Lightweight distributed soft bus communication method for unmanned cluster

Abstract

The technical scheme of the invention discloses a lightweight distributed soft bus communication method for an unmanned cluster. The invention discloses a lightweight unmanned cluster soft bus communication architecture, which enhances the flexibility and interoperability among different functional components in a system by designing a high-performance component integration method of an unmanned system. Moreover, an unmanned cluster lightweight communication scheme is introduced, and a real-time and rapid communication mode is provided for the unmanned cluster, so that cross-domain and cross-platform efficient integration of all unmanned system function components in the cluster is supported, and stable cooperation of the unmanned cluster in a complex environment is supported.

Description

Technical Field

[0001] The present invention relates to a lightweight distributed soft bus communication method for unmanned aerial vehicle (UAV) clusters, and belongs to the technical field of communications. Background Art

[0002] With the rapid development of unmanned technology, its applications in both military and civilian fields are becoming increasingly widespread. Unmanned swarms, as a new technological development direction, are becoming a hot topic of research due to their advantages of efficiency, flexibility, and low cost. In the civilian sector, unmanned swarms are being applied in scenarios such as logistics and distribution, agricultural plant protection, disaster relief, and film and television production, demonstrating high flexibility and cost-effectiveness. In the military, unmanned swarms can perform a variety of missions, including reconnaissance, surveillance, jamming, and attack. Their distributed nature enhances their survivability and mission efficiency.

[0003] Unmanned swarms can leverage their flexible internal networking, communication, and information sharing advantages to expand their situational awareness and mission execution capabilities, demonstrating greater coordination and intelligence. However, faced with increasingly complex mission environments and changing mission requirements, unmanned swarm systems face limitations such as tight coupling between missions and platforms and limited functional scalability, making it difficult to effectively address the challenges of cross-domain collaborative missions. Unmanned swarm platforms face greater pressure in terms of load carrying, endurance, computing power, and decision-making, highlighting their low capabilities, low efficiency, and lack of flexibility when executing missions. To enhance the ability of unmanned swarm platforms to operate swarm systems, a lightweight soft bus communication architecture needs to be designed to handle the more complex scheduling of cluster computing power, resources, tasks, communications, and other aspects.

[0004] At present, the field of soft bus research covers many aspects such as communication network construction, communication protocol design, and application scenario adaptation. These studies have laid a solid theoretical and technical foundation for soft bus research. At the same time, distributed soft buses have gradually attracted the attention of researchers. The design principles of distributed soft buses mainly revolve around achieving seamless communication, dynamic discovery, resource sharing, and collaborative work between devices. It connects multiple device ends through a "virtual bus" at the software level, enabling them to collaborate as efficiently as within the same system. The core goal of the distributed soft bus is to break the physical limitations between devices and achieve efficient communication and collaboration through wired / wireless networks. Its design includes the following key aspects:

[0005] (1) Virtualized communication channel

[0006] The soft bus simulates the functions of the hardware bus through software, creating a virtual communication channel, enabling devices to exchange data and call functions through networks such as Wi-Fi, Bluetooth, and the Internet. This virtualization design eliminates dependence on physical connections and enhances system flexibility.

[0007] (2) Service Discovery and Registration

[0008] When a new node joins the network, its services and functions are automatically registered in the Softbus system. Other nodes in the cluster dynamically find and connect to these services through the service discovery mechanism.

[0009] (3) Unified Communications Protocol

[0010] The soft bus must define a unified communication protocol to ensure that different devices can understand and process each other's data. This includes functions such as message encapsulation, routing updates, and timeout retransmission to ensure reliable and efficient communication.

[0011] (4) Resource sharing and collaboration

[0012] Unmanned systems can share resources and perception information via a soft bus. Distributed task execution is supported, such as assigning computing tasks to multiple devices for parallel processing, improving efficiency.

[0013] (5) Scalability and flexibility

[0014] The design of the distributed soft bus fully considers the scalability and adaptability of the system, and can cope with the growth of the number of devices and changes in the network environment.

[0015] Unmanned systems contain multiple independent functional modules, such as communication, positioning, and control. Each module focuses on a specific function and interacts with other modules through specific interfaces. Given the dynamic nature of device resource states, the soft bus must support the dynamic loading and unloading of functional modules. For example, when a device enters low-power mode, the system can unload non-essential modules to conserve resources; as mission demands increase, additional functional modules can be loaded. This dynamic adjustment capability ensures the efficient operation of the soft bus in diverse scenarios.

[0016] Communication between unmanned systems involves a variety of message types, including control commands, sensor data, and status updates. The diversity and complexity of these messages pose challenges to communication efficiency and resource utilization. To enable real-time communication within unmanned clusters, a standardized inter-software component communication and access mechanism must be provided to the upper layer. This unified encapsulation of communication modes between different processors and between different software components on the same processor ensures stable and reliable communication across multiple layers.

[0017] The design principle of a distributed soft bus is to achieve efficient connectivity and resource sharing between devices through mechanisms such as virtualized communication channels, service discovery, and unified protocols. However, because unmanned systems are typically resource-constrained, highly dynamic, and highly real-time, the design of a distributed soft bus for unmanned clusters must strike a balance between lightweight, low power consumption, and efficient collaboration to achieve efficient communication and resource sharing within resource constraints. Summary of the Invention

[0018] The technical problem to be solved by the present invention is that the existing unmanned cluster tasks are tightly coupled with the platform, the functional scalability is not strong, and it is difficult to meet the needs of cross-domain collaborative tasks in complex scenarios.

[0019] In order to solve the above technical problems, the technical solution of the present invention is to disclose a lightweight distributed soft bus communication method for unmanned clusters, which is characterized by comprising the following steps:

[0020] A distributed soft bus communication model is designed for unmanned clusters. The distributed soft bus communication model meets the requirements of efficient, real-time, and reliable communication between unmanned systems in the cluster.

[0021] Based on the distributed soft bus communication model, a high-performance component integration method for unmanned systems and a lightweight communication solution for unmanned clusters are designed, including:

[0022] The high-performance component integration method for unmanned systems provides modularization of functional components and efficient interaction of components, ensuring the autonomy and effectiveness of unmanned systems in performing complex tasks;

[0023] The unmanned cluster lightweight communication solution provides a set of standardized interfaces and protocols, and provides a standard communication and access mechanism between software components to the upper layer. It uniformly encapsulates the communication between different processors and between different functional components under the same processor, and completes stable and reliable communication transmission across layers.

[0024] Preferably, the construction of the distributed soft bus communication model includes the following steps:

[0025] Step 101: Build an unmanned cluster mutual discovery service, specifically including the following steps:

[0026] Step 101-1: Each node in the unmanned cluster registers its own information with the discovery server on the unmanned cluster soft bus upon startup, including participant information and data endpoint information. Participant information includes GUID, QoS policy, and transport protocol, while data endpoint information includes Topic, Type, endpoint QoS, endpoint GUID, and domain participant.

[0027] Step 101-2: After registration is completed, cluster information will be centrally managed on a master node. A mutual discovery topic route will be constructed based on the associated information. Nodes can obtain a list and detailed information of available services in the cluster by querying the mutual discovery topic route based on communication needs.

[0028] Step 101-3: When a node exits or a service instance is deregistered, it is automatically removed from the registry to ensure that other nodes do not access the unavailable service;

[0029] Step 102: Distributed soft bus communication mode selection

[0030] Step 102-1: Monitor the current communication environment of the unmanned system in real time;

[0031] Step 102-2: When the communication environment is highly loaded, the communication mode of topic communication is selected by default;

[0032] Step 102-3: Under normal communication conditions, based on the communication requirements of the current mission scenario, in high-concurrency communication scenarios, that is, when multiple functional components in the unmanned system need to receive the same type of data updates at the same time, select the topic communication mode and participate in the communication as a publisher or subscriber; in scenarios where immediate feedback or high-priority communication is required, switch to the request service communication mode; in scenarios where globally consistent parameter storage and updates are required, switch to the parameter server communication mode;

[0033] Step 103: Interaction between functional components of unmanned systems

[0034] Unmanned system transmitter:

[0035] Step 103-1A: Based on the information exchange requirements in the actual mission scenario, each functional component within the unmanned system encapsulates the messages to be transmitted;

[0036] Step 103-2A: Select the communication mode as needed according to step 102, and transmit the relevant message to the communication module within the system;

[0037] Step 103-3A: The communication module serializes various complex messages and uniformly encapsulates the data streams according to the standard real-time publish-subscribe protocol, abstracting and unifying any type of functional messages;

[0038] Step 103-4A: The communication module obtains other node information based on the mutual discovery route established in step 101, and sends the message to the corresponding topic via the unmanned cluster soft bus;

[0039] Unmanned system receiving end:

[0040] Step 103-1B: The communication module receives corresponding topic messages from the soft bus based on the other node information obtained from the mutual discovery route established in step 101 and the communication requirements in the actual task scenario;

[0041] Step 103-2B: The communication module performs deserialization processing on the data stream in the received message and restores it to a complex message corresponding to the functional component interface;

[0042] Step 103 - 3B: Select a communication mode as needed according to step 102 and transmit the converted message to the corresponding functional component.

[0043] Preferably, in the high-performance component integration method for unmanned systems, a unified abstract design of functional component interfaces is proposed. By modularizing the three core components of perception, decision-making, and control, efficient, coordinated, and intelligent operation of the system is achieved. High-performance components are modularized and linked to the soft bus in an interactive manner with communication modules. Unified interface standards and communication protocols simplify the interaction between different components and reduce integration complexity.

[0044] The efficient interaction method of unmanned system components specifically includes the following steps:

[0045] Step 201: Modularize the high-performance components of the unmanned system, specifically including the following steps:

[0046] Step 201-1: The modular design of the perception component is mainly targeted at different perception devices, and the information of different perception devices is encapsulated into a unified format for output;

[0047] Step 201-2: The modular design of the control component should focus on flight path planning and attitude control, output a time-constrained path point sequence, and provide real-time feedback of status data;

[0048] Step 201-3: The componentized decision module uses a sliding window mechanism to align the timestamps of the perception data, status data, and task instructions, eliminating the timing deviation caused by asynchronous transmission. It then integrates multi-source data based on the task scenario to update the decision and output standardized decision information.

[0049] Step 202: Interaction between functional modules in the unmanned system, specifically including the following steps:

[0050] Step 202-1: The communication module is started as a message agent to manage topic registration, message routing and distribution;

[0051] Step 202-2: The functional module to send the message is the publisher, which binds the modularized encapsulated message to the topic and sends it to the message broker;

[0052] Step 202-3: The function module to receive messages monitors the messages pushed by the message agent for the subscriber and receives messages on a specific topic;

[0053] Step 202-4: The message broker matches the registered subscriber list according to the topic and asynchronously distributes the message to all matching subscribers through the message push mechanism.

[0054] Preferably, in the unmanned cluster lightweight communication solution, a many-to-many one-way data exchange communication model is designed, in which the unmanned system to send the message publishes the message to the local cache of the unmanned system to receive the message, and the information flow is controlled by the service quality policy established between the entities responsible for data exchange; at the same time, communication can occur between multiple domains, that is, an isolated abstract plane connecting all distributed nodes that can communicate with each other, and only entities belonging to the same domain can communicate with each other. The matching between the entity subscribing to the data and the entity publishing the data is mediated by the topic, realizing agile combination and efficient interaction between the components in the cluster, so as to ensure that the unmanned cluster can improve its adaptability and collaboration capabilities in a complex and changing environment.

[0055] Preferably, the unmanned cluster lightweight communication structure includes the following steps:

[0056] Step 301: Unmanned system metadata registration and cluster information update, specifically including the following steps

[0057] Step 301-1: According to the distributed soft bus communication model, using the mutual discovery topic routing under the distributed soft bus communication model, when the unmanned system in the cluster starts, it actively connects to the pre-configured discovery server and registers its own metadata;

[0058] Step 301-2: The mutual discovery server stores the node information persistently to form a global mutual discovery topic route;

[0059] Step 301-3: When a new node joins, the server immediately distributes its metadata to the relevant registered nodes and pushes the information of existing nodes to the new node.

[0060] Step 302: Component information interaction within the unmanned cluster, specifically including the following steps:

[0061] Message sender:

[0062] Step 302-1A: After establishing a connection path based on the mutual discovery topic routing, the publisher, i.e., the unmanned system sending the message, transmits the message generated by the functional module to the communication module through the unmanned cluster lightweight communication solution;

[0063] Step 302-2A: The communication module uses an efficient serialization function to binary encode the message and calls a standardized interface to encapsulate the data into a domain-specific message object;

[0064] Step 302-3A: Verify format consistency through metadata pre-registered data types;

[0065] Message receiving end:

[0066] Step 302-1B: After establishing a connection path based on the mutually discovered topic routing, the subscriber, i.e., the unmanned system receiving the message, uses a communication module to monitor the specified topic message stream and asynchronously receive messages through a ring buffer;

[0067] Step 302-2B: Perform integrity check on the message based on the check code in the message message. If there is any inconsistency, the invalid data packet is discarded and a message retransmission request is triggered.

[0068] Step 302-3B: After the integrity verification is passed, the data stream in the message payload is deserialized and converted into a message format compatible with the functional module interface;

[0069] Step 302-4B: According to the unmanned system high-performance component integration method, the converted message is injected into the local cache queue and forwarded to the corresponding functional component interface.

[0070] This paper discloses a lightweight unmanned cluster soft bus communication architecture. By designing a high-performance component integration method for unmanned systems, it enhances the flexibility and interoperability between different functional components within the system. Furthermore, the present invention introduces a lightweight communication solution for unmanned clusters, providing real-time and fast communication for unmanned clusters. This supports efficient cross-domain and cross-platform integration of various unmanned system functional components within the cluster, and supports stable collaboration of unmanned clusters in complex environments.

[0071] Compared with the existing technical solutions, the present invention has the following beneficial effects:

[0072] 1) The present invention can realize module information interaction between unmanned clusters and supports rapid deployment. For any type of module message, it can forward it through the communication module to achieve multi-node information transmission;

[0073] 2) The present invention proposes a lightweight real-time communication method that supports multiple communication modes and flexible transmission content. Communication equipment can be freely selected according to communication needs and equipment selection, thereby improving the usability and applicability of the communication system.

[0074] 3) The lightweight distributed soft bus architecture proposed in this invention has good scalability and portability, is suitable for most unmanned cluster systems with cluster communication requirements, and is simple to configure and easy to operate. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 This is the structural diagram of the distributed soft bus communication model;

[0076] Figure 2 Schematic diagram for application scenarios of multiple communication modes;

[0077] Figure 3 Provide a structural diagram for efficient interaction methods between unmanned system components;

[0078] Figure 4 This is a diagram of the lightweight communication structure of an unmanned swarm;

[0079] Figure 5 To discover the routing principle diagram of the subject;

[0080] Figure 6 Publish a message flow chart for publishers;

[0081] Figure 7 Flowchart for subscriber message subscription. DETAILED DESCRIPTION

[0082] Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiments are only used to illustrate the present invention and are not used in limiting the scope of the present invention.In addition, should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall equally within the scope limited by the appended claims of the application.

[0083] The embodiment of the present invention discloses a lightweight distributed soft bus communication method for unmanned clusters, including the following contents:

[0084] A distributed soft bus communication model is designed for unmanned clusters. It meets the requirements for efficient, real-time, and reliable communication between unmanned systems (such as drones, unmanned vehicles, and unmanned boats) within the cluster. This distributed soft bus communication model creates virtualized communication channels and supports automated service discovery and registration mechanisms. This allows each unmanned system within the cluster to be dynamically identified and exchange data and call functional components over a network (such as Wi-Fi, Bluetooth, or 5G).

[0085] In an embodiment of the present invention, the construction of a distributed soft bus communication model includes the following steps:

[0086] Step 1: Build an unmanned cluster mutual discovery service

[0087] Step 1-1: Each node in the unmanned cluster, i.e., the unmanned system, registers its own information with the Discovery Server on the unmanned cluster's softbus upon startup, including participant information (DomainParticipant) and data endpoint information (DataWriter / DataReader). Participant information includes GUID (Globally Unique Identifier), QoS (Quality of Service) policy, and transport protocol. Data endpoint information includes Topic, Type, endpoint QoS, endpoint GUID, and domain participant.

[0088] Step 1-2: After registration is complete, cluster information is centrally managed on a master node, and a discovery topic route is constructed based on the associated information. Nodes query the discovery topic route to obtain a list and detailed information about available services in the cluster based on their communication needs.

[0089] Steps 1-3: When a node exits or a service instance is deregistered, it can be automatically removed from the registry to ensure that other nodes do not access the unavailable service.

[0090] Step 2: Distributed soft bus communication mode selection

[0091] Step 2-1: Monitor the current communication environment of the unmanned system in real time, including network traffic, load balancing, latency, packet loss rate, etc. Network traffic, load balancing, latency, and packet loss rate are obtained using the Netdata tool.

[0092] Step 2-2: When the communication environment is highly loaded, the topic communication mode is selected by default.

[0093] Step 2-3: Under normal communication conditions, based on the communication requirements of the current mission scenario, in high-concurrency communication scenarios, that is, when multiple functional components in the unmanned system need to receive the same type of data updates at the same time, select the topic communication mode and participate in the communication as a publisher or subscriber; in communication scenarios that require immediate feedback or high priority, switch to the request service communication mode; in scenarios that require global consistency in parameter storage and updates, switch to the parameter server communication mode.

[0094] Step 3: Interaction of functional components between unmanned systems

[0095] Unmanned system transmitter:

[0096] Step 3-1: Based on the information exchange requirements in the actual mission scenario, the functional components within the unmanned system, such as perception components, control components, decision-making components, etc., encapsulate the messages that need to be transmitted.

[0097] Step 3-2: Select the communication mode as needed according to step 2, and pass the relevant message to the communication module within the system.

[0098] Step 3-3: The communication module serializes a variety of complex messages and uniformly encapsulates the data stream according to the standard real-time publish-subscribe protocol, abstracting and unifying any type of functional messages.

[0099] Step 3-4: The communication module obtains other node information based on the mutual discovery route built in step 1, and relies on the unmanned cluster soft bus to send messages to the corresponding topics.

[0100] Unmanned system receiving end:

[0101] Step 3-1: The communication module obtains other node information based on the mutual discovery route built in step 1 and receives corresponding topic messages from the soft bus based on the communication requirements in the actual task scenario.

[0102] Step 3-2: The communication module deserializes the data stream in the received message and restores it to a complex message corresponding to the functional component interface.

[0103] Step 3-3: Select the communication mode as needed according to step 2 and transmit the converted message to the corresponding functional component.

[0104] Based on this communication model, a high-performance component integration method for unmanned systems and a lightweight communication model for unmanned swarms were designed. This high-performance component integration method provides core capabilities such as modularization of functional components and efficient component interaction, ensuring the autonomy and effectiveness of unmanned systems when performing complex tasks. The lightweight communication model for unmanned swarms provides a standardized set of interfaces and protocols, offering a standardized communication and access mechanism between software components to upper layers. It also provides unified encapsulation for communication between different processors and between different functional components within the same processor, achieving stable and reliable communication transmission across multiple layers.

[0105] This paper proposes a unified abstract design for functional component interfaces. By modularizing the three core components of perception, decision-making, and control, it achieves efficient, coordinated, and intelligent system operation. High-performance components are modularized and linked to the soft bus through interaction with communication modules. Unified interface standards and communication protocols simplify interaction between different components and reduce integration complexity.

[0106] In an embodiment of the present invention, a method for efficient interaction of unmanned system components includes the following steps:

[0107] Step 1: Modularization of high-performance components for unmanned systems

[0108] Step 1-1: For perception devices such as RGBD cameras, lidars, and infrared sensors, encapsulate device messages and unify the output format (including point clouds, semantic maps, dynamic target lists, and confidence labels).

[0109] Step 1-2: The modular design of the control component needs to focus on flight path planning and attitude control, output a sequence of path points with time constraints, and provide real-time feedback on status data such as attitude angle and position error.

[0110] Steps 1-3: The componentized decision module uses a sliding window mechanism to align the timestamps of perception data, status data, and task instructions, eliminating the timing deviation caused by asynchronous transmission. It integrates multi-source data to update the decision based on the task scenario (including search, tracking, and emergency obstacle avoidance), and outputs standardized decision information (including destination coordinates, heading angle adjustment, and speed setting).

[0111] Step 2: Interaction between functional modules within the unmanned system

[0112] Step 2-1: The communication module is started as a message broker to manage topic registration, message routing, and distribution.

[0113] Step 2-2: The functional module to send the message is the publisher, which binds the modularized encapsulated message to the topic and sends it to the message broker.

[0114] Step 2-3: The function module to receive messages listens to the messages pushed by the message agent as a subscriber and receives messages on specific topics.

[0115] Step 2-4: The message broker matches the registered subscriber list according to the topic and asynchronously distributes the message to all matching subscribers through the message push mechanism.

[0116] In an embodiment of the present invention, a many-to-many one-way data exchange communication model is designed, in which the unmanned system to send a message publishes the message to the local cache of the unmanned system to receive the message. The information flow is controlled by the Quality of Service (QoS) policy established between the entities responsible for data exchange. At the same time, communication can occur between multiple domains, that is, isolated abstract planes that connect all distributed nodes that can communicate with each other. Only entities belonging to the same domain can communicate with each other, and the matching between the entity that subscribes to the data and the entity that publishes the data is mediated by the topic, which realizes the agile combination and efficient interaction between the components in the cluster, so as to ensure that the unmanned cluster can improve its adaptability and collaboration capabilities in complex and changing environments.

[0117] The unmanned swarm lightweight communication structure includes the following steps:

[0118] Step 1: Unmanned system metadata registration and cluster information update

[0119] Step 1-1: Based on the distributed soft bus communication model described above, the mutual discovery of topic routing is used. When the unmanned system in the cluster starts, it actively connects to the pre-configured discovery server and registers its own metadata, such as participating domains, topic names, data types, and QoS policies.

[0120] Step 1-2: The mutual discovery server stores the node information persistently to form a global mutual discovery topic routing.

[0121] Steps 1-3: When a new node joins, the server immediately distributes its metadata to the relevant registered nodes (i.e., nodes in the same domain or with the same topic), and pushes the information of existing nodes to the new node.

[0122] Step 2: Information exchange between components within the unmanned cluster

[0123] Message sender:

[0124] Step 2-1: After establishing a connection path based on the mutual discovery topic routing, the publisher, i.e., the unmanned system that sends the message, passes the message generated by the functional module to the communication module through the method of content 3.

[0125] Step 2-2: The communication module uses an efficient serialization function to binary encode the message. It then calls a standardized interface to encapsulate the data into a domain-specific message object (including timestamp, data source ID, message payload, and checksum).

[0126] Step 2-3: Verify format consistency through metadata pre-registered data types.

[0127] Message receiving end:

[0128] Step 2-1: After establishing a connection path based on the mutual discovery topic routing, the subscriber, i.e., the unmanned system that receives messages, uses a communication module to listen to the specified topic message stream and asynchronously receive messages through a ring buffer.

[0129] Step 2-2: Perform integrity check on the message based on the checksum in the message. If they are inconsistent, the invalid data packet is discarded and a message retransmission request is triggered.

[0130] Step 2-3: After the integrity verification is passed, the data stream in the message payload is deserialized and converted into a message format compatible with the functional module interface.

[0131] Step 2-4: According to the method described above, inject the converted message into the local cache queue and forward it to the corresponding functional component interface.

Claims

1. A lightweight distributed soft bus communication method for unmanned clusters, characterized in that: The following steps are involved: A distributed soft bus communication model is designed for unmanned clusters. The distributed soft bus communication model meets the requirements of efficient, real-time, and reliable communication between unmanned systems in the cluster. Based on the distributed soft bus communication model, a high-performance component integration method for unmanned systems and a lightweight communication solution for unmanned clusters are designed, including: The high-performance component integration method for unmanned systems provides modularization of functional components and efficient interaction of components, ensuring the autonomy and effectiveness of unmanned systems in performing complex tasks; The unmanned cluster lightweight communication solution provides a set of standardized interfaces and protocols, and provides a standard communication and access mechanism between software components to the upper layer. It uniformly encapsulates the communication between different processors and between different functional components under the same processor, and completes stable and reliable communication transmission across layers.

2. A lightweight distributed soft bus communication method for unmanned clusters according to claim 1, characterized in that: The construction of the distributed soft bus communication model includes the following steps: Step 101: Build an unmanned cluster mutual discovery service, specifically including the following steps: Step 101-1: Each node in the unmanned cluster registers its own information with the discovery server on the unmanned cluster soft bus upon startup, including participant information and data endpoint information. Participant information includes GUID, QoS policy, and transport protocol, while data endpoint information includes Topic, Type, endpoint QoS, endpoint GUID, and domain participant. Step 101-2: After registration is completed, cluster information will be centrally managed on a master node. A mutual discovery topic route will be constructed based on the associated information. Nodes can obtain a list and detailed information of available services in the cluster by querying the mutual discovery topic route based on communication needs. Step 101-3: When a node exits or a service instance is deregistered, it is automatically removed from the registry to ensure that other nodes do not access the unavailable service; Step 102: Distributed soft bus communication mode selection Step 102-1: Monitor the current communication environment of the unmanned system in real time; Step 102-2: When the communication environment is highly loaded, the communication mode of topic communication is selected by default; Step 102-3: Under normal communication conditions, based on the communication requirements of the current mission scenario, in high-concurrency communication scenarios, that is, when multiple functional components in the unmanned system need to receive the same type of data updates at the same time, select the topic communication mode and participate in the communication as a publisher or subscriber; in scenarios where immediate feedback or high-priority communication is required, switch to the request service communication mode; in scenarios where globally consistent parameter storage and updates are required, switch to the parameter server communication mode; Step 103: Interaction between functional components of unmanned systems Unmanned system transmitter: Step 103-1A: Based on the information exchange requirements in the actual mission scenario, each functional component within the unmanned system encapsulates the messages to be transmitted; Step 103-2A: Select the communication mode as needed according to step 102, and transmit the relevant message to the communication module within the system; Step 103-3A: The communication module serializes various complex messages and uniformly encapsulates the data streams according to the standard real-time publish-subscribe protocol, abstracting and unifying any type of functional messages; Step 103-4A: The communication module obtains other node information based on the mutual discovery route established in step 101, and sends the message to the corresponding topic via the unmanned cluster soft bus; Unmanned system receiving end: Step 103-1B: The communication module receives corresponding topic messages from the soft bus based on the other node information obtained from the mutual discovery route established in step 101 and the communication requirements in the actual task scenario; Step 103-2B: The communication module performs deserialization processing on the data stream in the received message and restores it to a complex message corresponding to the functional component interface; Step 103 - 3B: Select a communication mode as needed according to step 102 and transmit the converted message to the corresponding functional component.

3. A lightweight distributed soft bus communication method for unmanned clusters according to claim 1, characterized in that: In the high-performance component integration method for unmanned systems, a unified abstract design of functional component interfaces is proposed. By modularizing the three core components of perception, decision-making, and control, efficient, coordinated, and intelligent operation of the system is achieved. High-performance components are modularized and linked to the soft bus in an interactive manner with communication modules. Unified interface standards and communication protocols simplify the interaction between different components and reduce integration complexity. The efficient interaction method of unmanned system components specifically includes the following steps: Step 201: Modularize the high-performance components of the unmanned system, specifically including the following steps: Step 201-1: The modular design of the perception component is mainly targeted at different perception devices, and the information of different perception devices is encapsulated into a unified format for output; Step 201-2: The modular design of the control component should focus on flight path planning and attitude control, output a time-constrained path point sequence, and provide real-time feedback of status data; Step 201-3: The componentized decision module uses a sliding window mechanism to align the timestamps of the perception data, status data, and task instructions, eliminating the timing deviation caused by asynchronous transmission. It then integrates multi-source data based on the task scenario to update the decision and output standardized decision information. Step 202: Interaction between functional modules in the unmanned system, specifically including the following steps: Step 202-1: The communication module is started as a message agent to manage topic registration, message routing and distribution; Step 202-2: The functional module to send the message is the publisher, which binds the modularized encapsulated message to the topic and sends it to the message broker; Step 202-3: The function module to receive messages monitors the messages pushed by the message agent for the subscriber and receives messages on a specific topic; Step 202-4: The message broker matches the registered subscriber list according to the topic and asynchronously distributes the message to all matching subscribers through the message push mechanism.

4. A lightweight distributed soft bus communication method for unmanned clusters according to claim 1, characterized in that: In the unmanned cluster lightweight communication scheme, a many-to-many one-way data exchange communication model is designed, in which the unmanned system to send a message publishes the message to the local cache of the unmanned system to receive the message, and the information flow is controlled by the service quality policy established between the entities responsible for data exchange; at the same time, communication can occur between multiple domains, that is, an isolated abstract plane connecting all distributed nodes that can communicate with each other. Only entities belonging to the same domain can communicate with each other. The matching between the entity subscribing to the data and the entity publishing the data is mediated by the topic, realizing agile combination and efficient interaction between the components within the cluster, so as to ensure that the unmanned cluster can improve its adaptability and collaboration capabilities in a complex and changing environment.

5. A lightweight distributed soft bus communication method for unmanned clusters according to claim 4, characterized in that: The unmanned swarm lightweight communication structure includes the following steps: Step 301: Unmanned system metadata registration and cluster information update, specifically including the following steps Step 301-1: According to the distributed soft bus communication model, using the mutual discovery topic routing under the distributed soft bus communication model, when the unmanned system in the cluster starts, it actively connects to the pre-configured discovery server and registers its own metadata; Step 301-2: The mutual discovery server stores the node information persistently to form a global mutual discovery topic route; Step 301-3: When a new node joins, the server immediately distributes its metadata to the relevant registered nodes and pushes the information of existing nodes to the new node. Step 302: Component information interaction within the unmanned cluster, specifically including the following steps: Message sender: Step 302-1A: After establishing a connection path based on the mutual discovery topic routing, the publisher, i.e., the unmanned system sending the message, transmits the message generated by the functional module to the communication module through the unmanned cluster lightweight communication solution; Step 302-2A: The communication module uses an efficient serialization function to binary encode the message and calls a standardized interface to encapsulate the data into a domain-specific message object; Step 302-3A: Verify format consistency through metadata pre-registered data types; Message receiving end: Step 302-1B: After establishing a connection path based on the mutually discovered topic routing, the subscriber, i.e., the unmanned system receiving the message, uses a communication module to monitor the specified topic message stream and asynchronously receive messages through a ring buffer. Step 302-2B: Perform integrity check on the message based on the checksum in the message message. If there is any inconsistency, the invalid data packet is discarded and a message retransmission request is triggered. Step 302-3B: After the integrity verification is passed, the data stream in the message payload is deserialized and converted into a message format compatible with the functional module interface; Step 302-4B: According to the unmanned system high-performance component integration method, the converted message is injected into the local cache queue and forwarded to the corresponding functional component interface.

Images