A large-scale networking method for cross-domain communication based on soft bus

Abstract

The application provides a large-scale networking method for cross-domain communication based on a soft bus, and relates to the field of communication networks.The method comprises the following steps: configuring roles for all nodes in the network, including a summary node, a group leader node and a group member node; each communication node performs initial networking according to its own role to form a tree-shaped network architecture; during the network operation, the networking topology is dynamically adjusted according to the network condition; and the connectivity state of the nodes is maintained through inter-layer heartbeat keep-alive.The application avoids the problem of limited networking scale caused by a single center node, realizes load sharing, can dynamically change the access topology, and can guarantee the service quality under the condition of large-scale networking.

Description

Technical Field

[0001] This invention relates to the field of communication networks, and in particular to a method for large-scale cross-domain communication networking based on a soft bus. Background Technology

[0002] As the integration of communication devices continues to increase, individual devices are equipped with more and more communication methods. Taking mobile phones and tablets as examples, they possess multiple communication methods such as 5G, Wi-Fi, and Bluetooth. Furthermore, smart home appliances often also feature Wi-Fi or Bluetooth. Given that physical connectivity is already available, establishing communication connections between these devices and achieving cross-device interoperability has become a crucial technical challenge that must be addressed.

[0003] The open-source HarmonyOS distributed soft bus technology aims to solve this problem. This technology provides a unified communication mechanism across devices by abstracting underlying network protocols and hardware differences. The open-source HarmonyOS distributed soft bus technology has a central node, which can be a mobile phone or a tablet, forming a star topology. However, the number of terminals supported by this structure is limited by the processing power and channel capacity of the central node, resulting in limited capacity and failing to meet the needs of large-scale networking. Furthermore, while there are actually more than one central node, it is not fully utilized. For example, in a home environment, multiple mobile phones or tablets exist simultaneously, but the open-source HarmonyOS distributed soft bus can only use one as the central node. The entire system is highly dependent on the stability of this central node, thus necessitating an expansion of the number of central nodes. Summary of the Invention

[0004] To address the issues of large-scale networking and centralized networking mentioned in the background section, this invention proposes a cross-domain communication large-scale networking method based on a soft bus.

[0005] The technical problem to be solved by this invention is achieved by the following technical solution:

[0006] A method for large-scale cross-domain communication networking based on a soft bus includes the following steps:

[0007] Step 1: Configure roles for all nodes in the network, including summary nodes, group leader nodes, and group member nodes;

[0008] Step 2: Each communication node performs initial networking according to its own role, forming a tree-shaped network architecture;

[0009] Step 3: During network operation, dynamically adjust the network topology according to network conditions;

[0010] Step 4: During network operation, the connectivity of nodes is maintained through inter-layer heartbeat keep-alive.

[0011] Furthermore, in step 1, the functions of the summary node, group leader node, and group member node are as follows:

[0012] Summary node: Located at the top layer of the network, it undertakes global coordination and management tasks, is responsible for communicating with the group leader nodes at the lower level, and handles the information summary, merging, and scheduling tasks from the group leader nodes;

[0013] Group leader node: Located in the middle layer of the network, each group leader node manages a subset of group member nodes, is responsible for communicating with the summary node, transmitting the status information and requests of the group members, and ensuring the synchronization and coordination of the group members.

[0014] Member nodes: Located at the bottom layer of the network, each member node is responsible for specific business or computing tasks and communicates with other nodes through the group leader node.

[0015] Furthermore, the specific method for step 2 is as follows:

[0016] Step 201: All nodes are initialized to their corresponding roles. Uninitialized nodes are treated as group member nodes by default.

[0017] Step 202: All nodes send a broadcast discovery request through the available communication link, carrying role information in the discovery request;

[0018] Step 202: Each node processes the discovery request according to its own role and decides whether to return a response; where:

[0019] The summary node only processes discovery requests from the group leader node. Upon receiving a discovery request, it replies with the corresponding discovery response, thereby establishing a connection between the summary node and the group leader node, and storing the connected device information in the device list.

[0020] The leader node only processes discovery requests from member nodes and summary nodes. Upon receiving a discovery request, it responds with the corresponding discovery response, thereby establishing a connection between the leader node and the summary node or between the leader node and member nodes, and storing the connected device information in the device list.

[0021] Member nodes only process discovery requests from the leader nodes. Upon receiving a discovery request, they select the leader node with the best link quality from all discovered leader nodes based on link quality, and reply with the corresponding discovery response, thereby establishing a connection between the member nodes and the leader nodes, and storing the connected device information in the device list.

[0022] Step 203: The nodes that have established connections synchronize information by synchronizing the connection status of their nodes with the connected devices. After synchronization is complete, the group leader node obtains the connection information of all group member nodes, and the summary node obtains the connection information of all group leader nodes and group member nodes.

[0023] Furthermore, the specific method for step 3 is as follows:

[0024] Step 301: Member nodes continuously discover backup group leader nodes, forming a list of group leader nodes;

[0025] Step 302: Member nodes periodically measure the channel link quality between themselves and each group leader node, and report the measurement results to the currently accessing group leader node, i.e., the source group leader node.

[0026] Step 303: When the processing capacity of the source group leader node reaches its limit, select a new group leader node with the best link quality as the destination group leader node for the subordinate group member nodes, notify the subordinate group member nodes to change the group leader node, and carry the destination group leader node identifier in the notification message.

[0027] Step 304: The member nodes check the availability of the destination group leader node and reconnect to the destination group leader node.

[0028] Furthermore, step 4 is specifically implemented as follows:

[0029] Step 401: When each node connects to the upper-layer node, the initial heartbeat cycle is negotiated.

[0030] Step 402: Member nodes and group leader nodes send heartbeat information to each other periodically according to the initial heartbeat cycle; group leader nodes and summary nodes send heartbeat information to each other periodically according to the initial heartbeat cycle; the heartbeat information includes the identifier and timestamp of the sending node.

[0031] Step 403: The upper-layer node adjusts the heartbeat cycle with the lower-layer node according to the current load, and notifies the lower-layer node to send heartbeat information according to the adjusted cycle.

[0032] Step 404: Each node determines the current heartbeat status. If the upper-level node determines that the heartbeat of the lower-level node is disconnected, it deletes the relevant information of the lower-level node and stops sending heartbeats. If the lower-level node determines that the heartbeat of the upper-level node is disconnected, it re-executes the access process.

[0033] Compared with the prior art, the present invention has the following advantages:

[0034] 1. Supports large-scale networking. This invention extends the network architecture by replacing the star topology with a tree-based network architecture, making full use of all central nodes and avoiding the limitation on network scale caused by a single central node.

[0035] 2. Node Role Differentiation. This invention changes the definition of the roles of central node and access node in the prior art, and defines the roles of communication nodes in the network as summary node, group leader node, and group member node. Each communication node adjusts its connection method and data forwarding strategy according to its role. Each group leader node only handles the traffic of the group member nodes it connects to, thereby realizing load sharing and ensuring service quality under large-scale networking conditions.

[0036] 3. Dynamic topology adjustment. When the network status changes, the group leader node can instruct member nodes to change the group leader node, avoiding too many devices connecting to a single node, thereby optimizing the network topology and improving the overall network throughput. Attached Figure Description

[0037] Figure 1 This is a network architecture diagram in an embodiment of the present invention.

[0038] Figure 2 This is a schematic diagram of the initial network configuration in an embodiment of the present invention.

[0039] Figure 3 This is a flowchart of the topology dynamic adjustment in an embodiment of the present invention.

[0040] Figure 4 This is a flowchart of the hierarchical heartbeat process in an embodiment of the present invention. Detailed Implementation

[0041] The present invention will now be further described with reference to the accompanying drawings.

[0042] A method for large-scale cross-domain communication networking based on a soft bus includes the following steps:

[0043] Step 1: Configure roles for all nodes in the network, including summary nodes, group leader nodes, and group member nodes. Specifically:

[0044] Summary Node: Located at the top layer of the network, it is responsible for communicating with the group leader nodes below and handling tasks such as summarizing, merging, and scheduling information from the group leader nodes. It undertakes global coordination and management tasks, requiring high processing power and reliability.

[0045] Group leader node: Located in the middle layer of the network. Each group leader node manages a subset of group member nodes, is responsible for communicating with the summary node, transmitting status information and requests of group members, and ensuring synchronization and coordination among group members. The number of group leader nodes is usually less than that of group member nodes, resulting in a higher load and requiring sufficient processing power.

[0046] Member nodes: Located at the bottom layer of the network, each member node is typically responsible for specific business or computing tasks and communicates with other nodes through the group leader node. Member nodes are relatively simple and have a light workload, primarily communicating and collaborating through the group leader node.

[0047] The summary node interacts only with the group leader node. The group leader node interacts with both the summary node and group member nodes. Group member nodes interact only with the group leader node. Group member nodes interact with each other through the group leader node. In the absence of a summary node, group leader nodes can interact with each other.

[0048] Network information is synchronously aggregated from bottom to top. Member nodes only have local access information, the group leader node has access information for all member nodes, and the aggregation node has access information for all nodes in the network.

[0049] Step 2: Each communication node performs initial network formation according to its own role, forming a tree-shaped network architecture, such as... Figure 1 As shown, in the tree network architecture, there is one summary node, and multiple group leader nodes and group member nodes.

[0050] The specific process of initial network setup is as follows: Figure 2 As shown, it includes:

[0051] (201) All communication nodes are initialized to their corresponding roles, including summary nodes, group leader nodes and group member nodes. Uninitialized nodes are treated as group member nodes by default.

[0052] (202) All communication nodes in the network send broadcast discovery requests through communication links, including WIFI, Star Flash, etc., and carry role information in the requests.

[0053] (203) The summary node only processes discovery requests from the group leader node. Upon receiving a discovery request, it replies with the corresponding discovery response, thereby establishing a connection between the summary node and the group leader node. At the same time, it stores the connected device information in the device list.

[0054] (204) The group leader node only processes discovery requests from group member nodes and summary nodes. Upon receiving a discovery request, it replies with the corresponding discovery response, thereby establishing a connection between the group leader node and the summary node or between the group leader node and group member nodes. At the same time, it stores the connected device information in the device list.

[0055] (205) Member nodes only process discovery requests from the group leader node. Upon receiving a discovery request, they reply with the corresponding discovery response, thereby establishing a connection between the member node and the group leader node. Simultaneously, they store the connected device information in the device list. Furthermore, member nodes should select one of the multiple group leader nodes for connection based on link quality.

[0056] (206) Each established communication node will synchronize information, sharing its connection status with the connected devices. After synchronization, the group leader node will have access to the connection information of all group member nodes, and the aggregation node will have access to the connection information of both the group leader node and the group member nodes.

[0057] Step 3: During network operation, dynamically adjust the network topology according to network conditions; such as... Figure 3 As shown, it specifically includes:

[0058] (301) The member node discovers the backup group leader node and forms a list of group leader nodes. The member node selects a group leader node (i.e., the source group leader node) to complete the initial access. The source group leader node discovers the destination group leader node.

[0059] (302) The member nodes periodically measure the channel link quality between themselves and each group leader node, and report the measurement results to the source group leader node.

[0060] (303) The source group leader node has reached its processing capacity limit and can no longer guarantee the quality of communication services between group member nodes.

[0061] (304) The source group leader node notifies the group member nodes to change the group leader node, and the notification message carries the destination group leader node identifier.

[0062] (305) The member node checks the availability of the destination node and reconnects to the destination group leader node.

[0063] Step 4: During network operation, the connectivity of nodes is maintained through inter-layer heartbeat keep-alive; such as... Figure 4 As shown, it specifically includes:

[0064] (401) Member nodes connect to the group leader node, and the group leader node connects to the summary node. The initial heartbeat cycle is negotiated during the connection process.

[0065] (402) The member nodes and the group leader node send heartbeat information to each other periodically according to the initial heartbeat cycle. The group leader node and the summary node send heartbeat information to each other periodically according to the initial heartbeat cycle.

[0066] (403) Between member nodes and the leader node, the leader node can adjust the heartbeat cycle according to the current load, and member nodes send heartbeats according to the adjusted cycle. Between the leader node and the aggregation node, the aggregation node can adjust the heartbeat cycle according to the current load, and the leader node sends heartbeats according to the adjusted cycle.

[0067] (404) Any node can determine the current heartbeat status. If the group leader node believes that a member's heartbeat is disconnected, it deletes the member's relevant information and stops sending heartbeats, thereby triggering the member to believe that the heartbeat is disconnected and execute the reconnection process. If the group leader node believes that the summary node's heartbeat is disconnected, it directly executes the reconnection process.

[0068] This invention addresses the needs of large-scale networking applications for smart devices by improving the technical architecture of distributed soft bus technology. It introduces aggregation nodes and group leader nodes, replaces the star topology with a tree-based networking architecture, and allows each communication node to adjust its connection method and data forwarding strategy according to its role. Each group leader node only handles the traffic of its connected group members, avoiding the problem of limited network scale caused by a single central node. This achieves load balancing, allows for dynamic changes in the access topology, and ensures service quality under large-scale networking conditions.

Claims

1. A method for large-scale cross-domain communication networking based on a soft bus, characterized in that, Includes the following steps: Step 1: Configure roles for all nodes in the network, including summary nodes, group leader nodes, and group member nodes; Step 2: Each communication node performs initial networking according to its own role, forming a tree-shaped network architecture; Step 3: During network operation, dynamically adjust the network topology according to network conditions; Step 4: During network operation, the connectivity of nodes is maintained through inter-layer heartbeat keep-alive.

2. The method for large-scale cross-domain communication networking based on a soft bus according to claim 1, characterized in that, In step 1, the functions of the summary node, group leader node, and group member node are as follows: Summary node: Located at the top layer of the network, it undertakes global coordination and management tasks, is responsible for communicating with the group leader nodes at the lower level, and handles the information summary, merging, and scheduling tasks from the group leader nodes; Group leader node: Located in the middle layer of the network, each group leader node manages a subset of group member nodes, is responsible for communicating with the summary node, transmitting the status information and requests of the group members, and ensuring the synchronization and coordination of the group members. Member nodes: Located at the bottom layer of the network, each member node is responsible for specific business or computing tasks and communicates with other nodes through the group leader node.

3. The method for large-scale cross-domain communication networking based on a soft bus according to claim 1, characterized in that, The specific method for step 2 is as follows: Step 201: All nodes are initialized to their corresponding roles. Uninitialized nodes are treated as group member nodes by default. Step 202: All nodes send a broadcast discovery request through the available communication link, carrying role information in the discovery request; Step 202: Each node processes the discovery request according to its own role and decides whether to return a response; where: The summary node only processes discovery requests from the group leader node. Upon receiving a discovery request, it replies with the corresponding discovery response, thereby establishing a connection between the summary node and the group leader node, and storing the connected device information in the device list. The leader node only processes discovery requests from member nodes and summary nodes. Upon receiving a discovery request, it responds with the corresponding discovery response, thereby establishing a connection between the leader node and the summary node or between the leader node and member nodes, and storing the connected device information in the device list. Member nodes only process discovery requests from the leader nodes. Upon receiving a discovery request, they select the leader node with the best link quality from all discovered leader nodes based on link quality, and reply with the corresponding discovery response, thereby establishing a connection between the member nodes and the leader nodes, and storing the connected device information in the device list. Step 203: The nodes that have established connections synchronize information by synchronizing the connection status of their nodes with the connected devices. After synchronization is complete, the group leader node obtains the connection information of all group member nodes, and the summary node obtains the connection information of all group leader nodes and group member nodes.

4. The method for large-scale cross-domain communication networking based on a soft bus according to claim 1, characterized in that, The specific method for step 3 is as follows: Step 301: Member nodes continuously discover backup group leader nodes, forming a list of group leader nodes; Step 302: Member nodes periodically measure the channel link quality between themselves and each group leader node, and report the measurement results to the currently accessing group leader node, i.e., the source group leader node. Step 303: When the processing capacity of the source group leader node reaches its limit, select a new group leader node with the best link quality as the destination group leader node for the subordinate group member nodes, notify the subordinate group member nodes to change the group leader node, and carry the destination group leader node identifier in the notification message. Step 304: The member nodes check the availability of the destination group leader node and reconnect to the destination group leader node.

5. The method for large-scale cross-domain communication networking based on a soft bus according to claim 1, characterized in that, The specific method for step 4 is as follows: Step 401: When each node connects to the upper-layer node, the initial heartbeat cycle is negotiated. Step 402: The member nodes and the group leader node send heartbeat information to each other periodically according to the initial heartbeat cycle; the group leader node and the summary node send heartbeat information to each other periodically according to the initial heartbeat cycle. The heartbeat information contains the identifier and timestamp of the sending node; Step 403: The upper-layer node adjusts the heartbeat cycle with the lower-layer node according to the current load, and notifies the lower-layer node to send heartbeat information according to the adjusted cycle. Step 404: Each node determines the current heartbeat status. If the upper-level node determines that the heartbeat of the lower-level node is disconnected, it deletes the relevant information of the lower-level node and stops sending heartbeats. If the lower-level node determines that the heartbeat of the upper-level node is disconnected, it re-executes the access process.

Images