Remote control method of unmanned aerial vehicle and remote control system of unmanned aerial vehicle

Abstract

The application discloses a kind of unmanned aerial vehicle remote control method and unmanned aerial vehicle remote control system, method includes: receiving unmanned aerial vehicle control intention from control front end, according to unmanned aerial vehicle control intention generation current to be executed task and the service instruction of each type of task business in current to be executed task;According to current to be executed task, the task session between communication terminal in open field is established, generates the service network demand corresponding to each type of task business;According to the task characteristic identification of each type of task business service network demand, according to the network state information of each candidate transmission path, determine the main transmission path from each candidate transmission path and determine the service flow transmission priority of each type of task business;Tunnel encapsulation is carried out to service instruction, task characteristic identification and session identification, according to the transmission strategy corresponding to service flow transmission priority, service message is transmitted to the open field communication terminal of unmanned aerial vehicle through the main transmission path via corresponding public network bearing.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) remote control technology, specifically to a UAV remote control method and a UAV remote control system. Background Technology

[0002] In recent years, the application of drones in inspection, logistics, security, and agricultural protection has expanded rapidly, with tasks becoming more widely distributed, more frequent, and more dispersed in time periods. The traditional paradigm of relying on on-site pilots and short-range links has revealed bottlenecks in terms of operational radius, manpower organization, and centralized management: personnel must be present, cross-location dispatch is costly, and short-range links are easily affected by the environment and terrain, making it difficult to support multiple drones operating concurrently and routine maintenance. Therefore, the industry urgently needs the ability to achieve remote centralized control and maintenance. With the gradual opening of low-altitude airspace resources and the advancement of low-altitude economic policies, high-density operation of multiple entities and multiple drone models in the same low-altitude airspace is becoming the norm, placing higher demands on the safety and controllability of flight activities, trajectory visibility, and accountability traceability. This further amplifies the demand for stable and predictable remote control links.

[0003] In related technologies, three main paths have emerged for the remote control of drones: The first is to deploy fixed-station systems such as airports and drone nests and pre-configure control center strategies. This approach has a high degree of automation, but it involves high initial construction and maintenance costs, high requirements for site and infrastructure, and strong reliance on vendor platforms for overall capabilities. The second approach is based on industry private networks or self-built dedicated lines. This approach can provide relatively stable communication connections within a local area, but its coverage and flexibility are limited, making it difficult to support cross-regional expansion. The third approach uses a single public network (mostly 4G / 5G) as the backhaul link, connecting the control terminal to a vendor or self-built platform to achieve remote operation and data backhaul. However, this approach is usually tied to a single operator and a single platform, and latency performance and the stability of remote control are difficult to guarantee when there are dynamic IPs or cross-network domain migrations. Summary of the Invention

[0004] This invention provides a method and system for remote control of unmanned aerial vehicles (UAVs), aiming to solve the technical problems of high cost, weak scalability, easy interruption of remote control, easy loss of commands, and insufficient stability of remote operation and data transmission in related UAV remote control solutions, such as network domain switching and dynamic changes in IP.

[0005] In a first aspect, the present invention provides a method for remote control of an unmanned aerial vehicle (UAV), applied to a remote control backend for an UAV, the method comprising: Receive drone control intent from the control front end, and generate business instructions for the current task to be executed and various task services in the current task to be executed based on the drone control intent; Establish a task session with the field communication terminal based on the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed. Based on the network requirements of various task services, the task characteristic identifiers corresponding to various task services are determined. Based on the task characteristic identifiers and the network status information of each candidate transmission path, the primary transmission path is determined from each candidate transmission path, and the service flow transmission priority of various task services is determined. The task characteristic identifier is used to identify the task service type of various task services in the current task to be executed. The candidate transmission path is an end-to-end transmission path between the remote control backend and the field communication terminal that can be used to carry the task session. Each candidate transmission path relies on a different public network and each candidate transmission path has a corresponding end-to-end logical tunnel. The service instructions, task characteristic identifiers, and session identifiers of various task services are tunnel-encapsulated, and the service messages of various task services are transmitted to the UAV's field communication terminal through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. The field communication terminal is used to tunnel-decapsulate the service messages and convert them into executable messages of the UAV flight control system, so as to drive the UAV to perform corresponding actions through the UAV flight control system.

[0006] In some embodiments, business instructions are generated based on the UAV control intent, including the current task to be executed and various task operations within the current task to be executed. The drone control intent is parsed to obtain drone flight control information, which includes the drone device identifier to be controlled, mission type, mission action sequence, mission network carrying requirements, and mission constraints. The UAV flight control information is encapsulated into a current task to be executed; The task action sequence in the current task to be executed is divided according to the task business type, and business instructions for various task businesses are generated.

[0007] In some embodiments, a task session is established with the field communication terminal based on the current task to be executed, and the service network requirements corresponding to various task services of the current task to be executed are generated, including: Match the corresponding UAV's field communication terminal by identifying the UAV device to be controlled in the current task to be executed; Establish an end-to-end logical task session between the remote control backend and the field communication terminal, and assign a corresponding session identifier; Based on the network carrying requirements of various tasks and services in the current tasks to be executed, generate the network requirements corresponding to the various tasks and services. The network carrying requirements include at least one of the following: latency tolerance range, jitter requirements, packet loss limit, minimum bandwidth requirements, and stability requirements. The method further includes: Aggregate the feedback information corresponding to the currently pending tasks, and perform consistency verification between the task execution plan of the currently pending tasks and the feedback information; When the consistency check passes, the task to be executed is determined to be completed, and the task session is released and the session state is terminated.

[0008] In some embodiments, based on the task characteristic identifier and the network status information of each candidate transmission path, a primary transmission path is determined from each candidate transmission path, and the service flow transmission priority of various task services is determined, including: Based on the network status information of each candidate transmission path, candidate transmission paths that meet the service network requirements corresponding to various task characteristic identifiers are selected from each candidate transmission path; the network status information includes at least one of the following: latency, jitter, packet loss rate, bandwidth, and stability indicators. The primary transmission path is determined from the candidate transmission paths that meet the business network requirements corresponding to various task characteristic identifiers; Based on the task characteristics of each type of task, the corresponding task service type is identified, and the corresponding service flow transmission priority is assigned to each type of task service. The mission service types include flight control type, status telemetry type, and payload return type. The service flow transmission priority of the flight control type is higher than that of the status telemetry type, and the service flow transmission priority of the status telemetry type is higher than that of the payload return type.

[0009] In some embodiments, the service instructions for various task services, the task characteristic identifier, and the session identifier are tunnel-encapsulated, including: A tunnel encapsulation layer is added to the outer layer of the original message of the service instructions in various task services; Add tunnel metadata to the tunnel encapsulation layer. The tunnel metadata carries the task characteristic identifier, the session identifier, and the tunnel identifier corresponding to the primary transmission path. The tunnel-encapsulated service message is obtained.

[0010] In some embodiments, the method further includes: During task execution, network status information of each candidate transmission path is collected in real time and uploaded to the perception and scheduling layer; Based on the network status information of each candidate transmission path, determine the backup transmission path for the current task to be executed from among the candidate transmission paths; In response to the network information of the primary transmission path corresponding to the current task to be executed meeting the preset path switching conditions, the primary and backup transmission paths of the current task to be executed are switched, or the service packets to be transmitted by the current task to be executed are copied and transmitted through multiple paths, and the field communication terminal processes the service packets that arrive first. During path switching or multi-path copy transmission, the session identifier and task characteristic identifier are quickly rebound to the switching or copying path; The preset path switching conditions include: the network quality of the primary transmission path corresponding to the current task to be executed is lower than a preset threshold, or the network domain-related parameters of the primary transmission path corresponding to the current task to be executed have changed.

[0011] Secondly, the present invention also provides a method for remote control of a drone, using the drone's field communication terminal, the method comprising: Receive a service message for a currently pending task sent by a remote control backend, wherein the service message is generated by the remote control backend and sent to the field communication terminal of the UAV through the UAV remote control method of any embodiment of the first aspect; The service messages are tunneled and decapsulated, and then converted into executable messages for various tasks and services. The executable messages for various mission operations are sent to the UAV flight control system, which then drives the UAV to perform corresponding actions.

[0012] In some embodiments, the method further includes: During mission execution, the system receives feedback information sent by the UAV after performing corresponding actions. The feedback information includes confirmation information, status telemetry information, or collected payload information, and carries associated session identifiers and mission characteristic identifiers. Match the corresponding backhaul transmission path and service flow transmission priority according to the task characteristic identifier and session identifier associated with the backhaul information; According to the transmission strategy corresponding to the matching service flow transmission priority, the backhaul information is transmitted to the remote control backend via the matching backhaul transmission path and the corresponding public network bearer.

[0013] In some embodiments, the returned information includes status telemetry information or payload information collected by the UAV, and the method further includes: Based on the real-time network status information of the matched backhaul transmission path, adaptive adjustment of the transmission parameters is performed on the backhaul information. The transmission parameters include at least one of the following parameters: rate, frequency, bit rate, and frame rate.

[0014] Thirdly, the present invention also provides a remote control system for unmanned aerial vehicles (UAVs). The system is divided into a remote control backend, a bearer network, and an field communication terminal according to the horizontal physical architecture, and the system is divided into a control orchestration layer, a perception scheduling layer, and a fusion network layer according to the vertical logical architecture. On the remote control backend side: The control orchestration layer is used to receive UAV control intentions from the control front end, and generate the current task to be executed and the service instructions of various task services in the current task to be executed according to the UAV control intentions; establish a task session with the field communication terminal according to the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed. The perception and scheduling layer is used to determine the task characteristic identifiers corresponding to various task services based on the service network requirements of various task services, and to determine the primary transmission path and the service flow transmission priority of various task services from the candidate transmission paths based on the task characteristic identifiers and the network status information of each candidate transmission path. The task characteristic identifiers are used to identify the task service types of various task services in the current task to be executed. The candidate transmission paths are end-to-end transmission paths between the remote control backend and the field communication terminal that can be used to carry task sessions. Each candidate transmission path relies on a different public network and each candidate transmission path has a corresponding end-to-end logical tunnel. The converged network layer is used to tunnel encapsulate the service instructions, task characteristic identifiers and session identifiers of various task services, and transmit the service messages of various task services to the UAV's field communication terminal through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. On the field communication end: The converged network layer is used to receive service messages for currently pending tasks sent by the remote control terminal and to perform tunnel decapsulation on the received service messages; The perception and scheduling layer is used to convert the decapsulated business messages into executable messages for various tasks and services. The control orchestration layer is used to send executable messages for various mission operations to the UAV flight control system, so that the UAV flight control system can drive the UAV to perform corresponding actions. The UAV remote control method according to embodiments of the present invention adopts a multi-public network candidate transmission path and independent logical tunnel architecture, reducing dependence on a single network, mitigating the impact of dynamic IP and cross-network domains, effectively preventing communication session interruption and control command loss, and adapting to UAV remote control in complex public network environments. By distinguishing service types through task characteristic identifiers and combining them with hierarchical transmission priorities, priority is given to ensuring the transmission of core flight control commands, which helps avoid non-critical services crowding out resources, ensuring UAV flight safety and control response speed from the transmission level. By transmitting service packets through end-to-end logical tunnels, the heterogeneous differences between different public networks and operators are shielded, breaking through the regional limitations of private networks and dedicated lines, supporting cross-regional remote operation of UAVs, and reducing dependence on sites and infrastructure. The entire process is coordinated by task sessions, dynamically selecting transmission paths and allocating bandwidth based on business network requirements, resulting in more rational network resource utilization. Simultaneously, integrated management of tasks, sessions, and transmission is achieved, creating a closed-loop process that facilitates anomaly tracing and maintenance management. A unified tunnel encapsulation format adapts to various public networks, ensuring strong compatibility. Attached Figure Description

[0015] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0016] Figure 1 A schematic diagram of the vertical logic architecture of a remote control system for unmanned aerial vehicles provided in an embodiment of the present invention; Figure 2 A schematic diagram of the horizontal physical architecture of a remote control system for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the workflow of the UAV remote control system in an embodiment of the present invention; Figure 4 A flowchart illustrating a remote control method for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention; Figure 5 This is a flowchart illustrating another method for remote control of a drone provided in an embodiment of the present invention. Detailed Implementation

[0017] 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.

[0018] It is understood that before using the technical solutions disclosed in the various embodiments of the present invention, users should be informed of the types, scope of use, and usage scenarios of the personal information involved in the present invention and their authorization should be obtained in accordance with relevant laws and regulations through appropriate means.

[0019] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0020] In related technologies, three main paths have emerged for remote control of drones: The first is to deploy fixed-station systems such as airports and drone nests and pre-configure control center strategies. This approach has a high degree of automation, but it involves high initial construction and maintenance costs, high requirements for site and infrastructure, and strong reliance on vendor platforms for overall capabilities. The second approach is based on industry private networks or self-built dedicated lines. This approach can provide relatively stable communication connections within a local area, but its coverage and flexibility are limited, making it difficult to support cross-regional expansion. The third approach uses a single public network (mostly 4G / 5G) as the backhaul link, connecting the control terminal to a vendor or self-built platform to achieve remote operation and data backhaul. However, this approach is usually tied to a single operator and a single platform, and latency performance and remote control stability are difficult to guarantee when there are dynamic IPs or cross-network domain migrations.

[0021] To address these issues, this invention provides a method and system for remote control of unmanned aerial vehicles (UAVs), aiming to effectively solve the technical problems of high cost, weak scalability, easy interruption of remote control, easy loss of commands, and insufficient stability of remote operation and data transmission in related UAV remote control solutions, especially when switching network domains or dynamically changing IP addresses.

[0022] Figure 1 This is a schematic diagram of the vertical logic architecture of a drone remote control system provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the horizontal physical architecture of a drone remote control system provided in an embodiment of the present invention, as shown below. Figure 1 and Figure 2As shown, the UAV remote control system runs on public network infrastructure. The system is divided into a control orchestration layer 101, a perception and scheduling layer 102, and a converged network layer 103 according to the vertical logical architecture. The system is also divided into a remote control backend 201, a bearer network 202, and a UAV field communication terminal 203 according to the horizontal physical architecture.

[0023] Specifically, on the remote control backend 201 side: The control orchestration layer 101 is used to receive UAV control intentions from the control front end, and generate the current task to be executed and the service instructions of various task services in the current task to be executed according to the UAV control intentions; establish a task session with the field communication terminal according to the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed. The perception and scheduling layer 102 is used to determine the task characteristic identifiers corresponding to various task services based on the service network requirements of various task services, and to determine the primary transmission path and the service flow transmission priority of various task services from the candidate transmission paths based on the task characteristic identifiers and the network status information of each candidate transmission path. Among them, the task characteristic identifier is used to identify the task service type of various task services in the current task to be executed, and the candidate transmission path is an end-to-end transmission path between the remote control backend 201 and the field communication terminal 203 that can be used to carry the task session. Each candidate transmission path relies on different public networks for carrying, and each candidate transmission path has a corresponding end-to-end logical tunnel. The converged network layer 103 is used to tunnel encapsulate the service instructions, task characteristic identifiers and session identifiers of various task services, and transmit the service messages of various task services to the UAV's field communication terminal 203 through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. On the 203 side of the field communication terminal: The converged network layer 103 is used to receive service messages of the currently pending task sent by the remote control terminal, and to perform tunnel decapsulation on the received service messages; The perception scheduling layer 102 is used to convert the decapsulated service messages into executable messages for various task services. The control orchestration layer 101 is used to send executable messages of various mission services to the UAV flight control system, so as to drive the UAV to perform corresponding actions through the UAV flight control system.

[0024] like Figure 1 As shown, the bearer network 102 is a public network bearer network, which may include, but is not limited to: 4G, 5G, CF-RAN, LEO, low-altitude ad hoc network, IPv6, IPv4, RSU, etc.

[0025] In embodiments of the present invention, such as Figure 1 As shown, the control orchestration layer 101 receives the UAV control intent and task request from the control front end, completes the conversion of control intent to executable task objects, task characteristic classification and other task orchestration processing, permission domain and task session meta information orchestration processing, and delivers the task's business network requirements and session identifier to the perception scheduling layer 102 to trigger policy generation and adjustment. The control front end can be an existing business platform, a monitoring platform or a dedicated control terminal, and can provide a front-end interface as an operation and observation entry point.

[0026] In some embodiments, on the remote control backend 201 side, at the control orchestration layer 101, business instructions for the current task to be executed and various task services in the current task to be executed are generated according to the UAV control intent. This includes: parsing the UAV control intent, obtaining UAV flight control information, which includes the UAV device identifier to be controlled, task type, task action sequence, network bearer requirements of the task, and task constraints; encapsulating the UAV flight control information into the current task to be executed; and dividing the task action sequence in the current task to be executed according to the task service type to generate business instructions for various task services.

[0027] Among them, the control orchestration layer 101 performs normalization processing after parsing the control intent, and abstracts the sequence of task actions of the control intent into a unified control semantic, such as: take-off, climb, move, hover observation, return to base, etc.

[0028] In some embodiments, at the control orchestration layer 101, a task session is established between the remote control backend and the field communication terminal based on the current task to be executed, and service network requirements corresponding to various task services of the current task to be executed are generated, including: matching the field communication terminal of the corresponding UAV through the UAV device identifier to be controlled in the current task to be executed; establishing an end-to-end logical task session between the remote control backend and the field communication terminal, and assigning a corresponding session identifier; generating service network requirements for various task services based on the network bearer requirements of various task services in the current task to be executed, wherein the network bearer requirements include at least one of the following: latency tolerance range, jitter requirements, packet loss limit, minimum bandwidth requirements, and stability requirements.

[0029] The task session is decoupled from the underlying IP address and port, so that the subsequent service message transmission is not restricted by the dynamic public network address and port status of the external field.

[0030] In this embodiment of the invention, the perception scheduling layer 102 relies on the real-time monitoring data of each parallel link in the lower layer to comprehensively evaluate the network status information (including indicators such as end-to-end latency, jitter, packet loss rate and available bandwidth) of each candidate transmission path, and maintains a set of candidate transmission paths for multi-operator parallel access.

[0031] Based on the service network requirements issued by the control orchestration layer 101, the perception scheduling layer 102 generates corresponding task characteristic identifiers for various task services of each task. The identifiers are used to identify the task service type of the task, so as to classify different task services of the task, such as flight control flow with low latency and high reliability, load flow with large bandwidth and normal state telemetry flow, etc.

[0032] During task scheduling, the perception scheduling layer 102 combines task characteristic identifiers with network status information of each candidate transmission path to jointly decide on the selection of the primary transmission path for the task and the allocation of service flow transmission priority.

[0033] In some embodiments, at the perception scheduling layer 102, based on the task characteristic identifier and the network status information of each current candidate transmission path, a primary transmission path is determined from each candidate transmission path, and the service flow transmission priority of each type of task service is determined. This includes: selecting candidate transmission paths from each candidate transmission path that meet the service network requirements corresponding to each type of task characteristic identifier based on the network status information of each current candidate transmission path; the network status information includes at least one of the following: latency, jitter, packet loss rate, bandwidth, and stability indicators; determining the primary transmission path from the candidate transmission paths that meet the service network requirements corresponding to each type of task characteristic identifier; and assigning corresponding service flow transmission priorities to each type of task service according to the task service type corresponding to each type of task characteristic identifier; wherein, the task service type includes flight control type, status telemetry type, and payload return type, and the service flow transmission priority of flight control type is higher than that of status telemetry type, and the service flow transmission priority of status telemetry type is higher than that of payload return type.

[0034] Specifically, the perception and scheduling layer 102 comprehensively evaluates the network status information (including indicators such as end-to-end latency, jitter, packet loss rate, and available bandwidth) of each candidate transmission path, assesses the network quality (health) of each candidate transmission path, and selects the candidate transmission path with the best network quality from the candidate transmission paths that meet the business network requirements corresponding to various task characteristic identifiers as the primary transmission path for the current task to be executed.

[0035] In some embodiments, the perceptual scheduling layer 102 pre-configures corresponding transmission strategies for the transmission priority of service flows corresponding to different task service types. For example, it gives the highest priority to critical service flows marked as low latency and high reliability and prioritizes keeping their paths open. For high bandwidth or low priority service flows, it can implement transmission strategies such as rate limiting, reducing bit rate, frame rate, or bounded degradation.

[0036] During task execution, the perception and scheduling layer 102 can continuously collect and analyze network status information of each candidate transmission path, and pre-construct backup transmission paths for the task. This allows for rapid switching between primary and backup transmission paths when the network quality of the primary transmission path deteriorates or network information changes, effectively maintaining the continuity of critical task service flows. Changes in network information refer to changes in IP addresses and network domains. These changes mainly originate from scenarios such as base station switching, NAT / CGN mapping changes, public network address changes, cross-carrier switching, and cellular / WiFi access switching.

[0037] During network fluctuations or path switching, the perception and scheduling layer 102, based on the aforementioned state perception and task characteristic identification scheduling mechanism, can ensure the end-to-end stability and continuity of the transmission path of critical service flows, while dynamically adjusting the resource usage of other services to maintain system performance balance.

[0038] In this embodiment of the invention, the converged network layer 103 is deployed on top of existing public networks, industry private networks, or satellite-ground converged bearer networks, providing end-to-end tunnels for task service packets, completing inbound tunnel encapsulation, tunnel identifier writing, and outbound restoration, making underlying address changes and network domain switching transparent to the upper layer. Inbound refers to entering the bearer network or end-to-end tunnel, and outbound refers to leaving the bearer network or end-to-end tunnel.

[0039] The task characteristic identifier from the perception scheduling layer 102 is written into the tunnel metadata during domain encapsulation. It remains unchanged within the end-to-end tunnel and does not alter the message body of the original service flow instructions. The backhaul confirmation, telemetry information, and payload information are carried back along the primary transmission path bound to the current task session, carrying this identifier, for task-level consistency aggregation at the control orchestration layer 101. The control orchestration layer 101 can maintain the backhaul transmission path relationship corresponding to the current task session through session management, and synchronously update its backhaul binding relationship when switching paths.

[0040] During task execution, the converged network layer 103 can map different service flows to different bearer categories and queue priorities based on task characteristic identifiers, and execute priority forwarding, shaping, or rate limiting according to the transmission policies issued by the upper layer. When policies change, transmission parameters are adjusted synchronously. Policies include transmission path selection, path switching, priority, and bounded degradation, while transmission parameters include egress selection, tunnel endpoints, priority, shaping, and rate limiting. During operation, the converged network layer 103 collects network status information such as latency, jitter, packet loss, and available bandwidth at ingress nodes, egress nodes, or transit measurement points on each transmission path, and reports this information to the perception and scheduling layer 102 after associating it with tunnel identifiers. Through the above encapsulation, transparent bearer, and status reporting mechanisms, the converged network layer 103 shields the changes in underlying addresses and network domains on the one hand, and provides the perception and scheduling layer 102 with consumable network observations on the other, laying the foundation for stable end-to-end transmission path bearing. Among them, transit observation points can be located at the ingress, relay, or egress positions of the transmission path, and are used to collect status information such as latency, jitter, packet loss, and available bandwidth.

[0041] The converged network layer 103 provides bearer-independent encapsulation, decapsulation, and tunnel identifier management over the public network. It uniformly encapsulates control, telemetry, and payload flows, maintaining transparency for dynamic IP, NAT / CGN, and cross-network domain / cross-carrier mobility handovers. It also supports service-type categorized bearer management, priority forwarding, and multi-path scheduling. Within the tunnel, categorized bearer management and priority governance are implemented based on category, and an adaptation identifier is bound upon entry into the domain. The tunnel is preferably implemented using a programmable data plane, but can also be implemented using an equivalent software repeater or a general-purpose server. Encapsulation format, field length, queue size, and shaping strategy are not limited.

[0042] In some embodiments, in the converged network layer 103, the service instructions, task characteristic identifiers and session identifiers of various task services are tunnel-encapsulated, including: adding a tunnel encapsulation layer to the outer layer of the original message of the service instructions of various task services; adding tunnel metadata to the tunnel encapsulation layer, wherein the tunnel metadata carries the task characteristic identifier, session identifier and the tunnel identifier corresponding to the primary transmission path; and obtaining the tunnel-encapsulated service message.

[0043] On the field communication terminal 203 side, the converged network layer 103 receives the service message of the current task to be executed sent by the remote control backend; the perception and scheduling layer 102 performs tunnel decapsulation on the service message and converts it into executable messages of various task services; the control and orchestration layer 101 sends the executable messages of various task services to the UAV flight control system of the UAV, so as to drive the UAV to perform corresponding actions through the UAV flight control system.

[0044] During mission execution, the converged network layer 103 receives feedback information sent by the UAV through the field network after performing corresponding actions. The feedback information includes service execution results, such as confirmation information for flight control services, status telemetry information for status telemetry services, or payload information for payload flow feedback services. The feedback information carries associated session identifiers and task characteristic identifiers. The perception and scheduling layer 102 matches the corresponding feedback transmission path and service flow transmission priority according to the task characteristic identifier and session identifier associated with the feedback information. The converged network layer 103 transmits the backhaul information to the remote control backend via the corresponding public network bearer through the matched backhaul transmission path according to the transmission strategy corresponding to the matching service flow transmission priority.

[0045] For example, for the confirmation information of the flight control stream transmitted back, the transmission strategy for the corresponding service stream transmission priority can be to transmit it with the highest priority and keep its transmission path unobstructed; for the status telemetry information of the status telemetry stream transmitted back and the payload information of the payload stream transmitted back, the transmission strategy for the corresponding service stream transmission priority can be to implement rate limiting, reduce bit rate, frame rate, or bounded degradation, etc.

[0046] In some embodiments, the backhaul information includes status telemetry information or payload information collected by the UAV. During the backhaul information transmission process, the fusion network layer 103 adaptively adjusts the transmission parameters of the backhaul information based on the real-time network status information of the matched backhaul transmission path. The transmission parameters include at least one of the following: rate, frequency, bit rate, and frame rate. For example, when the network quality of the backhaul transmission path deteriorates, the transmission rate, frequency, bit rate, or frame rate of the backhaul information is reduced, and the transmission parameters automatically recover when the network recovers.

[0047] In some embodiments, on the remote control terminal 201 side, the converged network layer 103 collects network status information of each candidate transmission path in real time during task execution and uploads it to the perception and scheduling layer 102; the perception and scheduling layer 102 determines the backup transmission path for the current task to be executed from each candidate transmission path based on the network status information of each candidate transmission path; in response to the network information of the primary transmission path corresponding to the current task to be executed meeting the preset path switching conditions, the perception and scheduling layer 102 performs primary and backup transmission path switching for the current task to be executed, or performs multi-path copy transmission of the service packets to be transmitted by the current task to be executed, and the field communication terminal processes the service packets that arrive first; during path switching or multi-path copy transmission, the session identifier, task characteristic identifier and the switched or copied path are quickly rebound; wherein, the preset path switching conditions include: the network quality of the primary transmission path corresponding to the current task to be executed is lower than a preset threshold, or the network domain related parameters of the primary transmission path corresponding to the current task to be executed have changed.

[0048] In some embodiments, the perceptual scheduling layer 102 can continuously collect and analyze the network status information of each candidate transmission path, comprehensively evaluate the network quality or network health of each candidate transmission path, and select a candidate transmission path other than the primary transmission path from each candidate transmission path as a backup transmission path.

[0049] In some embodiments, the control orchestration layer 101 is also used to aggregate the feedback information corresponding to the task to be executed, perform consistency verification between the task execution plan of the current task to be executed and the feedback information, and when the consistency verification passes, determine that the current task to be executed is completed, and release the task session and terminate the session state.

[0050] Figure 3 This is a schematic diagram of the workflow of the drone remote control system in an embodiment of the present invention, as shown below. Figure 3 As shown, the control intent enters the control orchestration layer 101 through the control entry point and forms a specific task. Then, a task session is established and the corresponding service network requirements are generated. Both are then exposed to the perception and scheduling layer 102.

[0051] The perception and scheduling layer 102 collects the current access environment, maintains the system candidate transmission path set, and after receiving the upper layer session request, performs service characteristic mapping on the corresponding business network requirements to form a task characteristic identifier. Then, it combines the network status information measured at the end or in transit by the converged network layer 103 to match the scheduling strategy and selects a route from the candidate transmission path set to send the request. The converged network layer 103 performs tunnel encapsulation on the business instructions, confirmations, telemetry and other control-related messages of the task, transparently carrying the task characteristic identifier in the end-to-end tunnel, and transmitting it to the field via the public network and other bearer networks. After the tunnel decapsulation is completed in the field, the message is sent to the control semantic parsing and task decomposition module on the field side. Based on the unified control semantics, the module generates control messages that can be executed by the terminal, which are then sent to the UAV flight control system via the field network and cluster networking, driving the UAV to perform corresponding actions and form flight attitude and task response behavior.

[0052] During task execution, the perception scheduling layer 102 adjusts the paths and priorities in a timely manner according to the strategy on the existing candidate transmission path set based on the network status update from the converged network layer. When necessary, it calls the path switching and bounded degradation mechanism to ensure that the low-latency, high-reliability control path remains continuously available during network fluctuations.

[0053] During mission execution, the confirmation and telemetry data generated by the field flight control are transmitted back along the predetermined tunnel direction, and the payload video is transmitted back adaptively based on the real-time network status, with the bit rate / frame rate adjusted accordingly.

[0054] In scenarios such as dynamic IP, NAT / CGN, and cross-network domain / carrier handover, the above control and feedback form a unified closed loop between the three layers, keeping the session uninterrupted until the task is completed and the closed loop is verified and converged at the control orchestration layer 101.

[0055] This invention, through the coordination of various layers of the UAV remote control system, can achieve public network penetration and maintain session continuity. It can perform path selection and switching based on control intent and network status, overcoming problems such as limited single network coverage, unstable sessions caused by dynamic addresses, and lack of hierarchical business governance. It provides stable and scalable cross-public network remote control for low-altitude application scenarios such as inspection, plant protection, and logistics.

[0056] According to an embodiment of the present invention, a method for remote control of a drone is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0057] This invention provides a method for remote control of a drone, applied to a remote control backend for a drone. Figure 4 This is a flowchart illustrating a remote control method for unmanned aerial vehicles (UAVs) provided in an embodiment of the present invention, as shown below. Figure 4 As shown, the remote control method for this drone includes the following steps: Step S401: Receive the UAV control intent from the control front end, and generate business instructions for the current task to be executed and various tasks and services in the current task to be executed based on the UAV control intent.

[0058] The control front end can be an existing business platform, a regulatory platform, or a dedicated control terminal. It can include a web-based control and visualization front end interface or an equivalent native client / automation interface for task orchestration, flight control, telemetry video display, log retention, and remote demonstration. This front end interface only serves the human-machine experience and operational efficiency and does not impose any restrictions on the underlying carrier, protocol, or implementation form. The above purposes can be fully achieved even without enabling the web interface.

[0059] The control intent of a drone refers to the control requirements of the drone, including but not limited to control requirements such as flight operation, operation instructions, parameter configuration, and status management.

[0060] In some embodiments, the process of generating business instructions for the current task to be executed and various task services within the current task to be executed based on the drone control intent includes: parsing the drone control intent and obtaining drone flight control information, which includes the drone device identifier to be controlled, task type, task action sequence, network bearer requirements for the task, and task constraints; encapsulating the drone flight control information into the current task to be executed; and dividing the task action sequence within the current task to be executed according to the task service type to generate business instructions for various task services.

[0061] The task type is a category of tasks based on the operational behaviors and flight actions that the UAV needs to perform. It is used to distinguish between the purpose of the operation and the form of execution, such as flight control, detection and data collection, operation execution, and status monitoring.

[0062] A mission sequence is a series of flight, operation, and control actions arranged in a preset order, forming a complete chain of execution steps.

[0063] The network requirements for a task are the various indicators and adaptation requirements imposed on the network during task operation, such as latency, jitter, packet loss, bandwidth, and stability.

[0064] Task constraints are the rules, threshold values, and boundary restrictions that limit the entire process of UAV mission execution. They are used to regulate operational behavior, ensure operational safety, and ensure compliance. Examples of constraints include airspace boundaries, flight altitude, speed limits, operation duration, execution sequence, action specifications, equipment operating conditions, environmental weather, safety and risk avoidance rules, resource allocation, and degradation constraints.

[0065] Step S402: Establish a task session with the field communication terminal based on the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed.

[0066] In some embodiments, a task session is established between the remote control backend and the field communication terminal based on the current task to be executed, and service network requirements corresponding to various task services of the current task to be executed are generated, including: matching the field communication terminal of the corresponding UAV through the UAV device identifier to be controlled in the current task to be executed; establishing an end-to-end logical task session between the remote control backend and the field communication terminal, and assigning a corresponding session identifier; generating service network requirements corresponding to various task services based on the network bearer requirements of various task services in the current task to be executed, wherein the network bearer requirements include at least one of the following: latency tolerance range, jitter requirements, packet loss limit, minimum bandwidth requirements, and stability requirements. Step S403: Determine the task characteristic identifiers corresponding to various task services based on the service network requirements of various task services, and determine the primary transmission path from each candidate transmission path and the service flow transmission priority of various task services based on the task characteristic identifiers and the network status information of each candidate transmission path.

[0067] Among them, the task characteristic identifier is used to identify the task service type of various task services in the current task to be executed. The candidate transmission path is an end-to-end transmission path between the remote control backend and the field communication terminal that can be used to carry the task session. Each candidate transmission path relies on different public networks for carrying, and each candidate transmission path has a corresponding end-to-end logical tunnel.

[0068] In some embodiments, based on the task characteristic identifier and the network status information of each current candidate transmission path, a primary transmission path is determined from each candidate transmission path, and the service flow transmission priority of each type of task service is determined. This includes: selecting candidate transmission paths from each candidate transmission path that meet the service network requirements corresponding to each type of task characteristic identifier based on the network status information of each current candidate transmission path; the network status information includes at least one of the following: latency, jitter, packet loss rate, bandwidth, and stability indicators; determining the primary transmission path from the candidate transmission paths that meet the service network requirements corresponding to each type of task characteristic identifier; and assigning corresponding service flow transmission priorities to each type of task service according to the task service type corresponding to each type of task characteristic identifier; wherein, the task service type includes flight control type, status telemetry type, and payload return type, and the service flow transmission priority of flight control type is higher than that of status telemetry type, and the service flow transmission priority of status telemetry type is higher than that of payload return type.

[0069] Step S404: The service instructions, task characteristic identifiers and session identifiers of various task services are tunnel-encapsulated, and the service messages of various task services are transmitted to the UAV's field communication terminal through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. The field communication terminal is used to tunnel-decapsulate the service messages and convert them into executable messages of the UAV flight control system, and drive the UAV to perform corresponding actions through the UAV flight control system.

[0070] The field communication terminal can be deployed on the drone or at the drone's operating site. It communicates with the drone's flight control system through a field network. The field network is the local access network between the field communication terminal and the drone's flight control system, and can be WiFi, Ethernet, serial port, short-range wireless, or other field links.

[0071] After receiving the service message of the current task, the field communication terminal performs tunnel decapsulation of the service message to obtain the service instructions, extracts the control semantics, generates an executable message that the UAV can execute according to the unified control semantics, and sends it to the UAV flight control system through the field network and cluster networking, driving the UAV to perform corresponding actions and form flight attitude and task response behavior.

[0072] The UAV remote control method according to embodiments of the present invention adopts a multi-public network candidate transmission path and independent logical tunnel architecture, reducing dependence on a single network, mitigating the impact of dynamic IP and cross-network domains, effectively preventing communication session interruption and control command loss, and adapting to UAV remote control in complex public network environments. By distinguishing service types through task characteristic identifiers and combining them with hierarchical transmission priorities, priority is given to ensuring the transmission of core flight control commands, which helps avoid non-critical services crowding out resources, ensuring UAV flight safety and control response speed from the transmission level. By transmitting service packets through end-to-end logical tunnels, the heterogeneous differences between different public networks and operators are shielded, breaking through the regional limitations of private networks and dedicated lines, supporting cross-regional remote operation of UAVs, and reducing dependence on sites and infrastructure. The entire process is coordinated by task sessions, dynamically selecting transmission paths and allocating bandwidth based on business network requirements, resulting in more rational network resource utilization. Simultaneously, integrated management of tasks, sessions, and transmission is achieved, creating a closed-loop process that facilitates anomaly tracing and maintenance management. A unified tunnel encapsulation format adapts to various public networks, ensuring strong compatibility.

[0073] In some embodiments, the service instructions, task characteristic identifiers, and session identifiers of various task services are tunnel-encapsulated, including: adding a tunnel encapsulation layer to the outer layer of the original message of the service instructions of various task services; adding tunnel metadata to the tunnel encapsulation layer, wherein the tunnel metadata carries the task characteristic identifier, session identifier, and tunnel identifier corresponding to the primary transmission path; and obtaining a tunnel-encapsulated service message.

[0074] In some embodiments, the UAV remote control method further includes: during task execution, collecting network status information of each candidate transmission path in real time; determining a backup transmission path for the current task to be executed from each candidate transmission path based on the network status information of each candidate transmission path; in response to the network information of the primary transmission path corresponding to the current task to be executed meeting a preset path switching condition, performing a primary / backup transmission path switch for the current task to be executed, or performing multi-path copy transmission of the service message to be transmitted by the current task to be executed, with the field communication terminal processing the service message that arrives first; and quickly rebinding the session identifier, task characteristic identifier, and the switched or copied path during path switching or multi-path copy transmission; wherein the preset path switching condition includes: the network quality of the primary transmission path corresponding to the current task to be executed is lower than a preset threshold, or the network domain-related parameters of the primary transmission path corresponding to the current task to be executed have changed.

[0075] In some embodiments, the drone remote control method further includes: aggregating the feedback information corresponding to the task to be executed, and performing a consistency check between the task execution plan of the current task to be executed and the feedback information; when the consistency check passes, determining that the current task to be executed is completed and releasing the task session and terminating the session state.

[0076] In some embodiments, the system supports parallel access from multiple operators and multi-path overlay, and can be incrementally deployed and smoothly expanded according to region or mission objectives; the remote control backend can be centrally deployed in the indoor data center, or it can be deployed to edge nodes or collaborate with the public cloud as needed; the transmission domain of the converged network layer can be located in any form such as a programmable data plane, software repeater, or general server; the field communication terminal can be integrated with the existing ground station or independently mounted on the UAV; the location of network status awareness and policy execution in the converged network layer can be flexibly configured at the entry side, relay node, or exit side of each transmission path, and can be aggregated by domain.

[0077] In some embodiments, the session anchoring method, keep-alive period, rebinding process and state synchronization means of the task session are not limited and can be implemented collaboratively by the application layer, transport layer or data plane.

[0078] In some application scenarios, such as routine inspections of roads, parks, or power lines, drones are typically deployed over a wide area, relying on public network infrastructure for access links, without fixed dedicated lines or airport system support. During operation, drones continuously generate status information and transmit it back in real time. The remote control backend then issues action commands, temporarily adjusts flight paths, or triggers a return-to-home mechanism based on this status feedback. The drone remote control method of this invention enables stable command issuance, session persistence, link penetration, and task closure in this typical application environment characterized by "crossing public networks, dynamic addresses, and lack of dedicated line support," allowing drones to operate under long-term, predictable, and controlled conditions within the public network.

[0079] Specifically, before the inspection task begins, the operator selects or inputs control intentions such as flight routes, locations, or observation tasks through the control front end at the remote control back end. The control orchestration layer normalizes these control intentions, abstracting the actions at the human-machine interaction level into unified control semantics, such as: takeoff, climb, move, hover observation, return, generate specific tasks, and establish a task session decoupled from the underlying IP / port system, so that subsequent control messages are not restricted by the dynamic public network address and port status of the field.

[0080] During mission execution, the UAV may be in any public network environment, such as cellular networks, WiFi, or shared hotspots. Its public network address may change dynamically, and the link quality may become unstable due to location changes. The converged network layer uniformly encapsulates the control flow, telemetry, and payload flow, making the underlying NAT / CGN and mobility handover transparent to the upper layer. Regardless of the fluctuation conditions of the external network, by setting high priority for critical control flows and limiting and adaptively adjusting non-critical telemetry or payload flows, it ensures that the UAV's attitude control is not affected by external link conditions, maintaining the continuity and consistency of operational response.

[0081] The field communication unit converts the unified control semantics of received service messages into executable messages for the UAV flight control system and drives the UAV to perform these actions. Confirmation information and status telemetry information generated after flight maneuvers are returned along the original path, and payload video is adaptively transmitted back based on network conditions, enabling the command center to monitor the UAV's status in real time. During mission execution, even if the public network link experiences short-term jitter or the public network address changes, the mission session remains valid, and service messages are not discarded or re-established due to address changes; telemetry data is adaptively transmitted back amidst link fluctuations, ensuring the continuous and controllable execution of the mission loop.

[0082] In some application scenarios, drones are tasked with moving across regions, experiencing rapidly changing coverage conditions, or performing missions at the intersection of different operator networks. Examples include long-distance inspections along highways, cross-regional hazard investigations, or mobile patrols on the outskirts of urban areas. Typical characteristics of such tasks include: drones may frequently experience base station handovers, link jitter, short-term weak coverage, or migration between different operator networks during flight; the latency, jitter, and available bandwidth of public network links fluctuate significantly with spatial location. Without a stable cross-network compensation mechanism, the control link is easily affected by jitter, handover, or brief disconnections, leading to delayed, intermittent, or inconsistent execution of control commands. Based on multi-path input, network status awareness, and policy scheduling, predictable adjustments are made to the path selection and priority of the control link under cross-operator and weak coverage conditions, thereby ensuring the continuity and controllability of remote control.

[0083] Before a cross-regional inspection or mobile patrol mission begins, the drone's field communication terminal can simultaneously have multiple public network access options, such as multiple cellular cards from different operators, or a combination of cellular and local WiFi access.

[0084] After a task is started, the converged network layer establishes a corresponding tunnel mapping for each available candidate transmission path and writes a task characteristic identifier, so that the same task has a switchable bearing foundation on multiple candidate transmission paths. The perception and scheduling layer pre-sets the highest priority and basic policies for the control flow according to the task type and service category. From the time the task starts, the system has the ability to observe network quality, complete decision conditions and prepare for path switching on the candidate transmission path set.

[0085] As drones fly along their routes, they may enter the edge of base station coverage, be obstructed by buildings, or cross cell boundaries of different operators, causing a continuous decline in the health of the current path. By continuously sensing changes in latency, jitter, and available bandwidth through in-transit measurements or end-side detection, and combining this with service type identifiers, the system assesses whether the current path is still suitable as the main control path. When it is determined that the current path cannot meet control requirements, the system can perform multi-link copying of the same control flow on multiple public network links. The field communication terminal then reconstructs the instruction sequence according to the first arriving messages, completing the path migration without interrupting the control session. In scenarios where terminal resources are limited or multi-link concurrency is not possible, the system degenerates into performing rapid rebinding of the session identifier with the new path and tunnel on the new path, maintaining the session identifier unchanged from before the path switch. Throughout the process, non-critical services such as payload video and telemetry are subject to quota limits, frequency reduction, or adaptive bitrate according to policies to avoid crowding out the control path within the switching window.

[0086] After the path switch is completed, the original control session and mission characteristic identifiers remain unchanged. The field communication terminal continues to drive flight control execution according to the unified control semantics, and commands remain continuous in semantics and timing. Confirmation information and telemetry data return along the latest valid path, and the payload video automatically adjusts transmission parameters based on the new network status information. Since the control session is not rebuilt before and after the switch, and the control flow is always guaranteed with the highest priority throughout the process, flight actions remain logically consistent. The mission can continue to advance under different public network conditions and multiple switch events, thus supporting the remote control requirements in highly dynamic scenarios.

[0087] In some application scenarios, such as multiple drones collaboratively performing tasks (e.g., a swarm of drones working together within a city to perform tasks like road inspection, environmental monitoring, or agricultural and forestry protection), the control unit needs to manage multiple task control intentions and control links simultaneously. Under normal public network conditions, each drone's control link must remain independent and stable. Multiple drones may share the same public network coverage area and uplink resources; without differentiated scheduling, the load traffic of a single device can easily interfere with the overall control plane. By using unified control orchestration, task characteristic identification, and independently schedulable tunnels, scalability and controllability can be achieved when multiple drones are remotely controlled simultaneously.

[0088] Before a multi-drone collaborative task begins, the remote control backend can set task content for each drone, such as different flight paths, key inspection sections, or observation targets. The control orchestration layer parses the control intentions of each drone separately and performs normalization processing to form independent control semantics and task characteristic identifiers. It also establishes an independent task session for each drone, decoupled from the public network address / port system. In this way, multiple devices maintain clear session boundaries and a basis for business division even in a shared public network environment, laying a unified entry point for subsequent differentiated scheduling.

[0089] When multiple drones operate simultaneously within the same public network coverage area, the traffic from each device will share limited link resources. At the converged network layer, the control, telemetry, and payload flows of each drone are encapsulated separately and classified based on task characteristics: the control flow of each device is given the highest priority in all sessions, the telemetry flow is shaped according to a preset period and quota, and the payload flow is adaptively adjusted in bit rate or frequency based on the network quality and transmission strategy of the overall transmission path. The perception and scheduling layer generates path selection and resource scheduling strategies for different drones based on health information collected independently for each session, ensuring that the control plane does not experience unpredictable jitter or congestion due to the payload traffic of other devices when multiple devices share the public network link.

[0090] On the field side, the field communication terminals of each UAV convert the unified control semantics into flight control executable messages and drive them independently; after the action is completed, the confirmation information and telemetry status are sent back to the remote control backend according to their respective sessions and paths.

[0091] The control orchestration layer aggregates and displays data transmitted from multiple UAVs based on task characteristic identifiers and session identifiers. This allows the command center to simultaneously monitor the progress, attitude, and health status of each UAV in the cluster and adjust task allocation or add scheduling commands accordingly. Even under conditions of multi-UAV concurrency, link fluctuations, or path switching of individual devices, each control session remains independent and controllable, and the overall mission can be continuously advanced in a collaborative manner in a public network environment.

[0092] This invention addresses the rapidly expanding applications of unmanned aerial vehicles (UAVs), eliminating reliance on on-site pilots and short-range links. Without building dedicated lines or binding to a single vendor's platform, it reuses existing public network infrastructure to provide a universal remote control and maintenance system that can be deployed across domains. This system emphasizes bearer independence and portability, maintaining end-to-end reachability and long-term sessions under conditions of high mobility and coverage boundaries, resolving continuity and security issues arising from dynamic addresses and cross-network domain migration. During congestion and handover, it provides deterministic constraints and prioritizes protection for the control plane, achieving measurable isolation and orderly degradation from payload video and telemetry. It upgrades from "pilot control" to "mission-level capability," supporting not only direct real-time UAV control but also route issuance, automatic cruise, and routine batch / cluster scheduling. It also provides a unified management and observation interface, facilitating integration with existing front-end and back-end systems and incremental deployment, reserving a technical foundation for integration with low-altitude airspace management and industry regulatory systems. By achieving the above objectives, this invention overcomes the limitations of related technical solutions in terms of cost, coverage / flexibility, versatility, and feasibility, and forms a remote unmanned aerial vehicle (UAV) control paradigm built on public network infrastructure.

[0093] This invention constructs a bearer-independent and implementation-independent remote drone control and maintenance platform on existing public network infrastructure. It maintains end-to-end reachability and session continuity under dynamic address and cross-network domain conditions. Combining service classification, state awareness, and programmable transport orchestration, it completes path migration and link switching without interruption or with minimal interruption, and implements bounded differentiated adjustments to control, telemetry, and payload flows. This platform is not limited to specific bearer networks or network equipment forms, nor to front-end access methods, and can be integrated with existing or future business platforms and regulatory platforms through open interfaces. The drone remote control and maintenance capabilities based on public network infrastructure can maintain control session continuity and differentiatedly adjust various service flows under dynamic IP and cross-network domain conditions. Without building new leased lines or binding to a single vendor platform or bearer network, it simultaneously achieves comprehensive goals such as portable coverage, vendor-free control, task-level expansion, and network optimization.

[0094] This invention provides a remote control method for unmanned aerial vehicles (UAVs), applied to the UAV's field communication terminal. Figure 5 This is a flowchart illustrating another method for remotely controlling a drone provided in an embodiment of the present invention, as shown below. Figure 5 As shown, the remote control method for this drone includes the following steps: Step S501: Receive the business message of the currently pending task sent by the remote control backend; Among them, the business message is generated by the remote control backend through the UAV remote control method of the above embodiment and sent to the UAV's field communication terminal; Step S502: Decapsulate the service message through the tunnel and convert it into executable messages for various task services; Step S503: Send executable messages for various mission services to the UAV flight control system so that the UAV can be driven to perform corresponding actions through the UAV flight control system.

[0095] In some embodiments, the drone remote control method further includes: During mission execution, the system receives feedback information sent by the UAV after performing corresponding actions. The feedback information includes confirmation information, status telemetry information, or collected payload information. The feedback information carries the associated session identifier and mission characteristic identifier. Match the corresponding backhaul transmission path and service flow transmission priority according to the task characteristic identifier and session identifier associated with the backhaul information; According to the transmission strategy corresponding to the matching service flow transmission priority, the return information is transmitted to the remote control backend through the matching return transmission path via the corresponding public network bearer.

[0096] In some embodiments, the backhaul information includes status telemetry information or payload information collected by the UAV. The UAV remote control method further includes: adaptively adjusting the transmission parameters of the backhaul information according to the real-time network status information of the matched backhaul transmission path. The transmission parameters include at least one of the following parameters: rate, frequency, bit rate, and frame rate.

[0097] For a detailed description of the UAV remote control method applied to the field communication terminal, please refer to the relevant descriptions in the above-mentioned UAV remote control method and UAV remote control system applied to the remote control backend, which will not be repeated here.

[0098] Although embodiments of the invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A method for remote control of an unmanned aerial vehicle (UAV), characterized in that, The method, applied to a remote control backend for unmanned aerial vehicles, includes: Receive drone control intent from the control front end, and generate business instructions for the current task to be executed and various task services in the current task to be executed based on the drone control intent; Establish a task session with the field communication terminal based on the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed. Based on the network requirements of various task services, the task characteristic identifiers corresponding to various task services are determined. Based on the task characteristic identifiers and the network status information of each candidate transmission path, the primary transmission path is determined from each candidate transmission path, and the service flow transmission priority of various task services is determined. The task characteristic identifier is used to identify the task service type of various task services in the current task to be executed. The candidate transmission path is an end-to-end transmission path between the remote control backend and the field communication terminal that can be used to carry the task session. Each candidate transmission path relies on a different public network and each candidate transmission path has a corresponding end-to-end logical tunnel. The service instructions, task characteristic identifiers, and session identifiers of various task services are tunnel-encapsulated, and the service messages of various task services are transmitted to the UAV's field communication terminal through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. The field communication terminal is used to tunnel-decapsulate the service messages and convert them into executable messages of the UAV flight control system, so as to drive the UAV to perform corresponding actions through the UAV flight control system.

2. The method according to claim 1, characterized in that, Based on the UAV control intent, business instructions are generated for the current task to be executed and various task operations within the current task to be executed, including: The drone control intent is parsed to obtain drone flight control information, which includes the drone device identifier to be controlled, mission type, mission action sequence, mission network carrying requirements, and mission constraints. The UAV flight control information is encapsulated into a current task to be executed; The task action sequence in the current task to be executed is divided according to the task business type, and business instructions for various task businesses are generated.

3. The method according to claim 1, characterized in that, Establish a task session with the external communication terminal based on the current task to be executed, and generate the service network requirements corresponding to various tasks and services of the current task to be executed, including: Match the corresponding UAV's field communication terminal by identifying the UAV device to be controlled in the current task to be executed; Establish an end-to-end logical task session between the remote control backend and the field communication terminal, and assign a corresponding session identifier; Based on the network bearer requirements of various tasks and services in the current tasks to be executed, generate the network requirements corresponding to the various tasks and services. The network bearer requirements include at least one of the following: latency tolerance range, jitter requirements, packet loss limit, minimum bandwidth requirements, and stability requirements. The method further includes: Aggregate the feedback information corresponding to the currently pending tasks, and perform consistency verification between the task execution plan of the currently pending tasks and the feedback information; When the consistency check passes, the task to be executed is determined to be completed, and the task session is released and the session state is terminated.

4. The method according to claim 1, characterized in that, Based on the task characteristic identifier and the network status information of each candidate transmission path, the primary transmission path is determined from each candidate transmission path, and the service flow transmission priority for various task services is determined, including: Based on the network status information of each candidate transmission path, candidate transmission paths that meet the service network requirements corresponding to various task characteristic identifiers are selected from each candidate transmission path; the network status information includes at least one of the following: latency, jitter, packet loss rate, bandwidth, and stability indicators. The primary transmission path is determined from the candidate transmission paths that meet the business network requirements corresponding to various task characteristic identifiers; Based on the task characteristics of each type of task, the corresponding task service type is identified, and the corresponding service flow transmission priority is assigned to each type of task service. The mission service types include flight control type, status telemetry type, and payload return type. The service flow transmission priority of the flight control type is higher than that of the status telemetry type, and the service flow transmission priority of the status telemetry type is higher than that of the payload return type.

5. The method according to claim 1, characterized in that, The business instructions for various tasks, the task characteristic identifiers, and the session identifiers are tunnel-encapsulated, including: A tunnel encapsulation layer is added to the outer layer of the original message of the service instructions in various task services; Add tunnel metadata to the tunnel encapsulation layer. The tunnel metadata carries the task characteristic identifier, the session identifier, and the tunnel identifier corresponding to the primary transmission path. The service message is obtained after tunnel encapsulation.

6. The method according to claim 1, characterized in that, The method further includes: During task execution, network status information of each candidate transmission path is collected in real time and uploaded to the perception and scheduling layer; Based on the network status information of each candidate transmission path, determine the backup transmission path for the current task to be executed from among the candidate transmission paths; In response to the network information of the primary transmission path corresponding to the current task to be executed meeting the preset path switching conditions, the primary and backup transmission paths of the current task to be executed are switched, or the service packets to be transmitted by the current task to be executed are copied and transmitted through multiple paths, and the field communication terminal processes the service packets that arrive first. During path switching or multi-path copy transmission, the session identifier and task characteristic identifier are quickly rebound to the switching or copying path; The preset path switching conditions include: the network quality of the primary transmission path corresponding to the current task to be executed is lower than a preset threshold, or the network domain-related parameters of the primary transmission path corresponding to the current task to be executed have changed.

7. A method for remotely controlling an unmanned aerial vehicle (UAV), characterized in that, The method, which utilizes the field communication terminal of a drone, includes: Receive a service message for a currently pending task sent by a remote control backend, wherein the service message is generated by the remote control backend and sent to the field communication terminal of the UAV by the UAV remote control method as described in any one of claims 1-7; The service messages are tunneled and decapsulated, and then converted into executable messages for various tasks and services. The executable messages for various mission operations are sent to the UAV flight control system, which then drives the UAV to perform corresponding actions.

8. The method according to claim 7, characterized in that, The method further includes: During mission execution, the system receives feedback information sent by the UAV after performing corresponding actions. The feedback information includes confirmation information, status telemetry information, or collected payload information, and carries associated session identifiers and mission characteristic identifiers. Match the corresponding backhaul transmission path and service flow transmission priority according to the task characteristic identifier and session identifier associated with the backhaul information. According to the transmission strategy corresponding to the matching service flow transmission priority, the backhaul information is transmitted to the remote control backend via the matching backhaul transmission path and the corresponding public network bearer.

9. The method according to claim 8, characterized in that, The returned information includes status telemetry information or payload information collected by the UAV, and the method further includes: Based on the real-time network status information of the matched backhaul transmission path, adaptive adjustment of the transmission parameters is performed on the backhaul information. The transmission parameters include at least one of the following parameters: rate, frequency, bit rate, and frame rate.

10. A remote control system for unmanned aerial vehicles (UAVs), characterized in that, The system is divided into a remote control backend, a bearer network, and an field communication terminal according to the horizontal physical architecture, and the system is divided into a control orchestration layer, a perception scheduling layer, and a fusion network layer according to the vertical logical architecture. On the remote control backend side: The control orchestration layer is used to receive UAV control intentions from the control front end, and generate the current task to be executed and the service instructions of various task services in the current task to be executed according to the UAV control intentions; establish a task session with the field communication terminal according to the current task to be executed, and generate the service network requirements corresponding to various task services of the current task to be executed. The perception and scheduling layer is used to determine the task characteristic identifiers corresponding to various task services based on the service network requirements of various task services, and to determine the primary transmission path and the service flow transmission priority of various task services from the candidate transmission paths based on the task characteristic identifiers and the network status information of each candidate transmission path. The task characteristic identifiers are used to identify the task service types of various task services in the current task to be executed. The candidate transmission paths are end-to-end transmission paths between the remote control backend and the field communication terminal that can be used to carry task sessions. Each candidate transmission path relies on a different public network and each candidate transmission path has a corresponding end-to-end logical tunnel. The converged network layer is used to tunnel encapsulate the service instructions, task characteristic identifiers and session identifiers of various task services, and transmit the service messages of various task services to the UAV's field communication terminal through the primary transmission path via the corresponding public network according to the transmission strategy corresponding to the service transmission priority. On the field communication end: The converged network layer is used to receive service messages for currently pending tasks sent by the remote control terminal and to perform tunnel decapsulation on the received service messages; The perception and scheduling layer is used to convert the decapsulated business messages into executable messages for various tasks and services. The control orchestration layer is used to send executable messages for various mission operations to the UAV flight control system, so that the UAV flight control system can drive the UAV to perform corresponding actions.

Images