Intelligent calculation center optical network optical path survivability-oriented air-ground cooperative unmanned aerial vehicle distributed optimization deployment method and system

Abstract

The invention relates to an intelligent calculation center optical network optical path survivability-oriented air-ground cooperative unmanned aerial vehicle distributed optimization deployment method and system, and belongs to the technical field of optical communication network protection. The invention aims to solve the technical problems of long switching time delay, limited coverage and poor stability of a traditional single ground or air protection mechanism. According to the technical scheme, an air-ground cooperative protection architecture is constructed, and firstly, an integer linear programming model with the aim of minimizing the total cost of a system is established to carry out precise deployment optimization of unmanned aerial vehicle nodes; in order to improve the solving efficiency, a spectral clustering and K-means clustering-based heuristic algorithm is designed to realize rapid intelligent division of the unmanned aerial vehicle service domain; the deployed unmanned aerial vehicle cluster adopts a master-slave architecture and has a failover mechanism, and a temporary relay network can be quickly established when a ground optical path fails. According to the invention, the survivability and fault recovery efficiency of the optical network are effectively improved, and good balance between deployment cost and protection performance is realized.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of optical communication network protection, and relates to an air-ground cooperative unmanned aerial vehicle distributed optimization deployment method and system for optical path survivability of an intelligent center optical network. BACKGROUND

[0002] With the deepening of the global digitalization process, the intelligent center optical network has become a key infrastructure for the normal operation of society, and the continuity and reliability of its communication links are crucial. Optical path protection is one of the core challenges to improve the survivability of the intelligent center optical network.

[0003] Traditional intelligent center optical networks mainly rely on ground backup optical paths to provide redundant protection. Although this mechanism has certain protection capabilities, it has obvious limitations in practical applications. On the one hand, the switching delay of backup routes is long, which is difficult to meet the strict requirements of real-time services. On the other hand, the capacity and coverage of ground routes are limited, and they often lack the ability to cope with large-scale regional failures.

[0004] At the same time, a single scheme relying on air network communication also faces significant challenges. Although unmanned aerial vehicle communication has the advantages of flexible deployment and rapid maneuvering, its transmission link is easily affected by complex terrain shielding and climate changes, resulting in a decline in transmission stability. In addition, uncertain factors such as airspace regulation policies also make it difficult to serve as a long-term stable independent communication guarantee scheme.

[0005] Therefore, there is a lack of an effective fusion of ground and air resource advantages in the prior art to achieve an efficient, reliable, and cost-controlled optical path survivability protection scheme. An innovative architecture and method are needed to overcome the inherent limitations of single-domain protection and meet the stringent requirements of intelligent center optical networks for high reliability and low latency. SUMMARY

[0006] Therefore, the purpose of the present application is to provide an air-ground cooperative unmanned aerial vehicle distributed optimization deployment method and system for optical path survivability of an intelligent center optical network.

[0007] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0008] An air-ground cooperative unmanned aerial vehicle distributed optimization deployment method for optical path survivability of an intelligent center optical network, comprising the following steps:

[0009] Obtaining topology structure information and high-failure-probability area data of the intelligent center optical network;

[0010] Based on the topology structure information and the high-failure-probability area data, the deployment position and number of the unmanned aerial vehicle cluster are calculated by an optimization algorithm;

[0011] Based on the deployment locations and quantities, a drone swarm is pre-deployed to construct a three-dimensional protection system;

[0012] And when the optical path fails, the UAV cluster is scheduled to provide communication relay services to realize the migration and reconstruction of the optical path.

[0013] Furthermore, the optimization algorithm includes constructing an Integer Linear Programming (ILP) model, which aims to minimize the total system cost and takes into account unmanned aerial vehicle (UAV) service coverage and capacity constraints.

[0014] Furthermore, the total system cost of the ILP model includes the basic operating cost of the UAV, the additional cost of the main UAV, the service latency cost of the ground node, and the backhaul link latency cost between UAV clusters. The ILP model also includes constraints that ensure each ground node is served by a UAV cluster, the UAV service capabilities meet the requirements, and the logical relationship of UAV cluster deployment.

[0015] Furthermore, the optimization algorithm includes the use of a heuristic clustering algorithm, which performs clustering based on optical network topology nodes to determine the UAV service domain.

[0016] Furthermore, the heuristic clustering algorithm is the K-means algorithm. The K-means algorithm identifies the node feature space through spectral clustering technology and uses K-means clustering to divide the UAV swarm.

[0017] Furthermore, the drone swarm adopts a master-slave architecture, in which the master drone is responsible for task coordination and backhaul connection, while the slave drones provide communication services, and includes a role transfer mechanism in case of master drone failure.

[0018] Furthermore, the role transfer mechanism includes: when the master drone fails, a new master drone is selected based on priority factors such as the battery level of the slave drone, signal strength, latency with other drones, and latency with ground nodes.

[0019] Furthermore, the calculation of the deployment location and quantity also considers the stepped cost function and load balancing constraints to achieve distributed optimization of the drone swarm.

[0020] An air-to-ground collaborative UAV deployment system is used to implement the aforementioned distributed optimized deployment method for air-to-ground collaborative UAVs oriented towards the survivability of optical paths in intelligent computing center optical networks, comprising:

[0021] The optimization deployment module is configured to execute the optimization algorithm to generate a drone deployment plan;

[0022] A drone swarm consists of multiple drone nodes, with each swarm employing a master-slave architecture.

[0023] Ground nodes communicate with the drone swarm;

[0024] And a scheduling controller, configured to activate the UAV cluster based on an optical path failure signal.

[0025] Furthermore, the scheduling controller is further configured to monitor the optical path status and automatically trigger the scheduling and optical path reconstruction process of the UAV cluster when a fault is detected.

[0026] The beneficial effects of this invention are as follows:

[0027] (1) By constructing an integrated air-ground three-dimensional protection system, this invention overcomes the inherent limitations of traditional single ground or air protection. When the ground optical path fails and the backup route fails, the system can quickly schedule the pre-deployed UAV cluster to work in coordination with the remaining ground nodes to form a temporary relay network, thereby achieving seamless migration and dynamic reconstruction of the optical path, which greatly enhances the network's self-healing ability and communication continuity in the event of a fault.

[0028] (2) This invention innovatively proposes an optimization method that combines an integer linear programming model with an efficient heuristic algorithm. The integer linear programming model can solve for the precise solution with the minimum number of drones, optimal deployment location, and lowest total system cost while ensuring service coverage quality, starting from a global perspective. The heuristic algorithm can significantly improve the solution efficiency while basically maintaining the deployment accuracy, providing a feasible path for rapid decision-making in large-scale network scenarios and achieving an effective trade-off between optimization effect and computational cost.

[0029] (3) The heuristic clustering algorithm proposed in this invention can intelligently divide the UAV service domain based on the actual topology of the intelligent computing center's optical network, embodying the core deployment principle of "distributed but not scattered". This distributed cluster partitioning method makes the network structure more flexible and better able to adapt to changes in failure probability and communication needs in different areas.

[0030] (4) This invention designs a master-slave drone architecture and a corresponding role transfer mechanism. This mechanism ensures that even if the master drone in the cluster fails, the system can quickly elect a new master node from the slave drones through an intelligent election algorithm based on multiple factors (such as power, signal strength, latency, etc.), maintaining the uninterrupted coordination function of the cluster, thereby ensuring the continuous and stable operation of the protection system.

[0031] (5) The optimization model of this invention comprehensively considers the basic cost of UAVs, communication latency, and tiered economies of scale, and achieves a reasonable allocation between ground nodes and service UAV clusters, as well as traffic load balancing between different clusters, through constraints. This not only avoids resource waste but also prevents local network congestion, further ensuring communication quality during the protection process.

[0032] Other advantages, objectives, and features of the invention will be set forth in part in the description which follows, and in part will be apparent to those skilled in the art from the following examination, or may be learned from practice of the invention. The objectives and other advantages of the invention can be realized and obtained through the following description. Attached Figure Description

[0033] To make the objectives, technical solutions, and advantages of the present invention clearer, the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings, wherein:

[0034] Figure 1 Diagram of drone network architecture;

[0035] Figure 2 A graph showing the relationship between the cost and quantity of drones;

[0036] Figure 3 A schematic diagram of the partitioning of a distributed drone swarm;

[0037] Figure 4 This is a performance comparison chart between ILP and heuristic algorithms. Detailed Implementation

[0038] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the illustrations provided in the following embodiments are only schematic representations of the basic concept of the present invention. Unless otherwise specified, the following embodiments and features can be combined with each other.

[0039] The accompanying drawings are for illustrative purposes only and are schematic diagrams, not actual pictures. They should not be construed as limiting the invention. To better illustrate the embodiments of the invention, some parts in the drawings may be omitted, enlarged, or reduced, and do not represent the actual product dimensions. It is understandable to those skilled in the art that some well-known structures and their descriptions may be omitted in the drawings.

[0040] In the accompanying drawings of the embodiments of the present invention, the same or similar reference numerals correspond to the same or similar components. In the description of the present invention, it should be understood that if terms such as "upper," "lower," "left," "right," "front," and "rear" indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, they are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, the terms used to describe positional relationships in the drawings are only for illustrative purposes and should not be construed as limiting the present invention. For those skilled in the art, the specific meaning of the above terms can be understood according to the specific circumstances.

[0041] 1. ILP Algorithm Design

[0042] 1.1 Unmanned Aerial Vehicle Network Architecture

[0043] In terms of the drone network architecture, this invention divides the entire drone network into multiple service domains, each consisting of a master drone and several slave drones, such as... Figure 1 As shown, the master drone is responsible for task coordination and backhaul connections within its region, thus achieving a group architecture with physically distributed deployment and logically centralized control. Furthermore, this invention designs a master-slave drone role transfer mechanism. When the master drone malfunctions or is damaged, and slave drones continuously fail to detect the master drone's communication signal, a master drone election is immediately conducted. A new master drone is selected based on the priority of the remaining slave drones (factors include: battery level, signal strength, latency with other drones, latency with ground nodes, etc.).

[0044] 1.2 Issues related to drone deployment

[0045] When deploying drones, the core principle is "distributed yet not scattered." We considered the basic cost of drone deployment, the communication latency between drones and ground nodes, and the backhaul link latency within the drone swarm. Based on these factors, this invention designs an integer linear programming model to meet the following requirements:

[0046] (1) A minimum number of drones are required;

[0047] (2) Divide all drones into multiple drone clusters;

[0048] (3) Deploy the drone swarm on the optimal network nodes;

[0049] (4) Achieve traffic load balancing between UAVs and ground nodes, as well as between different UAV clusters.

[0050] Specifically, this invention minimizes the total system cost while ensuring service coverage quality through design constraints, and also minimizes the number of drones and optimizes their deployment locations. To achieve a "distributed yet not scattered" deployment pattern, this invention also considers factors such as the failure frequency of ground optical paths and drone resource allocation to optimize drone deployment.

[0051] 1.3 ILP Model

[0052] (1) Parameters:

[0053]

[0054] (2) Variables:

[0055]

[0056] (3) Objective function:

[0057]

[0058] (4) Constraints:

[0059]

[0060] ∑ u∈U Y u ≥1 (6)

[0061]

[0062] Constraint (1-1) is the objective function, aiming to minimize the total system cost, including total hardware cost, additional cost of the main UAV, service latency cost of ground nodes, and latency cost between clusters. Constraint (1-2) ensures that each ground node g must be served by one and only one UAV cluster. Constraint (1-3) ensures that the total service capacity of all UAVs deployed on node u can meet the total demand of all ground nodes assigned to this node. Constraint (1-4) collectively defines the binary variable X. u With integer variable M u The logical relationship between them. Constraints (1-5) ensure that only nodes (X) that have deployed the drone swarm can access the drones. u =1) to be selected as the master drone. Constraint (1-6) ensures that there is at least one master drone in the entire network. Constraint (1-7) states that node u and node n both deploy a cluster (i.e., X) if and only if they have deployed a cluster. u =1 and X n When Z = 1), un It must be 1. Constraint (1-8) is to achieve the following: Figure 2 The linearization constraint of the step cost function is shown. Constraint (1-9) defines the range of values ​​for the variables.

[0063] 2 Heuristic Algorithms

[0064] To improve solution efficiency, this invention proposes a heuristic optimization algorithm for intelligent partitioning of UAV networks. It primarily utilizes the K-means clustering algorithm, whose basic idea is to divide all UAV nodes into K clusters, ensuring each UAV belongs to the cluster with the nearest cluster center, and minimizing the density of UAVs within a cluster while maintaining separation between clusters, thus achieving a "dispersed but not scattered" deployment strategy. Specifically, the optical network topology of the intelligent computing center is treated as a graph structure and identified using spectral clustering. An adjacency matrix is ​​constructed based on the communication costs between nodes in the topology. Then, the Laplacian matrix of the graph is calculated, and its dominant eigenvectors are obtained through eigenvalue decomposition. These eigenvectors map the original node coordinates to a new feature space that better reflects their connectivity. In this feature space, nodes that were originally closely connected will cluster in nearby locations. Finally, the K-means algorithm is used to cluster the projections of nodes in this feature space, thereby completing the subdomain partitioning of the entire intelligent computing center optical network and dividing the entire UAV cluster into different service domains. To determine the master UAV of each small UAV cluster, a greedy selection mechanism based on the shortest path strategy is employed. For each candidate node within a subdomain, the algorithm simulates and calculates the total cost required to connect all nodes within the subdomain via the shortest communication path, with that node as the "center". By traversing all candidate nodes and comparing their total communication costs, the algorithm ultimately selects the node that minimizes this total cost as the deployment location for the service resource.

[0065] 3 Performance Analysis

[0066] This invention presents a topology with 11 nodes and 26 edges, as follows: Figure 3 The diagram shows the drone clusters determined by ILP constraints and those obtained through heuristic algorithms. The heuristic algorithm yields drone domains 0-1-2-6 (where the master drone is at node 2, and the number of drones is 2), 8-10 (where the master drone is at node 8, and the number of drones is 1), and 3-4-5-7-9 (where the master drone is at node 5, and the number of drones is 2). The integer linear model yields drone domains 0-1-2-3 (where the master drone is at node 2, and the number of drones is 3), 6-7-9-10 (where the master drone is at node 9, and the number of drones is 3), and 4-5-8 (where the master drone is at node 5, and the number of drones is 2).

[0067] like Figure 4 As shown, the heuristic algorithm has fewer drones deployed than the ILP algorithm, and its cost is lower than that of the ILP algorithm. However, the ILP algorithm takes longer to compute.

[0068] Finally, it should be noted that 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 preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

Claims

1. A distributed optimization deployment method for air-ground cooperative unmanned aerial vehicles (UAVs) for improving the survivability of optical paths in intelligent computing center optical networks, characterized in that: Includes the following steps: Acquire topology information and high-failure-probability area data of the optical network in the intelligent computing center; Based on the topology information and the high-failure-probability area data, the deployment location and number of drone swarms are calculated using an optimization algorithm; Based on the deployment locations and quantities, a drone swarm is pre-deployed to construct a three-dimensional protection system; And when the optical path fails, the UAV cluster is scheduled to provide communication relay services to realize the migration and reconstruction of the optical path.

2. The distributed optimization deployment method for air-ground cooperative UAVs oriented towards optical path survivability in intelligent computing center optical networks according to claim 1, characterized in that: The optimization algorithm includes constructing an integer linear programming (ILP) model, which aims to minimize the total system cost and takes into account unmanned aerial vehicle (UAV) service coverage and capacity constraints.

3. The distributed optimization deployment method for air-ground cooperative UAVs oriented towards optical path survivability in intelligent computing center optical networks according to claim 2, characterized in that: The total system cost of the ILP model includes the basic operating cost of the UAV, the additional cost of the main UAV, the service delay cost of the ground node, and the backhaul link delay cost between UAV clusters. The ILP model also includes constraints that ensure each ground node is served by a UAV cluster, the UAV service capabilities meet the requirements, and the logical relationship of UAV cluster deployment.

4. The distributed optimization deployment method for air-ground cooperative UAVs for optical path survivability in intelligent computing center optical networks according to claim 1, characterized in that: The optimization algorithm includes the use of a heuristic clustering algorithm, which performs clustering based on optical network topology nodes to determine the UAV service domain.

5. The distributed optimization deployment method for air-ground cooperative UAVs for optical path survivability in intelligent computing center optical networks according to claim 4, characterized in that: The heuristic clustering algorithm is the K-means algorithm. The K-means algorithm identifies the node feature space through spectral clustering technology and uses K-means clustering to divide the UAV swarm.

6. The distributed optimization deployment method for air-ground cooperative UAVs oriented towards optical path survivability in intelligent computing center optical networks according to claim 1, characterized in that: The drone swarm adopts a master-slave architecture, in which the master drone is responsible for task coordination and backhaul connection, while the slave drones provide communication services, and include a role transfer mechanism in case of master drone failure.

7. The distributed optimization deployment method for air-ground cooperative UAVs for optical path survivability in intelligent computing center optical networks according to claim 6, characterized in that: The role transfer mechanism includes: when the master drone fails, a new master drone is selected based on priority factors such as the battery level of the slave drone, signal strength, latency with other drones, and latency with ground nodes.

8. The distributed optimization deployment method for air-ground cooperative UAVs for optical path survivability in intelligent computing center optical networks according to claim 1, characterized in that: The calculation of deployment locations and quantities also considers tiered cost functions and load balancing constraints to achieve distributed optimization of the drone swarm.

9. An air-ground cooperative UAV deployment system, used to implement the distributed optimized deployment method for air-ground cooperative UAVs based on the optical path survivability of optical networks in intelligent computing centers according to any one of claims 1 to 8, characterized in that: include: The optimization deployment module is configured to execute the optimization algorithm to generate a drone deployment plan; A drone swarm consists of multiple drone nodes, with each swarm employing a master-slave architecture. Ground nodes communicate with the drone swarm; And a scheduling controller, configured to activate the UAV cluster based on an optical path failure signal.

10. The air-ground cooperative UAV deployment system according to claim 9, characterized in that: The scheduling controller is further configured to monitor the optical path status and automatically trigger the scheduling and optical path reconstruction process of the UAV cluster when a fault is detected.

Images