
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

The existing single-KM centralized solution concentrates all network key management and path calculation functions in a single central node, lacking redundancy backup and fault tolerance mechanisms, resulting in a high risk of single point of failure.

Method used

A domain-aggregate quantum key management layer architecture is adopted, which divides the QKD network into multiple management domains. Each management domain is configured with a regional key management device. Digest status information is exchanged through domain-level link state announcement messages to maintain an inter-domain abstract graph and realize distributed path computation and key management.

Benefits of technology

It reduces the impact of single point of failure on the entire network, has redundant backup and fault-tolerant capabilities, and effectively avoids the paralysis of the entire network path calculation and key resource management caused by the failure of the central node.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122828A_ABST
    Figure CN123122828A_ABST
Patent Text Reader

Abstract

This invention discloses an end-to-end path computation method and system based on a domain-aggregate quantum key management layer architecture. It addresses the technical problem of existing single-KM centralized schemes lacking redundancy backup and fault-passing mechanisms, leading to a high risk of single-point failures. The method involves, under the domain-aggregate quantum key management layer architecture, constructing and maintaining an inter-domain abstract graph through limited flooding of D-LSA acquisition by key management devices in each region. The cross-domain end-to-end path computation is decomposed into two levels: inter-domain minimum cost domain sequence search and calculation of precise path segments within each management domain. First, the shortest domain sequence from the source domain to the destination domain is determined. Then, each regional key management device calculates the path segments from the entry point to the exit boundary node within its domain. Finally, the boundary nodes are concatenated to form the end-to-end path, transforming path computation from a fine-grained search across the entire network into a combination of coarse-grained inter-domain search and fine-grained local search within each domain.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of quantum key distribution network technology, and in particular to an end-to-end path calculation method and system based on a domain-aggregate quantum key management layer architecture. Background Technology

[0002] Quantum Key Distribution (QKD) generates a consistent symmetric key between communicating parties through quantum state preparation, transmission, reception, and post-processing. Fiber optic QKD links are typically deployed between adjacent QKD nodes and continuously generate link keys, which are stored in the key pool of the corresponding link or node. For end-to-end key service requests where the source and destination nodes are not adjacent, trusted relay nodes and multi-hop paths are usually required to establish the end-to-end key. Therefore, the implementation of end-to-end key services inevitably depends on a quantum key relay path from the source node through several trusted relay nodes to the destination node.

[0003] In QKD network engineering implementation, the key management layer, as the control core of end-to-end key services, typically maintains the key pool state, link state, task requirements, and key delivery interface, and is responsible for calculating and allocating quantum key relay paths. User end-to-end key service tasks usually include parameters such as source node, destination node, required key quantity, maximum acceptable latency, and service priority. Based on pre-configured paths, shortest paths, available key pool capacity, or dynamic routing strategies, the key management layer determines the QKD links and trusted relay nodes that the task should traverse, thereby calculating and allocating a quantum key relay path that meets the service requirements for the end-to-end key service.

[0004] The single-KM centralized scheme mainly relies on a single central KM across the entire network for end-to-end path calculation and key resource management. Its basic principle is that the centralized control and management entity uniformly grasps the QKD topology, link key pool, task queue and link load status of the entire network. After the user submits the end-to-end key service task, the central KM uniformly calculates the path from the source node to the destination node, uniformly issues key resource reservation instructions, and uniformly handles path failure, recalculation and resource release. In actual operation, the central KM is not only the sole aggregation point for network-wide status information but also the sole decision-making and execution point for path calculation and resource scheduling. Path calculation, key reservation, failure recalculation, and resource release for end-to-end key services across the network all rely on this single central node. The path solving and resource allocation for a single task depend on the operational status of the central KM. When processing a large number of tasks, the computational load and status synchronization load of the central KM increase in tandem with the task scale, and there are no other management entities across the network that can replace the functions of the central KM. If the central KM fails or crashes, path calculation, task scheduling, and key resource reservation across the entire network will simultaneously come to a standstill, and cannot be recovered by other nodes. Therefore, the single-KM centralized solution concentrates all network-wide key management and path calculation functions on a single central node, lacking redundancy backup and fault tolerance mechanisms, resulting in a high risk of single-point failure. Summary of the Invention

[0005] This invention provides an end-to-end path computation method and system based on a domain-aggregated quantum key management layer architecture. It solves the technical problem that the existing single-KM centralized scheme concentrates all key management and path computation functions of the entire network in a single central node, lacks redundancy backup and fault succession mechanism, and leads to a high risk of single point of failure.

[0006] The first aspect of this invention provides an end-to-end path computation method based on a domain-aggregate quantum key management layer architecture, applied to such an architecture. The domain-aggregate quantum key management layer architecture includes multiple management domains, each management domain comprising multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain corresponds to a regional key management device, which is communicatively connected to each QKD device within its respective management domain. Each regional key management device contains... A local key resource database is set up, which is connected to the regional key management device; each regional key management device is associated with a node management domain mapping table; adjacent management domains are connected through boundary nodes and cross-domain links, where the boundary nodes are QKD nodes belonging to two adjacent management domains and connected through cross-domain links; each regional key management device is connected to another region through restricted flooding via domain-level link status announcement messages; each regional key management device maintains an inter-domain abstract graph, where each management domain is a node and the cross-domain connection relationship between management domains is an edge; the method includes:

[0007] Acquire the QKD network topology data and user end-to-end key service task data, and perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, outputting the node management domain mapping table and the local key resource database corresponding to each regional key management device;

[0008] D-LSA digest generation and inter-domain restricted flooding processing are performed on the local key resource database corresponding to each of the aforementioned regional key management devices to output an inter-domain abstract graph.

[0009] The node management domain mapping table is used to determine the management domain to which the source node and destination node belong and to split the path calculation mode of the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers;

[0010] When the path calculation mode is identified as the intra-domain path calculation mode, based on the local key resource database corresponding to the source domain and the user end-to-end key service task data, the intra-domain optimal path is calculated for the path from the source node to the destination node, and the intra-domain optimal end-to-end path or intra-domain path unavailable determination result is output.

[0011] When the path calculation mode is identified as cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each regional key management device, and the user end-to-end key service task data, cross-domain end-to-end path calculation is performed on the path from the source domain to the destination domain, and the result of determining whether the cross-domain optimal end-to-end path or the cross-domain path is unavailable is output.

[0012] Optionally, the step of performing management domain division and regional key management device aggregation management configuration on the QKD network topology data, and outputting a node management domain mapping table and a local key resource database corresponding to each regional key management device, includes:

[0013] The QKD network topology data is divided into management domains to obtain multiple management domains;

[0014] Configure a regional key management device for each management domain, and establish a one-to-one correspondence between each management domain and its corresponding regional key management device to obtain the management scope configuration of the regional key management device for each management domain.

[0015] Based on the management scope configuration of each regional key management device, the key pool status, coding rate, link load, device status, task queue and domain topology of each QKD device in the management domain corresponding to each regional key management device are collected and aggregated for management, so as to obtain the domain key resource database corresponding to each regional key management device.

[0016] For each QKD node in each management domain, a correspondence is established between each QKD node and its respective management domain, resulting in a node management domain mapping table.

[0017] Optionally, the step of generating D-LSA digests and performing inter-domain restricted flooding processing on the local key resource database corresponding to each of the regional key management devices, and outputting an inter-domain abstract graph, includes:

[0018] Extract boundary node information and digest resource status from the local key resource database corresponding to each regional key management device, and encapsulate the boundary node information and digest resource status into a domain-level link status announcement message to obtain the D-LSA digest status message corresponding to each regional key management device.

[0019] The D-LSA digest status messages corresponding to the key management devices in each region are exchanged through a restricted flooding mechanism. The domain boundary connection relationships and digest resource status of each management domain obtained by the exchange are summarized and organized to obtain a set of inter-domain connection relationships.

[0020] Using each management domain as a node, the inter-domain connection relationships in the set of inter-domain connection relationships as edges, and carrying the summary resource status corresponding to each management domain, a domain-level topology structure is constructed to obtain an inter-domain abstract graph.

[0021] Optionally, the step of using the node management domain mapping table to determine the management domain to which the source and destination nodes belong and to split the path calculation mode for the user end-to-end key service task data, and outputting the source domain, destination domain, and path calculation mode identifiers, includes:

[0022] Extract the source node and destination node from the end-to-end key service task data;

[0023] Query the node management domain mapping table, and match the management domains of the extracted source and destination nodes to obtain the source domain and destination domain;

[0024] The source domain and the destination domain are compared for consistency. If they are in the same management domain, it is determined to be an intra-domain path calculation mode. If they are in different management domains, it is determined to be a cross-domain path calculation mode, and a path calculation mode identifier is obtained.

[0025] Optionally, when the path calculation mode is identified as an intra-domain path calculation mode, based on the local key resource database corresponding to the source domain and the user end-to-end key service task data, the intra-domain optimal path is calculated for the path from the source node to the destination node, and the intra-domain optimal end-to-end path or intra-domain path unavailable determination result is output, including:

[0026] Based on the key pool status and reserved key quantity of each link in the key resource database of the source domain, the remaining key quantity of each link corresponding to the source domain is calculated to obtain the remaining key quantity of multiple links corresponding to the source domain.

[0027] The remaining key quantity of each link corresponding to the source domain is compared and filtered with the key quantity required for the task in the end-to-end key service task data of the user, and links with a remaining key quantity less than the task requirement are excluded to obtain the set of available links corresponding to the source domain.

[0028] Weights are assigned to each link in the set of available links corresponding to the source domain to obtain a weighted topology map of available links in the source domain.

[0029] In the weighted source domain available link topology map, taking the source node in the user end-to-end key service task data as the starting point and the destination node as the ending point, a minimum weight path search is performed on the topology corresponding to the source domain to obtain the minimum weight path search result.

[0030] The existence of the minimum weight path search result is determined. If a feasible path exists, the optimal end-to-end path in the domain is obtained. If no feasible path exists, the path in the domain is deemed unusable.

[0031] Optionally, when the path calculation mode is identified as a cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each of the regional key management devices, and the user end-to-end key service task data, cross-domain end-to-end path calculation is performed on the path from the source domain to the destination domain, and the result of determining whether the cross-domain optimal end-to-end path or the cross-domain path is unavailable is output, including:

[0032] In the inter-domain abstraction graph, with the source domain as the starting point and the destination domain as the ending point, a shortest domain sequence search is performed on the inter-domain abstraction graph to obtain the shortest domain sequence from the source domain to the destination domain.

[0033] For each pair of adjacent management domains in the shortest domain sequence, the remaining key quantity of each pair of candidate cross-domain links between adjacent management domains is compared from the cross-domain link information stored in the local key resource database corresponding to each regional key management device. The cross-domain link with the largest remaining key quantity is selected, and the nodes at both ends of the link are determined as the boundary nodes of the adjacent management domains, thus obtaining the boundary node pair between adjacent management domains.

[0034] Based on the shortest domain sequence and the boundary node pair, the entry boundary node and exit boundary node of each transit management domain are determined. The intra-domain path calculation method is used to calculate the intra-domain path segment for each transit management domain to obtain the intra-domain path segment corresponding to each transit management domain.

[0035] According to the order of the shortest domain sequence, the intra-domain path segments corresponding to each management domain are connected sequentially through the corresponding boundary nodes to obtain the preliminary end-to-end path;

[0036] The availability of key resources and the connectivity of each link on the initial end-to-end path are verified. If the verification passes, the optimal cross-domain end-to-end path is obtained; if the verification fails, the cross-domain path is deemed unavailable.

[0037] The second aspect of this invention provides an end-to-end path computing system based on a domain-aggregate quantum key management layer architecture, applied to such an architecture. The domain-aggregate quantum key management layer architecture includes multiple management domains, each management domain comprising multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain corresponds to a regional key management device, which is communicatively connected to each QKD device within its respective management domain. Each regional key management device contains... A local key resource database is set up, which is connected to the regional key management device; each regional key management device is associated with a node management domain mapping table; adjacent management domains are connected through boundary nodes and cross-domain links, where the boundary nodes are QKD nodes belonging to two adjacent management domains and connected by cross-domain links; each regional key management device is connected to another region through restricted flooding via domain-level link status announcement messages; each regional key management device maintains an inter-domain abstract graph, where each management domain is a node and the cross-domain connection relationship between management domains is an edge; the system includes:

[0038] The acquisition module is used to acquire the QKD network topology data and user end-to-end key service task data, and to perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, and output the node management domain mapping table and the local key resource database corresponding to each regional key management device.

[0039] The output module is used to generate D-LSA digests and perform inter-domain restricted flooding processing on the local key resource database corresponding to each of the regional key management devices, and output an inter-domain abstract graph.

[0040] The splitting module is used to use the node management domain mapping table to determine the management domain to which the source node and destination node belong and to split the path calculation mode of the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers.

[0041] The intra-domain path determination module is used to calculate the optimal intra-domain path for the path from the source node to the destination node based on the local key resource database corresponding to the source domain and the user end-to-end key service task data when the path calculation mode is identified as intra-domain path calculation mode, and output the determination result of the optimal intra-domain end-to-end path or the intra-domain path being unavailable.

[0042] The cross-domain path determination module is used to perform cross-domain end-to-end path calculation on the path from the source domain to the destination domain when the path calculation mode is identified as cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each of the regional key management devices, and the user end-to-end key service task data, and output the determination result of the optimal cross-domain end-to-end path or the unavailable cross-domain path.

[0043] A third aspect of the present invention provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described above.

[0044] The fourth aspect of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed, it implements the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described above.

[0045] The fifth aspect of the present invention provides a computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions, wherein, when the program instructions are executed by a computer, the computer performs the steps of the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described above.

[0046] As can be seen from the above technical solutions, the present invention has the following advantages:

[0047] The present invention provides an end-to-end path computation method based on a domain-aggregated quantum key management layer architecture. This invention acquires the entire QKD network topology data and user end-to-end key service task data, divides the QKD network topology data into management domains, and configures the aggregation management of regional key management devices. The entire network of QKD devices, QKD nodes, and QKD links is divided into multiple management domains, and an independent regional key management device is configured for each management domain. A node management domain mapping table and a key resource database corresponding to each regional key management device are output, providing an architectural foundation for distributed key management and path computation. Subsequently, D-LSA digest generation and inter-domain restricted flooding processing are performed on the key resource databases corresponding to each regional key management device. This enables each regional key management device to exchange domain boundaries and digest resource states through domain-level link status announcement messages and independently maintain an inter-domain abstract graph, achieving distributed synchronization of inter-domain topology information. Finally, the node management domain mapping table is used to analyze the source node and... The system determines the management domain to which the destination node belongs and decentralizes the path calculation mode, outputting source domain, destination domain, and path calculation mode identifiers to provide domain-level routing basis for subsequent path calculations. When the path calculation mode identifier is an intra-domain path calculation mode, the regional key management device corresponding to the source domain independently completes the optimal intra-domain path calculation based on its own domain key resource database. When the path calculation mode identifier is a cross-domain path calculation mode, the regional key management devices corresponding to each management domain along the route collaboratively complete the intra-domain path segment calculation and end-to-end path splicing based on the inter-domain abstract graph and their own domain key resource database. This allows the end-to-end key path calculation function of the entire network to be distributed by multiple regional key management devices. A single regional key management device is only responsible for state maintenance and path calculation within its corresponding management domain. Its failure only affects the corresponding management domain and does not cause the entire network's path calculation and key resource management to be paralyzed. Thus, under the domain-aggregated quantum key management layer architecture, the distributed deployment of key management and path calculation functions is realized, with redundancy backup and fault succession capabilities, effectively reducing the risk of single point of failure. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0049] Figure 1 A schematic diagram of the domain-aggregate quantum key management layer architecture provided in an embodiment of the present invention;

[0050] Figure 2A flowchart illustrating the steps of an end-to-end path computation method based on a domain-aggregate quantum key management layer architecture, provided in an embodiment of the present invention;

[0051] Figure 3 A flowchart of an end-to-end path processing method provided in an embodiment of the present invention;

[0052] Figure 4 This is a structural block diagram of an end-to-end path computing system based on a domain-aggregate quantum key management layer architecture, provided for an embodiment of the present invention. Detailed Implementation

[0053] This invention provides an end-to-end path computation method and system based on a domain-aggregated quantum key management layer architecture. It solves the technical problem that the existing single-KM centralized scheme concentrates all key management and path computation functions of the entire network in a single central node, lacks redundancy backup and fault succession mechanism, and leads to a high risk of single point of failure.

[0054] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. It should be noted that in the optional embodiments of the present invention, the object information and other related data involved require the permission or consent of the object when the embodiments of the present invention are applied to specific products or technologies, and the collection, use, and processing of related data must comply with relevant laws, regulations, and standards. That is to say, if the embodiments of the present invention involve data related to the object, it needs to be obtained with the authorization and consent of the object, the authorization and consent of relevant departments, and in compliance with relevant laws, regulations, and standards. If personal information is involved in the embodiments, the acquisition of all personal information requires the consent of the individual. If sensitive information is involved, the separate consent of the information subject is required, and the embodiments also need to be implemented with the authorization and consent of the object.

[0055] Terminology Explanation:

[0056] 1. QKD (Quantum Key Distribution) device: A device used to execute quantum key distribution protocols, generate link keys, or participate in link key management.

[0057] 2. Region KM (Key Manager): A key management entity that manages multiple QKD devices, QKD links, and key pools within a management domain.

[0058] 3. Management Domain: A regional management area consisting of multiple QKD devices, QKD nodes, and QKD links. QKD devices refer to those deployed at QKD nodes, used to execute quantum key distribution protocols and generate link keys. They include quantum state transmitting units, quantum state receiving units, classical communication units, and post-processing units. Through the preparation, transmission, and measurement of quantum states on optical fiber channels, followed by post-processing processes such as basis vector comparison, error correction, and privacy amplification, a consistent symmetric key is generated at both ends of the link. QKD devices are the main generators of key resources in the QKD network. Their operational status, key generation rate, and key pool status are collected and managed by the regional key management equipment corresponding to their management domain.

[0059] A QKD node refers to a network node in a QKD network that deploys QKD devices and serves as a quantum key distribution endpoint or trusted relay point. It has a unique node identifier and can serve as a source node or destination node for end-to-end key service tasks, or as an intermediate relay node in a multi-hop quantum key relay path. It is a basic building block of the QKD network topology, and its correspondence with its management domain is registered in the node management domain mapping table.

[0060] A QKD link refers to a fiber optic quantum channel connecting two adjacent QKD nodes. The two ends are equipped with quantum state transmitting units and quantum state receiving units of QKD devices, respectively. By transmitting optical quantum signals in the optical fiber, the quantum key distribution protocol is continuously executed and a link key is generated. The generated link key is stored in the key pool corresponding to the QKD link. It is the transmission and generation channel of key resources in the QKD network. Its key pool status, reserved key quantity, and link availability status are stored in the key resource database of the corresponding management domain.

[0061] 4. D-LSA (Domain Link State Advertisement): A domain-level link state advertisement generated by the regional KM and flooded between KMs, carrying the domain boundary and summary resource status.

[0062] 5. Inter-domain abstract graph: A path computation graph constructed with regions KM or management domains as nodes and inter-domain connection relationships as edges.

[0063] 6. Boundary Nodes / Boundary Links: QKD nodes or QKD links used to connect adjacent management domains.

[0064] 7. End-to-end path: A QKD relay path from the task source node to the destination node, composed of one or more intra-domain path segments.

[0065] 8. KP (Key Pool): refers to the storage unit used to store the shared key resources generated by the QKD link.

[0066] This invention provides an end-to-end path computation method based on a domain-aggregate quantum key management layer architecture. The architecture comprises multiple management domains, each containing multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain has a corresponding regional key management device, which communicates with each QKD device within its domain. Each regional key management device has its own key resource database, which is connected to the regional key management device. Each regional key management device is associated with a node management domain mapping table. Adjacent management domains are connected via boundary nodes and cross-domain links. Boundary nodes are QKD nodes belonging to two adjacent management domains and connected via cross-domain links. Regional key management devices are connected via restricted flooding through domain-level link status announcement messages. Each regional key management device maintains an inter-domain abstract graph, with each management domain as a node and the cross-domain connections between management domains as edges.

[0067] like Figure 1 As shown, this application proposes a domain-based aggregated quantum key management layer architecture. The entire QKD network consists of multiple QKD nodes, QKD devices, and QKD links. Unlike the centralized architecture where only one central KM is set up for the entire network, and unlike the device-level distributed architecture where each QKD device has its own independent KM, this application divides multiple QKD devices into a management domain and configures a regional KM for each management domain.

[0068] Each regional KM is responsible for aggregating and managing multiple QKD devices, QKD links, and link key pools (KPs) within its domain. The regional KM maintains node information, link information, key pool status, reserved key quantity, link availability status, boundary node information, and cross-domain link information within its domain.

[0069] The system maintains the mapping relationship between nodes and management domains. Management domains can be connected through boundary nodes or cross-domain links. Regions (KMs) exchange digest state information through Domain-Level Link State Advertisement (D-LSA) to establish inter-domain abstract graphs and support cross-domain end-to-end key path calculation.

[0070] In practical implementation, the entire QKD network is first divided into management domains. The entire QKD network consists of multiple QKD nodes, QKD devices deployed at the QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link continuously generates link keys during the coding process and stores the generated link keys in the key pool corresponding to that link. Unlike the centralized architecture where only one central KM is set up for the entire network, and unlike the device-level distributed architecture where each QKD device has its own independent KM, this application divides multiple QKD devices, multiple QKD nodes, and multiple QKD links that are geographically adjacent or belong to the same administrative region according to the preset management domain division rules. The QKD link is divided into a management domain, thus dividing the entire QKD network into several management domains. Subsequently, a regional key management device is configured in each management domain. This regional key management device establishes a communication connection with each QKD device within that domain through a management communication interface, enabling it to collect the operational status and key resource status of each QKD device within its domain. Each regional key management device is configured with its own domain key resource database, which is connected to the data output terminal of the regional key management device. The regional key management device then displays the collected node information, link information, key pool status, reserved key quantity, and link availability status within its domain. Boundary node information and cross-domain link information are stored and maintained in the local key resource database. Simultaneously, a node management domain mapping table is established and maintained in the system, registering a one-to-one correspondence between the identifier of each QKD node and the identifier of its respective management domain. This mapping table is updated synchronously by the key management devices in each region after completing the node affiliation review. For two adjacent management domains, QKD nodes belonging to both domains and physically adjacent are selected as boundary nodes. Cross-domain links are deployed between these two boundary nodes to achieve physical connectivity between adjacent management domains. Cross-domain links also continuously generate link keys and store these keys in the corresponding key pool. Each region's key... Management devices establish restricted flooding connections for domain-level link status announcement messages through a management communication network. Each regional key management device periodically extracts boundary node information and digest resource status from its own domain's key resource database, encapsulates them into domain-level link status announcement messages, and sends them to neighboring regional key management devices. It also receives domain-level link status announcement messages from other regional key management devices. This allows each regional key management device to build and maintain an inter-domain abstract graph locally. The inter-domain abstract graph uses each management domain as a node, cross-domain connection relationships between management domains as edges, and carries the digest resource status of each management domain, providing a domain-level topology basis for subsequent cross-domain end-to-end key path calculation.

[0071] When performing end-to-end quantum key path computation using the aforementioned domain-aggregated quantum key management layer architecture, the entire QKD network is first divided into multiple independent management domains, and a regional key management device is configured for each management domain. This allows the key management and path computation functions of the entire network to be distributed among multiple regional key management devices, rather than being centralized in a single central node. Subsequently, each regional key management device independently maintains the precise state of its domain through its own domain key resource database, and exchanges domain boundary and digest resource states only with other regional key management devices through a restricted flooding mechanism of domain-level link state announcement messages, avoiding centralized synchronization of fine-grained states across the entire network. Upon receiving an end-to-end key service task, each regional key management device quickly determines the management domain to which the source and destination nodes belong through a node management domain mapping table. If the source and destination nodes are in the same management domain, the regional key management device corresponding to that management domain independently completes the optimal path computation within the domain based on its own domain key resource database. If nodes are located in different management domains, the shortest domain sequence is first determined by breadth-first search based on the inter-domain abstract graph. Then, the regional key management devices corresponding to the management domains along the shortest domain sequence calculate the path segments within their respective domains and concatenate them into an end-to-end path through the boundary nodes. Since the intra-domain path calculation is completed independently by the regional key management devices of the corresponding management domain, and the cross-domain path calculation is completed collaboratively by multiple regional key management devices based on the inter-domain abstract graph, when a single regional key management device fails, it only affects the path calculation and key resource management within its corresponding management domain. The regional key management devices of other management domains can still maintain the state of their respective domains and participate in the cross-domain path calculation. Thus, the distributed deployment of key management and path calculation functions is realized at the architecture level, with redundancy backup and fault tolerance capabilities. This effectively reduces the risk of a single point of failure in a single KM centralized scheme, where the entire network is paralyzed due to the failure of the only central node. At the same time, the design of direct management of precise intra-domain state and exchange of only digest state between domains also reduces the overhead of network-wide state synchronization.

[0072] like Figure 2 As shown, this invention proposes an end-to-end path computation method based on a domain-aggregate quantum key management layer architecture, specifically including the following steps:

[0073] Step 201: Obtain the QKD network topology data and user end-to-end key service task data, and perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, outputting the node management domain mapping table and the local key resource database corresponding to each regional key management device.

[0074] It should be noted that the QKD network topology data specifically includes the identifiers and location information of each QKD node, the identifiers and node information of each QKD device, the endpoint nodes and connection relationships of each QKD link, and the key pool status information corresponding to each QKD link. The user end-to-end key service task data specifically includes the source node identifier, the destination node identifier, and the number of keys required for the task. The regional key management device collects the QKD network topology data node by node from the entire QKD network through the management communication interface, and simultaneously receives user-submitted user end-to-end key service task data through the key service interface. Subsequently, the regional key management device divides the QKD network topology data into management domains according to preset management domain division rules, based on the geographical distribution or administrative affiliation of the QKD devices, thus dividing multiple QKD devices, QKD nodes, and QKD nodes into management domains. D-links are merged into a single management domain, and a regional key management device is configured for each management domain, establishing a one-to-one correspondence between management domains and regional key management devices. Each regional key management device collects and aggregates the key pool status, coding rate, link load, device status, task queue, and domain topology of each QKD device within its corresponding management domain. The collected node information, link information, key pool status, reserved key quantity, link availability status, boundary node information, and cross-domain link information are stored in the domain's key resource database. Simultaneously, based on the affiliation of each QKD node in the management domain division, a correspondence between each QKD node and its respective management domain is established and registered in the node management domain mapping table. This completes the configuration of management domain division and regional key management device aggregation management, providing a data foundation for subsequent D-LSA digest generation and inter-domain flooding.

[0075] Further, step 201 may include the following sub-steps:

[0076] S21. Divide the QKD network topology data into management domains to obtain multiple management domains;

[0077] S22. Configure a regional key management device for each management domain, and establish a one-to-one correspondence between each management domain and its corresponding regional key management device to obtain the management scope configuration of the regional key management device for each management domain.

[0078] S23. Based on the management scope configuration of each regional key management device, the key pool status, coding rate, link load, device status, task queue and domain topology of each QKD device in the management domain corresponding to each regional key management device are collected and aggregated for management, so as to obtain the domain key resource database corresponding to each regional key management device.

[0079] S24. Based on the affiliation relationship of each QKD node in each management domain, establish a correspondence between each QKD node and its management domain to obtain a node management domain mapping table.

[0080] It should be noted that when dividing the QKD network topology data into management domains, the regional key management device reads the geographical location and administrative affiliation information of each QKD device in the QKD network topology data, and groups them according to preset management domain division rules. For example, QKD devices, QKD nodes, and QKD links within the same city are grouped into the same management domain based on the administrative city boundary, or geographically adjacent devices are grouped into the same management domain based on the condition that the fiber optic link distance between QKD devices does not exceed a preset threshold. After traversing all QKD devices in the network and completing the grouping, multiple management domains are obtained. Subsequently, a regional key management device with a unique identifier is assigned to each of the divided management domains. The management communication interface of the regional key management device is paired and connected with the management interface of each QKD device in the management domain, establishing a one-to-one correspondence between the management domain identifier and the regional key management device identifier, clarifying the scope of QKD devices managed by each regional key management device, and obtaining the management domain for each management domain. The corresponding regional key management device is configured with a management scope. Based on the configuration of its management scope, each regional key management device sends a status collection command to each QKD device within its corresponding management domain. Upon receiving the command, each QKD device returns its key pool status, coding rate, link load, device status, task queue, and domain topology information to the corresponding regional key management device. The regional key management device aggregates and organizes the returned multi-device status data, classifying and storing it according to node information, link information, key pool status, reserved key quantity, link availability status, boundary node information, and cross-domain link information, thus obtaining the domain-specific key resource database for each regional key management device. Finally, based on the affiliation of each QKD node in each management domain, all QKD nodes in each management domain are traversed, and the QKD node identifier is paired and registered with the identifier of the management domain to which the node belongs, forming a set of correspondences between QKD nodes and management domains across the entire network, thus obtaining a node management domain mapping table.

[0081] Step 202: Perform D-LSA digest generation and inter-domain restricted flooding processing on the local key resource database corresponding to each regional key management device, and output the inter-domain abstract graph.

[0082] It should be noted that each regional key management device extracts the boundary node information and digest resource status of its respective domain from its local key resource database, and encapsulates the message according to the preset D-LSA encapsulation rules. For example, fields such as boundary node identifier, cross-domain link identifier, and link remaining key amount digest are filled into the message payload according to the preset message format to obtain the D-LSA digest status message. Each regional key management device sends the generated D-LSA digest status message to the adjacent regional key management device through a restricted flooding mechanism, and at the same time receives D-LSA digest status messages from other regional key management devices. The device summarizes and organizes the domain boundary connection relationships and digest resource status of each management domain carried in the message to obtain the inter-domain connection relationship set. Then, with each management domain as a node and the inter-domain connection relationships in the inter-domain connection relationship set as edges, and carrying the digest resource status of each management domain in the corresponding node attributes, a domain-level topology structure is constructed to obtain the inter-domain abstract graph, which provides a domain-level topology basis for subsequent cross-domain end-to-end path calculation.

[0083] Furthermore, step 202 may include the following sub-steps:

[0084] S21. Extract boundary node information and digest resource status from the local key resource database corresponding to each regional key management device, and encapsulate the boundary node information and digest resource status into a domain-level link status announcement message to obtain the D-LSA digest status message corresponding to each regional key management device.

[0085] S22. Exchange D-LSA digest status messages corresponding to key management devices in each region through a restricted flooding mechanism, and summarize and organize the domain boundary connection relationships and digest resource status corresponding to each management domain obtained by the exchange to obtain a set of inter-domain connection relationships.

[0086] S23. Using each management domain as a node, the inter-domain connection relationships in the set of inter-domain connection relationships as edges, and carrying the summary resource status corresponding to each management domain, construct the domain-level topology to obtain the inter-domain abstract graph.

[0087] Inter-domain connection set: refers to the set of cross-domain connection relationships between management domains across the entire network, which are exchanged and summarized by key management devices in each region through a restricted flooding mechanism. It includes the identifier of each pair of adjacent management domains, the identifier of the cross-domain link connecting the two, and the summary resource status of the link, and is used to construct the edge set of the inter-domain abstract graph.

[0088] Limited flooding mechanism: refers to the information propagation mechanism used when regional key management devices exchange D-LSA digest status messages. Each regional key management device only forwards the received D-LSA digest status messages to neighboring regional key management devices, rather than broadcasting them to all regional key management devices in the network. This is used to control management channel overhead while ensuring inter-domain state synchronization.

[0089] Summary Resource Status: Refers to the general key resource status information extracted by the regional key management device from the key resource database of this domain for inter-domain exchange. It includes the remaining key quantity and link availability status of cross-domain links, which have been summarized or quantified. It does not include the real-time status of the link key pool within this domain. It is used to provide the basis for judging the resource availability required for path calculation between domains.

[0090] It should be noted that each regional key management device queries its respective domain key resource database for boundary node information and cross-domain link status, extracting the identifiers of all boundary nodes in its domain, their respective management domains, the identifiers of connected cross-domain links, and the digest resource status such as the remaining key quantity and link availability of cross-domain links. Then, it encapsulates the message according to preset D-LSA encapsulation rules. For example, it sets the message type field to domain-level link status announcement, fills in the sender's management domain identifier, writes the boundary node information and digest resource status sequentially into the message payload in a type-length-value format, calculates the checksum, and fills it into the message tail, obtaining the D-LSA digest status message corresponding to each regional key management device. Each regional key management device sends the generated D-LSA digest status message to adjacent regional key management devices through the management communication network. After verifying the message checksum, the regional key management device extracts the domain boundary connection relationships and digest resource status carried in the message, and forwards the message to other adjacent regional key management devices except the sender, until the message is propagated to all regional key management devices in the network. Each regional key management device deduplicates and merges the domain boundary connection relationships received from different regional key management devices, and merges multiple cross-domain links between the same pair of management domains into one inter-domain connection relationship and associates it with the corresponding digest resource status, thus obtaining a set of inter-domain connection relationships. Each regional key management device uses all management domains in the network as graph nodes and sets a management domain identifier and digest resource status attribute for each node. It also uses each inter-domain connection relationship in the set of inter-domain connection relationships as a graph edge and sets a cross-domain link identifier and remaining key amount digest attribute for each edge, thus constructing a domain-level topology structure and obtaining an inter-domain abstract graph.

[0091] Step 203: Use the node management domain mapping table to determine the management domain to which the source node and destination node belong and to split the path calculation mode for the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers.

[0092] Source domain: refers to the management domain to which the source node belongs in an end-to-end key service task. It is the starting management domain of the end-to-end quantum key path and is used to determine the regional key management device that performs intra-domain path computation or participates in cross-domain path segment computation.

[0093] Destination domain: refers to the management domain to which the destination node belongs in an end-to-end key service task. It is the termination management domain of the end-to-end quantum key path and is used to determine the endpoint management domain of the end-to-end key path.

[0094] Path computation mode identifier: This is a flag parameter used to identify which path computation mode should be used for the end-to-end key service task. It includes two values: intra-domain path computation mode and cross-domain path computation mode. The value is determined based on whether the source domain and the destination domain are in the same management domain. It is used to guide the subsequent path computation process to select the intra-domain path computation branch or the cross-domain path computation branch.

[0095] It should be noted that the regional key management device extracts the source node identifier and destination node identifier from the end-to-end key service task data. Using the source node identifier and destination node identifier as query keywords, it searches the node management domain mapping table to find the management domain identifier corresponding to the source node identifier and the management domain identifier corresponding to the destination node identifier. The management domain corresponding to the source node identifier is determined as the source domain, and the management domain corresponding to the destination node identifier is determined as the destination domain. Subsequently, the management domain identifiers of the source domain and the destination domain are compared for consistency. If they are the same, it means that the source node and the destination node are in the same management domain, and the path calculation mode identifier is set to the intra-domain path calculation mode. If they are different, it means that the source node and the destination node are in different management domains, and the path calculation mode identifier is set to the cross-domain path calculation mode. This completes the determination of the management domain to which the source node and the destination node belong and the path calculation mode separation, providing a basis for subsequent selection of intra-domain path calculation or cross-domain path calculation based on the path calculation mode identifier.

[0096] Furthermore, step 203 may include the following sub-steps:

[0097] S31. Extract the source node and destination node from the end-to-end key service task data;

[0098] S32. Query the node management domain mapping table, and match the management domains of the extracted source node and destination node to obtain the source domain and destination domain;

[0099] S33. Perform a consistency comparison between the source domain and the destination domain. If they are in the same management domain, it is determined to be an intra-domain path calculation mode. If they are in different management domains, it is determined to be a cross-domain path calculation mode, and the path calculation mode identifier is obtained.

[0100] Source node: refers to the QKD node that initiates the key service request in the end-to-end key service task and serves as the starting endpoint of the quantum key relay path. Its identifier is included in the user's end-to-end key service task data and is used to determine the starting point of the end-to-end key path.

[0101] Destination node: refers to the QKD node that receives key services in the end-to-end key service task and serves as the termination endpoint of the quantum key relay path. Its identifier is included in the user's end-to-end key service task data and is used to determine the end point of the end-to-end key path.

[0102] It should be noted that after receiving the user's end-to-end key service task data, the regional key management device first parses the message structure of the task data, reads the source node identifier field and destination node identifier field from the task parameter fields, and temporarily stores the read source node identifier value and destination node identifier value in the local task context; then, the regional key management device performs an exact match search in the node management domain mapping table using the source node identifier as the key to obtain the management domain identifier corresponding to the source node identifier and determine it as the source domain; then, it performs an exact match search in the node management domain mapping table using the destination node identifier as the key to obtain the management domain identifier corresponding to the destination node identifier and determine it as the destination domain; finally... The regional key management device performs an equality comparison between the management domain identifier of the source domain and the management domain identifier of the destination domain. If they are equal, it is determined that the source node and the destination node are in the same management domain, and the path calculation mode identifier is set to the intra-domain path calculation mode. If they are not equal, it is determined that the source node and the destination node are in different management domains, and the path calculation mode identifier is set to the cross-domain path calculation mode. Through the above process of extracting the source node and destination node, querying and matching the node management domain mapping table, and comparing the consistency of the source and destination domains, the path calculation mode of the end-to-end key service task is quickly distributed, enabling subsequent path calculations to select the corresponding distributed computing branch based on whether the source and destination nodes are in the same domain.

[0103] Step 204: When the path calculation mode is identified as intra-domain path calculation mode, based on the local key resource database corresponding to the source domain and the end-to-end key service task data of the user, perform intra-domain optimal path calculation on the path from the source node to the destination node, and output the intra-domain optimal end-to-end path or intra-domain path unavailable determination result.

[0104] Optimal end-to-end path within the domain: When the source node and the destination node are located in the same management domain, the QKD relay path calculated by the regional key management device corresponding to the management domain based on the key resource database of the domain is the one from the source node to the destination node with the smallest sum of link weights. The remaining key quantity of each link on this path meets the key quantity requirements of the task.

[0105] Intra-domain path unavailable determination result: This refers to the determination result output by the regional key management device when the source node and the destination node are located in the same management domain but a connected path from the source node to the destination node cannot be found in the set of available links. It is used to inform that there is no end-to-end key relay path that meets the key resource constraints in the current domain.

[0106] It should be noted that when the path calculation mode is identified as intra-domain path calculation mode, the regional key management device corresponding to the source domain reads the key pool status and reserved key quantity of each link from the key resource database of this domain. The remaining key quantity of each link is obtained by subtracting the reserved key quantity from the current total key quantity in the key pool status. Then, the remaining key quantity of each link is compared with the key quantity required for the task in the end-to-end key service task data, eliminating links whose remaining key quantity is less than the required key quantity for the task, thus obtaining the set of available links corresponding to the source domain. Next, a preset link weight calculation formula is used to assign weights to each link in the available link set. For example, the link weight is the reciprocal of the sum of the remaining key quantity and a preset positive number, where the preset positive number is used to avoid the denominator being zero, resulting in a weighted source domain available link topology. Then, in the weighted source domain available link topology, with the source node in the end-to-end key service task data as the starting point and the destination node as the ending point, the SPF (Shortest Path First) / Dijkstra algorithm is used. The algorithm (Dixtra algorithm) performs a minimum weight path search to obtain the minimum weight path search result. Finally, it checks the existence of the minimum weight path search result. If a connected path from the source node to the destination node is found, the optimal end-to-end path in the domain is output. If no connected path is found, the path in the domain is deemed unusable. Thus, the end-to-end path calculation in the domain is completed independently on the local area key management device corresponding to the source domain, without relying on the area key management devices of other management domains. This reflects the distributed and autonomous characteristics of path calculation in the domain under the domain-aggregated architecture.

[0107] Furthermore, step 204 may include the following sub-steps:

[0108] S41. Based on the key pool status and reserved key quantity of each link in the key resource database of the source domain, calculate the remaining key quantity of each link in the source domain to obtain the remaining key quantity of multiple links in the source domain.

[0109] S42. Compare and filter the remaining key quantity of each link corresponding to the source domain with the key quantity required for the task in the end-to-end key service task data of the user, exclude links with a remaining key quantity less than the key quantity required for the task, and obtain the set of available links corresponding to the source domain.

[0110] S43. Assign weights to each link in the set of available links corresponding to the source domain to obtain a weighted topology map of available links in the source domain.

[0111] S44. In the weighted source domain available link topology graph, with the source node in the user end-to-end key service task data as the starting point and the destination node as the ending point, perform a minimum weight path search on the topology corresponding to the source domain to obtain the minimum weight path search result.

[0112] S45. Perform an existence check on the search results of the minimum weight path. If a feasible path exists, the optimal end-to-end path in the domain is obtained. If no feasible path exists, the path in the domain is deemed unusable.

[0113] It should be noted that the end-to-end key service task for users includes the source node, the destination node, and the amount of key required for the task. For a given task, the area KM first obtains the source node identifier, the destination node identifier, and the amount of key required for this task.

[0114] Unlike data forwarding in ordinary communication networks, end-to-end key service tasks in QKD networks require consuming a corresponding number of link keys on each QKD link traversed along the path. Therefore, path calculation first requires determining whether the remaining key quantity on each link meets the task requirements.

[0115] In the available links, this application sets the link weight to the reciprocal of the remaining key quantity:

[0116] ;

[0117] in, Link weight; This represents the remaining key quantity; The value is positive to avoid a denominator of zero. Therefore, the larger the remaining key amount in a link, the smaller the link weight; conversely, the smaller the remaining key amount in a link, the larger the link weight. When using SPF to calculate paths, paths with smaller sums of weights are preferred, i.e., paths with more remaining key amounts are preferred.

[0118] Furthermore, in the quantum key relay process, the end-to-end key service task needs to reach the destination node from the source node through one or more relay nodes. When the source node and the destination node are located in the same management domain, the shortest path is calculated using Dijkstra's algorithm with the remaining key quantity as the link weight.

[0119] The intra-domain path calculation process is as follows:

[0120] 1. The regional KM receives the task request and obtains the source node, destination node, and the amount of key required for the task.

[0121] 2. The regional KM queries the local key resource database to calculate the remaining key quantity for each link.

[0122] 3. Region KM excludes links where the remaining key quantity is less than the key quantity required by the task.

[0123] 4. Set weights for the remaining available links, and the link weights are given by equation (1).

[0124] 5. In the domain topology composed of available links, the SPF algorithm is used to calculate the minimum weight path from the source node to the destination node.

[0125] If a feasible path is found, output that path; otherwise, determine that the path within the domain is unavailable.

[0126] In this embodiment, the regional key management device corresponding to the source domain first reads the key pool status and reserved key quantity of each QKD link within the source domain from the key resource database of this domain. The key pool status records the total number of keys currently generated and stored for that link, and the reserved key quantity records the number of keys that have been reserved for other end-to-end key service tasks but have not yet been consumed. The regional key management device subtracts the reserved key quantity from the current total key quantity in the key pool status of each link to obtain the remaining key quantity currently available for allocation for that link. After traversing all links within the source domain, the remaining key quantities of multiple links corresponding to the source domain are obtained. Because the end-to-end key service in the QKD network... The task requires consuming a link key equal to the required key amount on each QKD link along the path. The regional key management device compares the remaining key amount of each link corresponding to the source domain with the required key amount in the end-to-end key service task data. Links with a remaining key amount less than the required key amount are marked as unavailable and excluded from the candidate links. Links with a remaining key amount greater than or equal to the required key amount are retained as available links. All available links form the available link set corresponding to the source domain. Subsequently, the regional key management device assigns weights to each link in the available link set corresponding to the source domain, according to the formula... Calculation, where Let e ​​be the link weight at time t. Let be the remaining key amount for link e at time t. To avoid using a pre-defined positive number with a denominator of zero, links with larger remaining key quantities have smaller link weights, and links with smaller remaining key quantities have larger link weights. After assignment, each available link and its weight is integrated with the QKD node topology in the source domain to obtain a weighted source domain available link topology map. Then, the regional key management device uses the source node in the user end-to-end key service task data as the search start point and the destination node as the search end point in the weighted source domain available link topology map, and performs minimum weight path search using the SPF / Dijkstra algorithm. The algorithm iteratively selects the path from the source node. The system selects the unvisited node with the smallest cumulative weight and relaxes and updates the path weights of its neighboring nodes until the destination node is reached or all reachable nodes have been traversed, thus obtaining the minimum weight path search result. Finally, the regional key management device performs an existence determination on the minimum weight path search result. If the search result contains a connected path from the source node to the destination node, the path is determined as the optimal end-to-end path in the domain. If the search result does not contain a connected path from the source node to the destination node, the system outputs a decision that the path in the domain is unavailable. Thus, the regional key management device corresponding to the source domain independently completes the calculation of the end-to-end path in the domain.

[0127] After obtaining the optimal end-to-end path within the domain, the regional key management device corresponding to the source domain sends a key resource reservation instruction to the key pool corresponding to each QKD link along the path, based on the key quantity required for the task consumed by each QKD link on the path. This marks the corresponding number of link keys from the available state to the reserved state. After completing the key resource reservation for the end-to-end path, the device distributes the end-to-end shared key to the source and destination nodes through the key delivery interface, thus completing the end-to-end key service task within the domain. If the domain path is deemed unavailable, the regional key management device corresponding to the source domain returns a notification to the user that the end-to-end key service task cannot be completed temporarily, or adds the task to a waiting queue. Once the key pool of the relevant QKD links within the domain continuously generates and replenishes keys through the QKD coding process, the domain path calculation process is retried to attempt to obtain a feasible optimal end-to-end path within the domain.

[0128] Step 205: When the path calculation mode is identified as cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each area key management device, and the user end-to-end key service task data, perform cross-domain end-to-end path calculation on the path from the source domain to the destination domain, and output the cross-domain optimal end-to-end path or cross-domain path unavailable determination result.

[0129] It should be noted that when the path calculation mode is identified as cross-domain path calculation mode, the regional key management device corresponding to the source domain first performs a breadth-first search (BFS) algorithm in the inter-domain abstract graph, starting from the source domain and ending at the destination domain. This involves traversing the management domain nodes in the inter-domain abstract graph layer by layer and recording the predecessor management domains reaching each management domain until the destination domain is found. Then, the predecessor relationships are traced back to obtain the shortest domain sequence from the source domain to the destination domain. Subsequently, for each pair of adjacent management domains in the shortest domain sequence, all candidate cross-domain links between the pair are queried from the cross-domain link information stored in the key resource database of each regional key management device's corresponding domain. The remaining key quantity of each candidate cross-domain link is compared, and the cross-domain link with the largest remaining key quantity is selected. The QKD nodes at both ends of this link are determined as the boundary nodes of the pair of adjacent management domains, thus obtaining the boundary node pairs between adjacent management domains. Finally, based on the shortest domain sequence and the boundary node pairs, the entry and exit boundary nodes of each passing management domain are determined. For each boundary node, the intra-domain path calculation method is used to calculate the intra-domain path segments from the entry boundary node to the exit boundary node for each management domain. Then, in the order of the shortest domain sequence, the intra-domain path segments corresponding to each management domain are connected end to end through the corresponding boundary nodes to obtain the preliminary end-to-end path. Finally, the key resource availability and path connectivity of each link on the preliminary end-to-end path are verified to confirm that the remaining key quantity of each link on the path meets the key quantity required by the task and that each path segment is connected without error. If the verification passes, the optimal cross-domain end-to-end path is obtained; if the verification fails, the cross-domain path is determined to be unavailable. Thus, the cross-domain end-to-end path calculation is completed by combining coarse-grained search of the inter-domain abstract graph with fine-grained calculation within each management domain.

[0130] Among them, the shortest domain sequence refers to the sequence of management domains obtained by searching from the source domain to the destination domain in the inter-domain abstract graph using the BFS algorithm. There are cross-domain links between two adjacent management domains in the sequence, and the sequence contains the fewest number of management domains. It is used to determine the management domains that need to be passed through and the order of passing through them in the cross-domain end-to-end path.

[0131] Boundary node pair: refers to the QKD nodes at both ends of the cross-domain link with the largest remaining key quantity selected after comparing the remaining key quantities of candidate cross-domain links between each pair of adjacent management domains in the shortest domain sequence. One node belongs to the previous management domain and the other node belongs to the next management domain. It is used to connect intra-domain path segments of adjacent management domains.

[0132] Passing through management domains: These are the intermediate management domains that the cross-domain end-to-end path needs to pass through in the shortest domain sequence, excluding the source and destination domains. Each passing through management domain requires calculating the intra-domain path segment from the ingress boundary node to the egress boundary node.

[0133] Entry boundary node: For a given management domain, the boundary node that the cross-domain end-to-end path passes through when entering the management domain from the adjacent previous management domain is the starting point for the calculation of the path segment within the management domain.

[0134] Exit boundary node: For a given management domain, the boundary node that the cross-domain end-to-end path passes through when traveling from that management domain to the next adjacent management domain is the endpoint of the path segment calculation within that management domain.

[0135] Intra-domain path segment: refers to the QKD relay path calculated by the regional key management devices corresponding to each management domain using the intra-domain path calculation method within this management domain, from the ingress boundary node to the egress boundary node, and is a component of the cross-domain end-to-end path.

[0136] Preliminary end-to-end path: refers to the cross-domain end-to-end path obtained by connecting the intra-domain path segments corresponding to each management domain in the order of the shortest domain sequence through the corresponding boundary nodes. It has not yet undergone the final key resource availability and path connectivity verification.

[0137] Cross-domain optimal end-to-end path: refers to the QKD relay path from the source node to the destination node when the source node and the destination node are located in different management domains. This path is obtained by searching the shortest domain sequence of the inter-domain abstract graph, selecting boundary nodes, calculating and splicing path segments within each management domain, and finally verifying the results. The path passes through multiple management domains and the key resources of each link meet the task requirements.

[0138] Cross-domain path unavailable determination result: This refers to the determination result output by the regional key management device when the source node and the destination node are located in different management domains but an end-to-end path that meets the key resource constraints and connectivity requirements cannot be found during the cross-domain path calculation process. It is used to inform that there is currently no feasible cross-domain end-to-end key relay path.

[0139] Furthermore, step 205 may include the following sub-steps:

[0140] S51. In the inter-domain abstraction graph, with the source domain as the starting point and the destination domain as the ending point, perform a shortest domain sequence search on the inter-domain abstraction graph to obtain the shortest domain sequence from the source domain to the destination domain.

[0141] S52. For each pair of adjacent management domains in the shortest domain sequence, compare the remaining key amount of each pair of candidate cross-domain links from the cross-domain link information stored in the local key resource database corresponding to each regional key management device, select the cross-domain link with the largest remaining key amount, and determine the nodes at both ends of the link as the boundary nodes of the adjacent management domains to obtain the boundary node pair between adjacent management domains.

[0142] S53. Based on the shortest domain sequence and boundary node pairs, determine the entry boundary node and exit boundary node of each transit management domain, and use the intra-domain path calculation method to calculate the intra-domain path segments for each transit management domain to obtain the intra-domain path segments corresponding to each transit management domain.

[0143] S54. Following the order of the shortest domain sequence, connect the intra-domain path segments corresponding to each management domain through the corresponding boundary nodes in sequence to obtain the preliminary end-to-end path.

[0144] S55. Verify the availability of key resources and path connectivity of each link on the initial end-to-end path. If the verification passes, the optimal cross-domain end-to-end path is obtained; if the verification fails, the cross-domain path is deemed unavailable.

[0145] It should be noted that, as Figure 3 As shown, when the source node and destination node are located in different management domains, the region KM first combines the "node-management domain" mapping table to abstract each management domain and cross-domain connections into "points" and "edges" of an inter-domain graph. Then, it uses the BFS algorithm on this abstract graph to calculate the shortest domain sequence from the source domain to the destination domain. Subsequently, for adjacent management domains, the boundary nodes at both ends are established by prioritizing the link with the largest remaining key quantity. Finally, each management domain calculates its own path segments and collaboratively combines them into an end-to-end path.

[0146] The cross-domain path calculation process is as follows:

[0147] 1. Determine the source domain and destination domain.

[0148] 2. Use Breadth-First Search (BFS) to compute the shortest domain sequence. In the inter-domain abstract graph, use Breadth-First Search (BFS) to compute the shortest domain sequence from the source domain to the destination domain.

[0149] 3. Determine the boundary nodes between adjacent management domains. For any two adjacent management domains in the shortest domain sequence, if there are multiple candidate cross-domain links, select the cross-domain link with the largest remaining key quantity, and use the nodes at both ends of this cross-domain link as the boundary nodes between the adjacent management domains.

[0150] 4. Use the intra-domain path calculation method to calculate the path segments passing through each management domain.

[0151] 5. Assemble end-to-end paths.

[0152] 6. Perform path verification.

[0153] Specifically, when the path calculation mode is identified as a cross-domain path calculation mode, the regional key management device corresponding to the source domain first performs a breadth-first search on the inter-domain abstract graph, using the source domain as the search starting point and the destination domain as the search ending point. Starting from the source domain node, it visits adjacent management domain nodes that have cross-domain connections with the current management domain node layer by layer, recording the predecessor management domain node of each visited management domain node, until the destination domain node is visited for the first time. Then, by backtracking the predecessor management domain node relationships, it obtains the shortest domain sequence from the source domain to the destination domain. Subsequently, for each pair of adjacent management domains in the shortest domain sequence, the regional key management device... The management device queries the cross-domain link information stored in the local key resource database corresponding to each regional key management device for all candidate cross-domain links between adjacent management domains. It reads the remaining key quantity of each candidate cross-domain link and compares them, selecting the candidate cross-domain link with the largest remaining key quantity as the connection link between the adjacent management domains. The QKD nodes at both ends of this cross-domain link, belonging to the two management domains respectively, are determined as the boundary nodes between the adjacent management domains, thus obtaining the boundary node pairs between adjacent management domains. Then, based on the order of the shortest domain sequence and the boundary node pairs between each pair of adjacent management domains, the entry edge of each passing management domain is determined. Boundary nodes and exit boundary nodes are defined, where the entry boundary node is the boundary node between the passing management domain and the previous adjacent management domain, and the exit boundary node is the boundary node between the passing management domain and the next adjacent management domain. An intra-domain path calculation method is used to calculate the intra-domain path segment from the entry boundary node to the exit boundary node for each passing management domain, obtaining the intra-domain path segment corresponding to each passing management domain. Then, according to the order of management domains in the shortest domain sequence, the intra-domain path segments corresponding to each passing management domain are sequentially connected end-to-end through the corresponding boundary nodes, i.e., the endpoint of the intra-domain path segment of the previous management domain is connected to the endpoint of the intra-domain path segment of the next management domain. The starting point of each intra-domain path segment is the same boundary node, resulting in a preliminary end-to-end path. Finally, the availability of key resources and the connectivity of each link on the preliminary end-to-end path are verified. It is confirmed that the remaining key quantity of each QKD link on the path is greater than or equal to the key quantity required by the task, and that the connection between each intra-domain path segment and the cross-domain link is correct. If the verification passes, the preliminary end-to-end path is determined as the optimal cross-domain end-to-end path. If the verification fails, the cross-domain path is deemed unavailable. Thus, the cross-domain end-to-end path calculation is completed by combining coarse-grained domain sequence search of the inter-domain abstract graph with fine-grained path segment calculation within each management domain.

[0154] Among them, candidate cross-domain links refer to all cross-domain links that exist between each pair of adjacent management domains in the shortest domain sequence and can be used to connect the two management domains. They are the comparison objects when selecting boundary nodes. Each candidate cross-domain link is stored in the local key resource database of the corresponding area key management device and records the remaining key quantity.

[0155] Key resource availability: This refers to whether the remaining key quantity of each QKD link on the initial end-to-end path meets the key quantity requirements of the end-to-end key service task. It is one of the core verification items in the final verification of cross-domain paths and is used to ensure that each link on the path has enough keys to consume.

[0156] Path connectivity: refers to whether the intra-domain path segments and cross-domain links in the initial end-to-end path are correctly connected through boundary nodes, and whether the entire path is continuous without breaks from the source node to the destination node. It is one of the core verification items for the final verification of cross-domain paths, and is used to ensure that the path is physically passable.

[0157] After obtaining the optimal cross-domain end-to-end path, the regional key management devices corresponding to the management domains traversed on the shortest domain sequence send key resource reservation instructions to the key pools corresponding to the QKD links traversed by the path segments within their respective domains and the key pools corresponding to the cross-domain links, based on the key quantity required for the task consumed by each QKD link in the optimal cross-domain end-to-end path. This marks the corresponding number of link keys from available to reserved. After reserving key resources within their respective domains, each regional key management device collaboratively distributes the end-to-end shared key to the source and destination nodes through the key delivery interface, completing this cross-domain end-to-end key service task. If a cross-domain path unavailable determination is obtained, the regional key management device corresponding to the source domain returns a notification to the user that this end-to-end key service task cannot be completed temporarily, or adds the task to a waiting queue. Once the QKD link key pools within the relevant management domains continuously generate and replenish keys through the QKD coding process, the cross-domain path calculation process is re-triggered to attempt to obtain a feasible optimal cross-domain end-to-end path.

[0158] For comparison of technical effects, existing technologies can be referenced. Quantum Key Distribution (QKD) generates a consistent symmetric key between communicating parties through quantum state preparation, transmission, reception, and post-processing. Fiber optic QKD links are typically deployed between adjacent QKD nodes and continuously generate link keys, which are stored in the key pool of the corresponding link or node. For end-to-end key service requests where the source and destination nodes are not adjacent, trusted relay nodes and multi-hop paths are usually required to establish end-to-end keys.

[0159] In QKD network engineering implementations, the key management layer typically maintains key pool status, link status, task requirements, and key delivery interfaces. User end-to-end key service tasks typically include parameters such as source node, destination node, required key quantity, maximum acceptable latency, and service priority. The key management layer determines the QKD links and trusted relay nodes that the task should traverse based on pre-configured paths, shortest paths, available key pool capacity, or dynamic routing policies.

[0160] Unlike data routing in conventional communication networks, path selection in QKD networks must consider not only topology hop count or link latency, but also the generation and consumption characteristics of quantum key resources. Specifically, each QKD link on an end-to-end key relay path consumes a number of link keys matching the task's key requirements; the link key pool dynamically increases during the QKD coding process and dynamically decreases during service reservations or key distribution. Therefore, the optimal path in a QKD network is not necessarily the shortest topological path, but rather a quantum key service path that satisfies key resource requirements.

[0161] Furthermore, in a single-KM centralized scheme, a central hub is set up across the entire network, or a centralized control and management entity manages the entire network's QKD topology, link key pool, task queue, and link load status. After a user submits an end-to-end key service task, the central KM uniformly calculates the path from the source node to the destination node, uniformly issues key resource reservation instructions, and uniformly handles path failures, recalculations, and resource releases.

[0162] These schemes typically rely on dynamic routing based on the overall network state to achieve end-to-end key relay. When a user node's key pool falls below a threshold or receives a key service request, the central KM calculates a feasible path based on a preset path or the overall network topology and link key pool availability. Existing QKD dynamic routing research has also begun to utilize link key state and network state for path selection. However, as the scale of QKD devices, links, and end-to-end tasks grows, the central KM is prone to computational bottlenecks, state synchronization bottlenecks, and single-point-of-failure risks.

[0163] In a device-level distributed key manager (KM) scheme, an independent KM is set up near each QKD device. Each KM device mainly maintains the key pool status, key generation status, key synchronization information, and local key delivery information related to its own device or link. When an end-to-end task crosses multiple QKD links, multiple device-level KMs need to interact hop-by-hop to complete local state synchronization.

[0164] The advantage of this type of scheme is that a single KM has a relatively small management scope and strong local autonomy. However, since the number of KMs increases rapidly with the number of QKD devices or links, end-to-end tasks across devices, links, and regions require a large amount of signaling interaction between KMs. If KMs only exchange a small amount of synchronization information, or even no real-time status, problems such as outdated status, fragmented path calculation, and inconsistent failure rollback may occur.

[0165] Existing QKD network research typically uses Shortest Path First (SPF) to calculate the end-to-end shortest path, and further incorporates quantum key resource factors such as remaining key quantity, key generation rate, and link load into path weights or path feasibility assessments. This type of scheme primarily focuses on path security, link load balancing capabilities, or online routing efficiency.

[0166] However, existing research has not clearly proposed how to calculate the end-to-end key path under the key management layer architecture of "multiple QKD devices corresponding to one region KM".

[0167] In summary, the disadvantages of existing technologies include:

[0168] 1. The centralized KM solution suffers from central bottlenecks and single points of failure. The central KM needs to continuously maintain the status of all QKD devices, link key pools, task queues, and paths across the network. As the number of QKD devices, links, management domains, and tasks increases, the computational, storage, status collection, and command issuance pressures on the central KM increase significantly, easily becoming a system performance bottleneck. If the central KM fails, network-wide path calculation, task scheduling, and key resource reservation may become impossible.

[0169] 2. The centralized single-KM scheme incurs significant overhead for network-wide status synchronization. To ensure accurate path calculation, the central KM needs to frequently retrieve the remaining key pool information, coding rate, link load, and fault status of the entire network links. For large-scale QKD networks, uploading the entire status can easily lead to high management channel overhead, high status update latency, and inconsistent status.

[0170] 3. Device-level distributed KM solutions suffer from excessive KM numbers and fragmented end-to-end path computation. If each QKD device corresponds to one KM, end-to-end tasks spanning multiple links require multiple device-level KMs to participate hop-by-hop, leading to an increase in signaling interactions. Furthermore, a single device-level KM cannot form a holistic view of multiple QKD devices, multiple links, and resources within the region.

[0171] 4. Device-level distributed KM schemes are prone to inconsistencies in boundary link states. When multiple device-level KMs are not sufficiently synchronized, they may make inconsistent judgments on the same boundary link or the same key pool. If a segment of a cross-domain path becomes unavailable, the resources already occupied or reserved by the preceding link need to be released, increasing control complexity.

[0172] 6. Existing research has not solved the problem of calculating the end-to-end optimal path under the domain-aggregate quantum key management layer architecture. Although existing research can provide dynamic routing ideas, it does not provide a complete method for calculating the end-to-end path for the architecture where "multiple QKD devices correspond to one region KM".

[0173] To address the aforementioned problems, this invention provides an end-to-end path calculation method based on a domain-aggregated quantum key management layer architecture, relating to the field of quantum key distribution network technology, and particularly to a quantum key management layer architecture, a regional key management device, an end-to-end quantum key service path calculation method, a quantum key distribution network, an electronic device, and a computer-readable storage medium.

[0174] More specifically, this application relates to a domain management architecture in which multiple QKD devices are aggregated and managed by a region KM, and a method for calculating an end-to-end quantum key path based on the precise quantum key resource state within the domain and by splicing path fragments within the domain under this architecture.

[0175] In this invention, multiple QKD devices, QKD nodes, QKD links and their key pools are divided into several management domains, and each management domain is configured with a region KM.

[0176] Each regional KM is responsible for managing the key pool status, coding rate, link load, device status, task queue, and intra-domain topology of multiple QKD devices within its domain, and calculating intra-domain or cross-domain path segments based on the precise status of its domain. Limited flooding between regional KMs is achieved through domain-level link state announcements, and each regional KM maintains an inter-domain abstract graph based on the received D-LSA.

[0177] For end-to-end key service tasks, this application first determines the management domain to which the source node and destination node belong. If they belong to the same management domain, the KM of that region directly calculates the optimal intra-domain path based on the precise state of the domain. If they belong to different management domains, the optimal domain sequence is first calculated on the inter-domain abstract graph, and then the KM of each region on the domain sequence calculates the intra-domain path segments respectively. Finally, the segments are spliced ​​into an optimal end-to-end path through boundary nodes or cross-domain links.

[0178] As can be seen from the above, the present invention provides a domain-aggregated quantum key management layer architecture, which enables a regional KM to aggregate and manage multiple QKD devices, QKD links and link key pools within its domain. By "direct reporting of accurate state within the domain and flooding of D-LSA digest state between domains", the centralized management bottleneck is reduced while the large amount of signaling interaction brought about by device-level distributed KM is reduced.

[0179] Meanwhile, this invention provides an end-to-end quantum key path calculation method applicable to the above-mentioned domain-aggregated architecture. It decomposes cross-domain key path calculation into inter-domain abstract graph search, boundary node selection, intra-domain path segment calculation, and end-to-end path splicing. Furthermore, it incorporates QKD network characteristics such as link key pool reserve, reserved amount, security key rate, key waiting time, trusted relay constraints, quantum link quality, and fault status into the path cost and feasibility judgment, thereby calculating the optimal end-to-end path that satisfies edge-level key resource constraints and service latency constraints.

[0180] Compared to a centralized single-KM solution, this application reduces the bottleneck of central computing and the risk of single-point failure. This application divides all QKD devices in the network into multiple management domains. Each regional KM only maintains the precise state of its own domain and is responsible for calculating paths or path segments within that domain. Cross-domain communication is achieved solely through domain-level summary states, enabling the exchange of information required for path searching. Therefore, it eliminates the need for a single central KM to continuously maintain the state of all fine-grained links across the entire network.

[0181] Compared to centralized state synchronization across the entire network, this application reduces state propagation overhead. It avoids propagating the entire link key pool and device-level state across the entire network.

[0182] Compared to device-level distributed KM solutions, this application reduces the number of KMs and the complexity of hop-by-hop signaling. Multiple QKD devices are aggregated and managed by a single regional KM, and the regional KM uniformly calculates paths for multiple links within the domain. Cross-domain tasks only require path segment calculation and result return at the regional KM level.

[0183] In summary, the cross-domain path computation method proposed in this application, under a domain-aggregated quantum key management architecture, constructs and maintains an inter-domain abstract graph based on D-LSA-restricted flooding among the key management devices in each region. This decomposes the cross-domain end-to-end path computation into two levels: searching for the minimum-cost domain sequence on the inter-domain abstract graph and calculating the precise path fragment within each passing management domain. The cross-domain path first determines the shortest domain sequence from the source domain to the destination domain at the level of the key management devices. Then, each key management device calculates the precise path fragment from the entry boundary node to the exit boundary node within its own domain based on the shortest domain sequence. The path fragments are then pieced together into an end-to-end path through boundary nodes. This transforms path computation from a fine-grained search across the entire network to a combination of coarse-grained search between domains and fine-grained search within a domain. This avoids the computational bottlenecks and single-point-of-failure risks associated with a single central node maintaining the fine-grained state of the entire network and uniformly calculating paths in a centralized single-KM scheme. It also avoids the state inconsistencies and path fragmentation caused by numerous hop-by-hop interactions between device-level KMs in a device-level distributed scheme. This provides a feasible technical approach for efficient and reliable path computation for end-to-end key services in large-scale QKD networks.

[0184] Please see Figure 4 , Figure 4 This is a structural block diagram of an end-to-end path computing system based on a domain-aggregate quantum key management layer architecture, provided for an embodiment of the present invention.

[0185] This invention provides an end-to-end path computing system based on a domain-aggregate quantum key management layer architecture. The system is applied to this architecture, which includes multiple management domains. Each management domain contains multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain also has a corresponding regional key management device, which is communicatively connected to each QKD device within its respective management domain. Each regional key management device is equipped with... The system includes: a local key resource database connected to the regional key management devices; each regional key management device associated with a node management domain mapping table; adjacent management domains connected via boundary nodes and cross-domain links, where the boundary nodes are QKD nodes belonging to two adjacent management domains and connected via cross-domain links; limited flooding connections between regional key management devices via domain-level link status announcement messages; and each regional key management device maintaining an inter-domain abstract graph, where each management domain is a node and the cross-domain connection relationships between management domains are edges.

[0186] The acquisition module 401 is used to acquire the QKD network topology data and user end-to-end key service task data, and to perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, and output the node management domain mapping table and the local key resource database corresponding to each regional key management device.

[0187] Output module 402 is used to generate D-LSA digests and perform inter-domain restricted flooding processing on the local key resource database corresponding to each of the regional key management devices, and output an inter-domain abstract graph.

[0188] The splitting module 403 is used to use the node management domain mapping table to determine the management domain to which the source node and destination node belong and to split the path calculation mode of the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers.

[0189] The intra-domain path determination module 404 is used to calculate the intra-domain optimal path for the path from the source node to the destination node based on the local key resource database corresponding to the source domain and the user end-to-end key service task data when the path calculation mode is identified as intra-domain path calculation mode, and output the intra-domain optimal end-to-end path or intra-domain path unavailable determination result.

[0190] The cross-domain path determination module 405 is used to perform cross-domain end-to-end path calculation on the path from the source domain to the destination domain based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each of the regional key management devices, and the user end-to-end key service task data when the path calculation mode is identified as cross-domain path calculation mode, and output the determination result of the optimal cross-domain end-to-end path or the unavailable cross-domain path.

[0191] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the system and modules described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0192] This invention also provides a computer device, including a memory and a processor, wherein the memory stores a computer program; when the computer program is executed by the processor, the processor performs the steps of the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described in the above embodiments.

[0193] This invention also provides a computer-readable storage medium storing a computer program / instructions thereon, which, when executed by a processor, implements the steps of the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described in the above embodiments.

[0194] This invention also provides a computer program product, including a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, wherein when the program instructions are executed by a computer, the computer performs the steps of the end-to-end path computation method based on the domain-aggregate quantum key management layer architecture as described in the above embodiments.

[0195] In the embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0196] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0197] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0198] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0199] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. An end-to-end path computation method based on a domain-aggregate quantum key management layer architecture, characterized in that, This invention relates to a domain-aggregated quantum key management layer architecture, which includes multiple management domains. Each management domain contains multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain has a corresponding regional key management device, which is communicatively connected to each QKD device within its domain. Each regional key management device has a local key resource database, which is connected to the device itself. Each regional key management device is associated with a node management domain mapping table. Adjacent management domains are connected through boundary nodes and cross-domain links. The boundary nodes are QKD nodes belonging to two adjacent management domains and connected by cross-domain links. The key management devices in each region establish restricted flooding connections through domain-level link status announcement messages; Each regional key management device maintains an inter-domain abstraction graph, wherein each management domain is a node and the cross-domain connection relationships between management domains are edges; the method includes: Acquire the QKD network topology data and user end-to-end key service task data, and perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, outputting the node management domain mapping table and the local key resource database corresponding to each regional key management device; D-LSA digest generation and inter-domain restricted flooding processing are performed on the local key resource database corresponding to each of the aforementioned regional key management devices to output an inter-domain abstract graph. The node management domain mapping table is used to determine the management domain to which the source node and destination node belong and to split the path calculation mode of the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers; When the path calculation mode is identified as the intra-domain path calculation mode, based on the local key resource database corresponding to the source domain and the user end-to-end key service task data, the intra-domain optimal path is calculated for the path from the source node to the destination node, and the intra-domain optimal end-to-end path or intra-domain path unavailable determination result is output. When the path calculation mode is identified as cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each regional key management device, and the user end-to-end key service task data, cross-domain end-to-end path calculation is performed on the path from the source domain to the destination domain, and the result of determining whether the cross-domain optimal end-to-end path or the cross-domain path is unavailable is output.

2. The end-to-end path computation method based on a domain-aggregate quantum key management layer architecture according to claim 1, characterized in that, The step of dividing the QKD network topology data into management domains and configuring the aggregation management of regional key management devices, outputting a node management domain mapping table and a key resource database for each regional key management device, includes: The QKD network topology data is divided into management domains to obtain multiple management domains; Configure a regional key management device for each management domain, and establish a one-to-one correspondence between each management domain and its corresponding regional key management device to obtain the management scope configuration of the regional key management device for each management domain. Based on the management scope configuration of each regional key management device, the key pool status, coding rate, link load, device status, task queue and domain topology of each QKD device in the management domain corresponding to each regional key management device are collected and aggregated for management, so as to obtain the domain key resource database corresponding to each regional key management device. For each QKD node in each management domain, a correspondence is established between each QKD node and its respective management domain, resulting in a node management domain mapping table.

3. The end-to-end path computation method based on a domain-aggregate quantum key management layer architecture according to claim 1, characterized in that, The step of generating D-LSA digests and performing inter-domain restricted flooding processing on the local key resource database corresponding to each of the aforementioned regional key management devices, and outputting an inter-domain abstract graph, includes: Extract boundary node information and digest resource status from the local key resource database corresponding to each regional key management device, and encapsulate the boundary node information and digest resource status into a domain-level link status announcement message to obtain the D-LSA digest status message corresponding to each regional key management device. The D-LSA digest status messages corresponding to the key management devices in each region are exchanged through a restricted flooding mechanism. The domain boundary connection relationships and digest resource status of each management domain obtained by the exchange are summarized and organized to obtain a set of inter-domain connection relationships. Using each management domain as a node, the inter-domain connection relationships in the set of inter-domain connection relationships as edges, and carrying the summary resource status corresponding to each management domain, a domain-level topology structure is constructed to obtain an inter-domain abstract graph.

4. The end-to-end path computation method based on a domain-aggregate quantum key management layer architecture according to claim 1, characterized in that, The step of using the node management domain mapping table to determine the management domain of the source and destination nodes and to perform path calculation mode splitting on the user end-to-end key service task data, and outputting source domain, destination domain, and path calculation mode identifiers, includes: Extract the source node and destination node from the end-to-end key service task data; Query the node management domain mapping table, and match the management domains of the extracted source and destination nodes to obtain the source domain and destination domain; The source domain and the destination domain are compared for consistency. If they are in the same management domain, it is determined to be an intra-domain path calculation mode. If they are in different management domains, it is determined to be a cross-domain path calculation mode, and a path calculation mode identifier is obtained.

5. The end-to-end path computation method based on a domain-aggregate quantum key management layer architecture according to claim 1, characterized in that, When the path calculation mode is identified as an intra-domain path calculation mode, based on the local key resource database corresponding to the source domain and the user end-to-end key service task data, the intra-domain optimal path is calculated for the path from the source node to the destination node, and the intra-domain optimal end-to-end path or intra-domain path unavailable determination result is output, including: Based on the key pool status and reserved key quantity of each link in the key resource database of the source domain, the remaining key quantity of each link corresponding to the source domain is calculated to obtain the remaining key quantity of multiple links corresponding to the source domain. The remaining key quantity of each link corresponding to the source domain is compared and filtered with the key quantity required for the task in the end-to-end key service task data of the user, and links with a remaining key quantity less than the task requirement are excluded to obtain the set of available links corresponding to the source domain. Weights are assigned to each link in the set of available links corresponding to the source domain to obtain a weighted topology map of available links in the source domain. In the weighted source domain available link topology map, taking the source node in the user end-to-end key service task data as the starting point and the destination node as the ending point, a minimum weight path search is performed on the topology corresponding to the source domain to obtain the minimum weight path search result. The existence of the minimum weight path search result is determined. If a feasible path exists, the optimal end-to-end path in the domain is obtained. If no feasible path exists, the path in the domain is deemed unusable.

6. The end-to-end path computation method based on a domain-aggregate quantum key management layer architecture according to claim 1, characterized in that, When the path calculation mode is identified as a cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each of the regional key management devices, and the user end-to-end key service task data, a cross-domain end-to-end path calculation is performed on the path from the source domain to the destination domain, and the result of determining whether the cross-domain optimal end-to-end path or the cross-domain path is unavailable is output, including: In the inter-domain abstraction graph, with the source domain as the starting point and the destination domain as the ending point, a shortest domain sequence search is performed on the inter-domain abstraction graph to obtain the shortest domain sequence from the source domain to the destination domain. For each pair of adjacent management domains in the shortest domain sequence, the remaining key quantity of each pair of candidate cross-domain links between adjacent management domains is compared from the cross-domain link information stored in the local key resource database corresponding to each regional key management device. The cross-domain link with the largest remaining key quantity is selected, and the nodes at both ends of the link are determined as the boundary nodes of the adjacent management domains, thus obtaining the boundary node pair between adjacent management domains. Based on the shortest domain sequence and the boundary node pair, the entry boundary node and exit boundary node of each transit management domain are determined. The intra-domain path calculation method is used to calculate the intra-domain path segment for each transit management domain to obtain the intra-domain path segment corresponding to each transit management domain. According to the order of the shortest domain sequence, the intra-domain path segments corresponding to each management domain are connected sequentially through the corresponding boundary nodes to obtain the preliminary end-to-end path; The availability of key resources and the connectivity of each link on the initial end-to-end path are verified. If the verification passes, the optimal cross-domain end-to-end path is obtained; if the verification fails, the cross-domain path is deemed unavailable.

7. An end-to-end path computing system based on a domain-aggregate quantum key management layer architecture, characterized in that, This invention relates to a domain-aggregated quantum key management layer architecture, which includes multiple management domains. Each management domain contains multiple QKD devices, multiple QKD nodes, and QKD links connecting adjacent QKD nodes. Each QKD link corresponds to a key pool. Each management domain has a corresponding regional key management device, which is communicatively connected to each QKD device within its domain. Each regional key management device has a local key resource database, which is connected to the device itself. Each regional key management device is associated with a node management domain mapping table. Adjacent management domains are connected through boundary nodes and cross-domain links. The boundary nodes are QKD nodes belonging to two adjacent management domains and connected by cross-domain links. The key management devices in each region establish restricted flooding connections through domain-level link status announcement messages; Each regional key management device maintains an inter-domain abstraction graph, which uses each management domain as a node and the cross-domain connection relationship between management domains as an edge. The system includes: The acquisition module is used to acquire the QKD network topology data and user end-to-end key service task data, and to perform management domain division and regional key management device aggregation management configuration on the QKD network topology data, and output the node management domain mapping table and the local key resource database corresponding to each regional key management device. The output module is used to generate D-LSA digests and perform inter-domain restricted flooding processing on the local key resource database corresponding to each of the regional key management devices, and output an inter-domain abstract graph. The splitting module is used to use the node management domain mapping table to determine the management domain to which the source node and destination node belong and to split the path calculation mode of the user end-to-end key service task data, and output the source domain, destination domain and path calculation mode identifiers. The intra-domain path determination module is used to calculate the optimal intra-domain path for the path from the source node to the destination node based on the local key resource database corresponding to the source domain and the user end-to-end key service task data when the path calculation mode is identified as intra-domain path calculation mode, and output the determination result of the optimal intra-domain end-to-end path or the intra-domain path being unavailable. The cross-domain path determination module is used to perform cross-domain end-to-end path calculation on the path from the source domain to the destination domain when the path calculation mode is identified as cross-domain path calculation mode, based on the source domain and destination domain, the inter-domain abstract graph, the local key resource database corresponding to each of the regional key management devices, and the user end-to-end key service task data, and output the determination result of the optimal cross-domain end-to-end path or the unavailable cross-domain path.

8. An electronic device, characterized in that, The system includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor causes the processor to perform the steps of the end-to-end path computation method based on a domain-aggregate quantum key management layer architecture as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed, it implements the end-to-end path computation method based on a domain-aggregate quantum key management layer architecture as described in any one of claims 1-6.

10. A computer program product, characterized in that, The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions, wherein, when the program instructions are executed by a computer, the computer causes the computer to perform the steps of the end-to-end path computation method based on a domain-aggregate quantum key management layer architecture as described in any one of claims 1-6.