A Decentralized Variable Topology Networking System and Method Based on a Hierarchical Management Architecture

By dividing the Ad-hoc network into master station, base station, and slave station layers, with the master station layer handling routing and addressing, and the base station and slave station layers not triggering network-wide routing updates, the problems of network storms and limited node capacity are solved, achieving a highly stable and low-complexity networking scheme suitable for complex geographical environments and emergency communications.

CN123075593APending Publication Date: 2026-11-10BEIJING SIDA RUIKANG TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611587230.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-10-14
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Existing Ad-hoc networks suffer from network storms, limited node capacity, low communication rates, and high system complexity due to frequent node joining or leaving, especially under poor channel conditions, making it difficult to meet the needs of large-scale deployment.

Method used

The Ad-hoc network is divided into a master station layer, a base station layer, and a slave station layer. The master station layer is responsible for routing and addressing functions. The base station and slave station layers do not trigger network-wide routing table updates. Roles are determined by device identification number segments. A unified hardware platform is adopted. Master station nodes exchange routing information and perform distributed takeover in abnormal situations.

Benefits of technology

It significantly improves the node carrying capacity and stability of the network under poor channel conditions, reduces system complexity and operation and maintenance costs, and ensures rapid recovery and communication continuity of the network when local nodes fail.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123075593A_ABST
    Figure CN123075593A_ABST
Patent Text Reader

Abstract

This invention discloses a decentralized, variable-topology Ad-hoc network system based on a hierarchical management architecture, relating to the field of wireless communication network technology. The system divides the network into a master station layer, a base station layer, and a slave station layer. Master station nodes in the master station layer constitute a decentralized, variable-topology Ad-hoc network, undertaking routing and addressing functions and maintaining the network-wide routing table. Base station nodes and slave nodes connect to either the master station node or the base station node, respectively, without undertaking routing and addressing functions. Furthermore, their network entry and exit status changes are only processed within the local star connection range and do not trigger updates to the network-wide routing table. This invention, by separating routing information maintenance from terminal access, effectively suppresses network storms caused by frequent node entry and exit, improves the stability and node carrying capacity of large-scale self-organized networking under narrowband and harsh channel conditions, and is suitable for scenarios such as mountainous areas, maritime environments, and disaster emergency response.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of wireless communication network technology, and in particular to an Ad-hoc decentralized variable topology networking system and method based on a hierarchical management architecture. Background Technology

[0002] Wireless ad hoc networks (VOCs) are communication network architectures that do not rely on pre-existing infrastructure and are formed by individual nodes. Due to their flexible deployment and strong resilience, they have broad application prospects in areas such as long-distance communication in mountainous and forested areas, disaster emergency communication, maritime information access, long-distance IoT transmission, and vehicle-to-everything (V2X) and ship-to-everything (S2X) networks. Ad-hoc networks, as a typical representative of VOCs, support dynamic node joining and leaving, and can adapt to frequent changes in network topology, making them an important technical means to achieve reliable communication in the aforementioned scenarios. However, existing Ad-hoc networking technologies still face several prominent problems in practical applications.

[0003] Conventional UHF / UHF Ad-hoc networks typically employ a single-layer flat architecture, meaning all nodes in the network are equal and can participate in the exchange and maintenance of network-wide routing information. In this architecture, the addition or departure of any node triggers an update of the entire network's routing table. In real-world deployment environments, due to poor channel conditions and frequent node movement, the rate of network topology change often far exceeds the convergence rate of routing protocols. A large number of routing update messages compete and collide in narrowband channels, easily triggering network storms. Once a network storm occurs, it significantly slows down network convergence and, in severe cases, can even cause network paralysis. This phenomenon directly limits the node capacity of a single-layer Ad-hoc network, typically supporting only about twenty nodes operating stably, making it difficult to meet the needs of large-scale deployments.

[0004] Furthermore, existing routing mechanisms in ad-hoc networks suffer from over-reliance on core nodes. In conventional network communication, data transmission between any two nodes requires routing to determine the forwarding path, heavily relying on the core node responsible for address resolution and packet forwarding. In complex geographical environments and highly dynamic, challenging channel conditions, terminal nodes and core nodes often cannot communicate directly due to channel obstruction or excessive distance. This means that even if two terminal nodes are within each other's wireless coverage area, they cannot bypass the core node for direct communication, severely reducing network throughput. Simultaneously, all communication depends on core node addressing and forwarding, resulting in double the consumption of network and channel resources, further exacerbating the problem of limited node capacity in narrowband channels.

[0005] On the other hand, the design of existing wireless ad hoc network equipment also increases system complexity and operation and maintenance costs. In traditional solutions, different network roles typically correspond to different hardware designs. For example, base stations, relay equipment, and user terminals each employ different hardware and software architectures, resulting in a wide variety of equipment, complex system integration, and high construction and maintenance costs. This hinders rapid deployment and flexible application in scenarios such as emergency communications and remote areas. Furthermore, traditional wireless networking protocols generally only apply to a single ground wave line-of-sight transmission mode, failing to integrate the advantages of multiple transmission methods such as skywave and ground wave, leading to short transmission distances and low communication rates in complex geographical environments such as mountains and forests. In summary, existing technologies suffer from limited ad-hoc network node capacity, low communication rates, and high system complexity, necessitating a networking solution that can effectively suppress network storms, increase node capacity, and reduce system complexity. Summary of the Invention

[0006] The purpose of this invention is to provide a decentralized, variable topology ad-hoc networking system and method based on a hierarchical management architecture. By dividing the ad-hoc network into a master layer, a base station layer, and a slave layer, the master layer is responsible for routing and maintaining the entire network routing table. Base station nodes and slave nodes do not undertake routing, and their entry and exit status changes do not trigger updates to the entire network routing table. This improves the routing stability and node carrying capacity of the network under frequent topology changes, and solves the problem of network storms caused by frequent node entry and exit triggering of network routing updates in single-layer ad-hoc networks, resulting in severely limited node capacity.

[0007] To address the aforementioned technical problems, a first aspect of this invention provides a decentralized, variable-topology Ad-hoc network system based on a hierarchical management architecture, comprising a master station layer, a base station layer, and a slave station layer. The master station nodes in the master station layer constitute a decentralized, variable-topology Ad-hoc network and undertake routing and addressing functions, generating and maintaining a network-wide routing table by exchanging routing information. The base station nodes in the base station layer connect to the master station nodes but do not undertake routing and addressing functions, and changes in the network entry and exit states of the base station nodes do not trigger updates to the network-wide routing table. The slave station nodes in the slave station layer connect to either the master station nodes or the base station nodes, do not undertake routing and addressing functions, and changes in the network entry and exit states of the slave station nodes do not trigger updates to the network-wide routing table.

[0008] Furthermore, the master node, the base station node, and the slave node are all configured with a device identification number. The device identification number is divided into non-overlapping master number segments, base station number segments, and slave number segments. Each node determines itself as the master node, the base station node, or the slave node based on the number segment to which its device identification number belongs.

[0009] Furthermore, in the master station layer, each master station node is configured to execute a self-organizing networking protocol, discover neighboring master station nodes by sending probe signal frames, and establish the Ad-hoc network with other master station nodes. In the Ad-hoc network, routing information is exchanged to generate and maintain the network-wide routing table.

[0010] Furthermore, the network-wide routing table only contains the address information of other master nodes in the master station layer, and does not contain the address information of the base station nodes or the slave nodes; the master station node maintains the local access tables of its subordinate base station nodes and slave nodes, and the local access tables are not included in the network-wide routing table.

[0011] Furthermore, the base station node is configured to send a network access request to a pre-designated master station node and establish a star connection to extend the signal coverage of the pre-designated master station node; the network access and disconnection status changes of the base station node are only processed within the star connection range and do not trigger the update of the entire network routing table; when the base station node detects an anomaly in the master station node it has accessed, the base station node scans and accesses other master station nodes.

[0012] Furthermore, the slave node is configured to send a network access request to a pre-specified master node or a pre-specified base station node by default and establish a star connection; when roaming is enabled and the pre-specified master node or the pre-specified base station node is unreachable, it automatically selects an accessible master node or base station node and establishes a star connection; the network access and disconnection status changes of the slave node are only processed within the scope of the star connection and do not trigger the update of the entire network routing table.

[0013] Furthermore, the master node, the base station node, and the slave node all use the same core hardware platform and start different working modes by writing device identification numbers of different number segments, thereby respectively undertaking the network roles of master station, base station, and slave station.

[0014] Furthermore, the base station node and the slave node do not participate in the routing information exchange between the master station nodes; the routing information exchange between the master station nodes is performed only by the master station nodes in the master station layer.

[0015] Furthermore, the master node, base station node, and slave node can all scan neighboring nodes that can communicate directly in one hop, generate and maintain a neighboring node table. When the destination address of the data to be sent exists in the neighboring node table, the node device can directly send the data to the corresponding neighboring node without going through the master node for routing and forwarding.

[0016] A second aspect of this invention provides an Ad-hoc decentralized variable topology networking method based on a hierarchical management architecture, applied to the aforementioned networking system, and executed by the master station node in the networking system, comprising the following steps: Step S1: During operation, the master node exchanges routing information with neighboring master nodes. The routing information includes a summary of the entire network routing table. Based on the received summary information, the master node maintains a routing shadow table of neighboring master nodes locally. The routing shadow table contains at least some entries of the entire network routing table maintained by the neighboring master nodes. The routing shadow table contains routing entries, and each routing entry includes the destination master node identifier, next hop, hop count, and version number. Step S2: When the master node detects an abnormal neighboring master node, it initiates a succession voting request to other master nodes. The succession voting request includes the identifier of the abnormal master node and the version information of the routing shadow table maintained by the master node that initiated the vote. Step S3: The master station node that receives the replacement voting request determines whether it has the ability to take over based on its own maintained routing shadow table and returns a voting response; the master station node that initiated the vote counts the voting responses, and when the number of votes agreeing to take over exceeds a preset ratio, the master station node with the most votes is determined as the replacement master station node; Step S4: The taking over master node generates an updated network routing table based on its own maintained network routing table and the routing information of the subordinate nodes of the abnormal master node in the routing shadow table, and broadcasts the routing table update to other master nodes; the subordinate nodes of the abnormal master node include base station nodes and slave nodes that are connected to the abnormal master node.

[0017] Further, step S3, which involves determining whether a node has the capability to take over based on its own maintained routing shadow table, includes: S311: The master node determines whether it has the capability to take over based on the coverage rate of routing entries for subordinate nodes of the abnormal master node in its own maintained routing shadow table; the routing entry coverage rate is determined based on the routing entries for subordinate nodes of the abnormal master node already recorded in the routing shadow table; S312: When the routing entry coverage rate meets the preset succession capability condition, the master node returns a vote response agreeing to take over to the master node that initiated the vote; S313: When the routing entry coverage rate does not meet the preset succession capability condition, the master node returns a vote response rejecting succession to the master node that initiated the vote, and carries the version information of its own routing shadow table and the missing routing entry information in the vote response, so that the master node that initiated the vote can coordinate with other master nodes to supplement the missing routing entries or select other candidate master nodes.

[0018] The above-described technical solutions of the embodiments of the present invention have the following beneficial technical effects: 1. By dividing the Ad-hoc network into a master station layer, a base station layer, and a slave station layer, the master station nodes in the master station layer form a decentralized, variable topology Ad-hoc network and undertake routing and addressing functions. They generate and maintain the entire network routing table by exchanging routing information. Base station nodes and slave nodes only connect to master station nodes or base station nodes and do not undertake routing and addressing functions. Furthermore, their network entry and exit status changes are only processed within their respective star connection ranges and do not trigger updates to the entire network routing table. This achieves the separation of static and dynamic routing information at the architectural level, strictly limiting routing update behavior that may cause network storms to a limited number of relatively stable master station nodes. This significantly reduces the probability of network storms and significantly improves the node carrying capacity of large-scale self-organized networking under narrowband and poor channel conditions. It solves the problem of slow network convergence speed and severely limited node capacity caused by frequent node entry and exit triggering of network-wide routing updates in conventional single-layer Ad-hoc networks. 2. By configuring device identification numbers for all node devices and dividing these identification numbers into non-overlapping master station number segments, base station number segments, and slave station number segments, each node automatically determines itself as a master station node, base station node, or slave station node based on the number segment to which its device identification number belongs. Using the same core hardware platform, different network roles can be assumed by writing device identification numbers from different number segments. This achieves hardware platform unification and software definition of network roles, eliminating the problems of numerous device types, complex system integration, and high construction and maintenance costs caused by different hardware designs for different network roles in traditional solutions. This enables the network to be quickly deployed and flexibly applied in scenarios such as emergency communication and remote areas. 3. By exchanging routing information containing a summary of the entire network routing table among master nodes and maintaining a routing shadow table for adjacent master nodes locally, when a master node detects an anomaly in an adjacent master node, it initiates a succession voting request. Other master nodes determine whether they have the capability to succeed based on the coverage of routing entries for subordinate nodes of the anomaly master node in their own maintained routing shadow tables. When the number of votes agreeing to succeed exceeds a preset proportion, the successor master node is determined. The successor master node generates an updated entire network routing table based on its own maintained entire network routing table and the routing information for subordinate nodes of the anomaly master node in the routing shadow tables, and broadcasts the updated routing table. This achieves distributed and rapid succession and smooth migration of routing tables when a master node fails in a decentralized network architecture, ensuring that the network can still maintain routing reachability to base station nodes and slave nodes under the anomaly master node when local nodes fail. This solves the problem of routing interruption and communication blind spots caused by sudden anomalies of master nodes in high-dynamic channel environments. Attached Figure Description

[0019] Figure 1This is a schematic diagram of the network topology provided in an embodiment of the present invention; Figure 2 This is a flowchart of the Ad-hoc decentralized variable topology networking method based on a hierarchical management architecture provided in this embodiment of the invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments and the accompanying drawings. It should be understood that these descriptions are merely exemplary and not intended to limit the scope of the invention. Furthermore, descriptions of well-known structures and techniques are omitted in the following description to avoid unnecessarily obscuring the concept of the invention.

[0021] Please refer to Figure 1 The first aspect of this invention provides a decentralized, variable-topology Ad-hoc network system based on a hierarchical management architecture, comprising a master station layer, a base station layer, and a slave station layer. In the master station layer, master station nodes constitute a decentralized, variable-topology Ad-hoc network, undertaking routing and addressing functions, and generating and maintaining a network-wide routing table by exchanging routing information. In the base station layer, base station nodes connect to master station nodes, do not undertake routing and addressing functions, and their network entry and exit status changes do not trigger updates to the network-wide routing table. In the slave station layer, slave station nodes connect to master station nodes or base station nodes, do not undertake routing and addressing functions, and their network entry and exit status changes do not trigger updates to the network-wide routing table.

[0022] This network architecture divides network nodes into three functional layers: master station layer, base station layer, and slave station layer. The master station layer consists of multiple master station nodes forming a decentralized, variable-topology ad-hoc network. Master station nodes exchange routing information and generate and maintain a network-wide routing table through self-organization, undertaking the routing and addressing functions of the entire network. Base station nodes in the base station layer connect to a master station node in a star topology to extend the signal coverage of that master station node, but they do not participate in routing and addressing themselves. Their network entry and exit status changes are only processed within the scope of the star topology they connect to and do not trigger updates to the network-wide routing table of the master station layer. Slave station nodes in the slave station layer can connect to master station nodes or base station nodes as user terminal equipment. Similarly, they do not undertake routing and addressing functions, and their network entry and exit status changes are also only processed within the scope of a local star topology and do not trigger updates to the network-wide routing table. In scenarios lacking basic network coverage, such as mountainous areas, forested areas, maritime locations, or disaster emergency situations, master nodes can be deployed at high points or command centers, base station nodes at relay locations, and numerous slave nodes as mobile terminals. Master nodes form a backbone grid network via wireless channels, while base stations and slave nodes connect in a star topology for proximity access, thus constructing a hierarchically managed wireless communication network. This architecture strictly limits the maintenance and updating of routing information to a limited and relatively stable master node layer, while isolating the state changes of the most numerous and frequently changing slave and base station nodes within local star topologies. This fundamentally avoids network storms caused by repeated updates to the entire network routing table due to frequent entry and exit of terminal nodes, significantly improving the stability and carrying capacity of large-scale self-organized networking under narrowband and harsh channel conditions.

[0023] In practical implementation, the ad-hoc network between master nodes can adopt a self-organizing networking protocol. Each master node discovers neighboring master nodes by sending probe signal frames and exchanges routing information based on channel quality and topology changes, forming a decentralized, variable-topology grid-like backbone network. Base station nodes connect to a master node through pre-configuration or automatic selection, establish a star connection, and forward downlink data from the master node and uplink data from slave nodes, thereby expanding the coverage of the master node. Slave nodes connect according to pre-specified master or base station, or automatically select accessible nodes to establish a star connection when roaming is enabled. Since the joining and leaving of base station nodes and slave nodes only involves updating the local access table of the connected master or base station and does not propagate routing update messages to the master layer backbone network, the network-wide routing table maintained by the master node only needs to reflect the reachability between master nodes, without needing to establish routing entries for each base station or slave. This separation of routing and non-routing nodes ensures the stability of the backbone routing even when a large number of terminals frequently go online and offline. It avoids the slow convergence speed and waste of channel resources caused by network-wide routing updates triggered by any node entering or leaving the network, which is common in traditional single-layer Ad-hoc networks. This provides a highly reliable and low-overhead networking solution for application scenarios that require large-scale terminal access and highly dynamic topology, such as emergency communication, remote IoT, and vehicle networking.

[0024] Specifically, master nodes, base station nodes, and slave nodes are all configured with device identification numbers. These device identification numbers are divided into non-overlapping master, base station, and slave number segments. Each node determines itself as a master, base station, or slave node based on the number segment to which its device identification number belongs. Upon power-on or network access, each node automatically determines its network role based on its device identification number's number segment, without manual intervention or additional role configuration procedures. Master nodes, base station nodes, and slave nodes can be implemented on the same core hardware platform. They can undertake different network functions simply by writing device identification numbers with different number segments, thus achieving software definition of network roles and hardware platform unification. This significantly reduces the types of equipment, lowers system integration complexity and construction and maintenance costs, and facilitates rapid deployment and flexible application in scenarios such as emergency communications and remote areas.

[0025] Specifically, in the master station layer, each master station node is configured to execute a self-organizing networking protocol. It discovers neighboring master station nodes by sending probe frames and establishes an Ad-hoc network with other master station nodes. Within the Ad-hoc network, routing information is exchanged to generate and maintain the network-wide routing table. After startup, master station nodes periodically send probe frames to discover neighboring master station nodes. These probe frames carry the sending node's device identification number and message type identifier, transmitted in a narrowband channel with a minimal frame structure. When a master station node receives a probe frame from another master station node, it confirms the existence of that neighboring master station node and records it in its local neighbor table. Through continuous interaction of probe frames, master station nodes gradually establish mutual discovery relationships, thus forming a decentralized, variable-topology Ad-hoc network composed of master station nodes. After the ad-hoc network is established, each master node encapsulates its information on neighboring master nodes, as well as the reachability information of other master nodes obtained through neighboring master nodes, into route update messages according to the rules of the self-organizing routing protocol, and propagates them throughout the network, gradually building and maintaining a network-wide routing table. When a master node goes offline due to channel changes or equipment failure, its neighboring master nodes determine that it is offline after the probe signal frame reception times out and trigger the routing table update process, deleting the routing entry of the offline node and notifying other master nodes. When a new master node joins or an existing master node resumes communication, the same mechanism injects the routing information of the new node into the network-wide routing table, thereby enabling the master layer backbone network to have the ability to self-organize and adapt to topology changes, providing highly reliable backbone transmission support for the entire three-layer network architecture.

[0026] Specifically, the network-wide routing table only contains the address information of other master nodes in the master station layer, excluding the address information of base station nodes and slave nodes. Master nodes maintain local access tables for their subordinate base station nodes and slave nodes; these local access tables are not included in the network-wide routing table. After a base station node and a slave node connect to a master node in a star topology, their address information is only recorded in the local access table of the master node they connect to, used for local forwarding and terminal management within the star topology range. When a slave node needs to communicate with another slave node, it first uses a probe to determine if it is directly reachable. If directly reachable, it skips the addressing process and completes direct communication between nodes; otherwise, it sends the data to the master node it connects to. The master node then queries the network-wide routing table to determine the target master node to which the destination slave node belongs, forwards the data along the master station layer Ad-hoc network to the target master node, and then the target master node delivers the data to the destination slave node according to its own local access table. Since the network-wide routing table only reflects reachability between master nodes, its number of entries is only related to the number of master nodes and does not increase with the number of base station nodes and slave nodes. Therefore, the storage space overhead and routing computation complexity of the routing table do not increase significantly with the expansion of the network scale. At the same time, the entry and exit of base station nodes and slave nodes only cause updates to the local access tables of the master nodes they connect to, without triggering updates to the network-wide routing table. This isolates the state changes of a large number of terminal nodes within a local star topology, avoiding network-wide routing updates and network storms caused by frequent entry and exit of terminals. This significantly improves the stability and carrying capacity of large-scale node networks under narrowband and harsh channel conditions.

[0027] Specifically, base station nodes are configured to send network access requests to pre-designated master nodes and establish star connections to extend the signal coverage of the pre-designated master nodes. Network access and disconnection status changes of base station nodes are processed only within the star connection range and do not trigger updates to the entire network routing table. When a base station node detects an anomaly in a master node it has accessed, it scans for and accesses other master nodes. Upon receiving network access confirmation from a master node, the base station node forwards downlink data from the accessed master node and uplink data from subordinate slave nodes within the star connection range, thereby extending the signal coverage of the master node to areas that were previously not directly covered, allowing slave nodes outside the master node's coverage area to indirectly access the network through the base station node. Network access and disconnection status changes of base station nodes are local access events, involving online status information exchange and local access table updates only within the star connection range between the base station node and the master node it has accessed. They do not involve routing information exchange in the master-level ad-hoc network, nor do they trigger updates to the entire network routing table between master nodes. When a base station node detects that the master station node it is connected to is abnormal due to channel degradation or equipment failure, the base station node actively scans for other master station nodes that can be connected in the vicinity and sends network access requests to the scanned master station nodes to establish a new star connection. This completes the reconstruction of the access relationship in a local area, ensuring the communication continuity of the base station node and its subordinate slave nodes. At the same time, this change in access relationship does not trigger an update of the entire network routing table, thus avoiding the impact on the stability of the master station layer backbone routing.

[0028] Specifically, slave nodes are configured to send network access requests to a pre-specified master node or a pre-specified base station node and establish a star connection by default. When roaming is enabled and the pre-specified master node or pre-specified base station node is unreachable, an accessible master node or base station node is automatically selected and a star connection is established. Changes in the network access and disconnection status of slave nodes are processed only within the star connection and do not trigger updates to the entire network routing table. The default access method allows network planners to specify a suitable access point for each slave node based on coverage requirements and service distribution, thereby ensuring the determinism and controllability of the initial network deployment. When roaming is enabled on a slave node and the pre-specified master node or base station node is unreachable due to channel degradation, equipment failure, or being outside the coverage area, the slave node automatically scans for accessible master nodes or base station nodes in the vicinity and selects one with signal quality that meets the access conditions to establish a star connection. This ensures that communication can still be maintained when the specified access point is unavailable, improving the connectivity and communication continuity of terminals in complex geographical environments and under highly dynamic channel conditions. The changes in the network entry and exit status of a slave node are processed only within the star connection between itself and the master or base station node it connects to. The master or base station node updates the record of the slave node in its local access table. This does not propagate to the master layer Ad-hoc network, nor does it trigger an update of the network-wide routing table between master nodes. In this way, the status changes of the most numerous and frequently changing slave nodes are isolated within the local access range, avoiding network-wide routing updates and network storms caused by frequent online and offline of a large number of terminals. This significantly improves the stability and network carrying capacity of large-scale terminal access under narrowband and poor channel conditions.

[0029] Specifically, the master node, base station node, and slave node all utilize the same core hardware platform and activate different operating modes by writing device identification numbers from different number segments, thereby assuming the network roles of master, base station, and slave, respectively. The three network roles are not differentiated in terms of hardware architecture; they all possess the same RF front-end, baseband processing, network control and processing units, and interface circuits. They can assume the network roles of master, base station, and slave simply by writing device identification numbers from different number segments. In practical implementation, before leaving the factory or being deployed, the node device uses a configuration tool to burn the device identification number belonging to the master, base station, or slave number segment into its internal non-volatile storage medium. After power-on, the node device reads this device identification number and determines its corresponding network function, thereby executing the appropriate network protocol and access procedure without requiring the replacement of hardware modules or redesign of circuits. This approach, which combines hardware platform unification with network role software definition, allows the same device to be flexibly configured as a master node, base station node, or slave node according to actual deployment needs. This effectively reduces the types of equipment and spare parts, lowers system integration complexity, manufacturing costs, and the difficulty of on-site construction and subsequent operation and maintenance, and facilitates rapid deployment and flexible expansion in scenarios such as emergency communication and remote areas.

[0030] Specifically, base station nodes and slave nodes do not participate in the exchange of routing information between master nodes; the exchange of routing information between master nodes is only performed by the master nodes in the master layer. In the entire network architecture, the generation, updating, and dissemination of routing information are strictly limited to the master layer. Base station nodes and slave nodes only connect to master nodes or base station nodes through a star topology, and complete data forwarding and terminal management within their local area of ​​access, without participating in routing calculations and routing table maintenance in the master layer Ad-hoc network. With this design, the scope and number of routing update messages exchanged between master nodes are only related to the number of master nodes, and are not affected by the number or status changes of base station nodes and slave nodes. This concentrates routing control overhead in a limited and relatively stable master layer, avoiding the spread of routing update messages throughout the network caused by frequent online / offline cycles of numerous terminal nodes. It also reduces competition and collisions of control messages in narrowband channels, lowers the routing processing burden on each node, improves the utilization efficiency of channel and network resources, and provides a guarantee for stable networking of large-scale nodes under harsh channel conditions.

[0031] Based on the aforementioned three-layer network architecture, the master station nodes in the master station layer can employ general network control and processing technologies, enabling the master station layer Ad-hoc network to simultaneously support three transmission modes: meteor trail communication, ionospheric scattering communication, and ultra-shortwave ground wave over-the-line communication. Since these three types of channels share common characteristics such as poor channel stability and weak signal quality, the master station nodes can flexibly select or combine the use of meteor trail, ionospheric scattering, and ground wave over-the-line transmission based on channel conditions through a unified signal waveform and network control processing mechanism. This extends the coverage distance and connectivity reliability of the master station layer backbone network in complex geographical environments such as mountainous and forested areas. Base station nodes and slave nodes, after connecting to the master station node or base station node, can also adapt to the aforementioned multi-mode transmission methods, ensuring the entire hierarchical network maintains a high communication rate even under narrowband and harsh channel conditions.

[0032] Furthermore, the master node, base station node, and slave node can all scan their directly communicating neighboring nodes, generate and maintain a neighboring node table. When the destination address of the data to be sent exists in the neighboring node table, the node device can directly send the data to the corresponding neighboring node without going through the master node for routing and forwarding. This neighboring node pass-through mechanism, in the three-layer architecture, serves as a supplement to the master layer's routing and forwarding, enabling geographically adjacent nodes to transmit data using an efficient direct communication path. Even if the master node partially loses connection due to channel degradation, neighboring nodes can still maintain basic communication capabilities, thereby improving user communication rates under complex, highly dynamic, and adverse channel conditions, and reducing network and channel resource consumption.

[0033] The master node, base station node, and slave nodes can all utilize the same core hardware platform. The network control and processing units within this platform can be implemented using a CPU control module and an FPGA control module. The CPU control module is responsible for communication network protocol processing, including control command parsing and network parameter setting, probe signal frame generation and transmission, routing table generation and management, data frame packetization, transmission queue management, data frame parsing, message recovery and receive queue management, transmit / receive control, and routing processing. The FPGA control module is responsible for interface adaptation with other units, connecting to the data receiving unit, frequency synthesis unit, and power amplifier unit via interface adapter control circuits to achieve functions such as frequency setting and transmit / receive control. The CPU control module and FPGA control module interact via serial port data, transmit / receive clock signals, transmit / receive data signals, parameter configuration signals, and transmit / receive control signals. The FPGA control module can also connect to a handle and operating terminal via an interface conversion circuit to achieve human-machine interaction and parameter configuration. This hardware implementation provides a unified hardware platform to support the functional implementation of each node role in the three-layer network architecture, enabling master nodes, base station nodes, and slave nodes to assume their respective network roles simply by writing device identification numbers of different number segments.

[0034] Under the aforementioned three-layer network architecture, the operation of each node device can include processes such as initialization, parameter setting, channel state detection, transmitting probe signals, neighbor node scanning, establishing routing tables, data reception and address resolution, packet recovery, data transmission and packaging, transmission queue management, channel state detection, and backoff / retransmission. The master node completes self-organizing networking and the generation and maintenance of the entire network routing table through the above processes, while base station nodes and slave nodes complete network access to the master node or base station node and local data forwarding through the same processes. This workflow is applicable to long-range beyond-line-of-sight (BLS) networking communication in complex geographical environments such as mountainous and forested areas; emergency communication in major disaster environments such as earthquakes and floods; communication and information access in areas without basic information network coverage, such as at sea; long-range beyond-line-of-sight information transmission for disaster monitoring; long-range IoT information transmission for industrial field control; and beyond-line-of-sight vehicle-to-everything (V2X) and ship-to-everything (S2X) information interconnection scenarios. It provides a complete system implementation foundation for large-scale node self-organizing networking under narrowband and harsh wireless channels.

[0035] Please refer to Figure 2 The second aspect of this invention provides an Ad-hoc decentralized variable topology networking method based on a hierarchical management architecture, applied to the above-mentioned networking system, executed by the master station node in the networking system, including the following steps.

[0036] S1: During operation, the master node exchanges routing information with neighboring master nodes. This routing information includes a summary of the entire network routing table. Based on the received summary information, the master node maintains a routing shadow table for neighboring master nodes locally. The routing shadow table contains at least a portion of the entries in the entire network routing table maintained by the neighboring master nodes. Each routing entry includes the destination master node identifier, next hop, hop count, and version number.

[0037] In scenarios such as long-distance communication in mountainous areas or information access at sea, master station nodes are interconnected through an ad-hoc network. Each master station node maintains its own full network routing table and periodically exchanges routing information with neighboring master station nodes. The exchanged routing information is not the complete full network routing table, but rather a simplified version containing a summary of the full network routing table. This summary includes at least the identifier of each destination master station node, the next hop to that destination master station node, the hop count, and the version number of the routing entry. Upon receiving the summary information from neighboring master station nodes, each master station node maintains a routing shadow table locally for each neighboring master station node. This routing shadow table is essentially a partial copy or cache of the full network routing table maintained by the neighboring master station node, and its content is updated with the received summary information. Through this summary exchange and shadow table maintenance method, each master station node can locally grasp the general routing information known by neighboring master station nodes without increasing channel overhead, providing a basis for quickly determining the successor master station node in case of failure. Since the exchange of routing information between master nodes only involves summary information rather than complete routing tables, and the shadow table only stores at least a portion of the entries in the routing tables of adjacent master nodes, it does not significantly increase the routing control overhead of the master layer ad-hoc network, nor does it change the basic architecture that the entire network routing table is maintained only by the master nodes.

[0038] S2: When a master node detects an abnormal neighboring master node, it initiates a succession voting request to other master nodes. The succession voting request includes the identifier of the abnormal master node and the version information of the routing shadow table maintained by the master node that initiated the vote.

[0039] In complex geographical environments and under highly dynamic channel conditions, master nodes may malfunction due to channel obstruction, equipment failure, or energy depletion, rendering them unable to continue performing routing functions. When a master node detects an anomaly in a neighboring master node through probe signal frame exchanges or route summary information updates, it initiates a succession vote request to other master nodes in the master-level Ad-hoc network to begin the succession process for the malfunctioning master node. The succession vote request includes the identifier of the malfunctioning master node, indicating the target to be replaced; it also includes the version information of the routing shadow table maintained by the initiating master node, indicating to other master nodes the current uptime of the routing information about the malfunctioning master node's subordinate nodes. Upon receiving the succession vote request, other master nodes can use this information to determine whether the initiator is qualified and has the necessary conditions to initiate a succession vote, avoiding insufficient basis for succession decisions due to an outdated routing shadow table. This method, initiating voting requests proactively by the master node that detects an anomaly, does not rely on any central control node and adapts to the self-organizing characteristics of decentralized Ad-hoc networks.

[0040] S3: Upon receiving the succession voting request, the master node determines whether it has the capability to succeed the other node based on its own maintained routing shadow table and returns a voting response. The master node that initiated the vote tallies the voting responses, and when the number of votes agreeing to succeed exceeds a preset proportion, the master node with the most votes is determined as the successor master node.

[0041] Upon receiving the succession vote request, other master nodes each determine their own capability to succeed the abnormal master node based on their locally maintained routing shadow tables. This determination is based on whether the master node's routing shadow table contains sufficient routing information about the abnormal master node's subordinate nodes. If its routing shadow table records a large number of routing entries for the abnormal master node's subordinate nodes, it indicates that the master node can quickly restore its routing forwarding capabilities to the abnormal master node's subordinate base station nodes and slave nodes after succession, and therefore can return a vote agreeing to succeed. If its routing shadow table contains little routing information about the abnormal master node's subordinate nodes, it indicates that it does not yet meet the succession conditions and can return a vote rejecting succession. The master node that initiated the vote tallies the vote responses received from each master node. When the number of votes agreeing to succeed exceeds a preset proportion, it indicates that a majority of master nodes in the master layer recognize the conditions for initiating the succession process. At this point, the master node with the highest number of votes among those agreeing to succeed is selected as the successor master node. The successor node is determined by voting, which ensures that the succession decision is based on the joint judgment of multiple master nodes. This avoids the risk of misjudgment that may be caused by a single node designating the successor, and also adapts to the characteristic of a decentralized network where there is no centralized control node.

[0042] S4: The succeeding master node generates an updated network routing table based on its own maintained network-wide routing table and routing shadow table containing routing information about the subordinate nodes of the abnormal master node, and broadcasts the routing table update to other master nodes. Subordinate nodes of the abnormal master node include base station nodes and slave nodes connected to the abnormal master node.

[0043] Once the replacement master node is determined, it merges its own maintained network-wide routing table with the routing information of the abnormal master node's subordinate nodes in the routing shadow table to generate an updated network-wide routing table. Since the abnormal master node's subordinate nodes include base station nodes and slave nodes connected to it, the replacement master node, using the relevant routing information already in the routing shadow table, can continue to provide routing reachability to these subordinate nodes after taking over. This allows service data originally forwarded by the abnormal master node to be forwarded by the replacement master node instead, thus maintaining network communication services for the abnormal master node's subordinate base station nodes and slave nodes when the master node fails. After generating the updated network-wide routing table, the replacement master node broadcasts the updated routing table to other master nodes in the master-level Ad-hoc network, causing them to synchronously update their network-wide routing tables, thereby completing the reconvergence of routing information within the master-level layer. The above process, without relying on any central control node, achieves distributed and rapid succession when the master station node fails through digest exchange, shadow table maintenance, succession voting, and routing table merging among master station nodes. This ensures that the master station layer Ad-hoc network can still maintain routing reachability to subordinate base station nodes and slave nodes when local nodes fail, and improves the routing stability and communication continuity of the decentralized variable topology network under complex geographical environments and highly dynamic channel conditions.

[0044] Steps S1 to S4 above exchange routing information containing a summary of the entire network routing table among master nodes and maintain routing shadow tables of neighboring master nodes locally. This allows each master node to grasp the general routing information known by neighboring master nodes without relying on a central control node. When a master node detects an anomaly in a neighboring master node, the detected master node initiates a succession voting request. Other master nodes determine their succession capability based on their own maintained routing shadow tables and return voting responses. When the number of votes in favor of succession exceeds a preset proportion, the master node with the most votes is determined as the successor master node. The successor master node further generates an updated entire network routing table based on its own maintained entire network routing table and routing information about the subordinate nodes of the anomaly master node in the routing shadow tables, and broadcasts the updated routing table to other master nodes. This ensures that the base station nodes and slave nodes under the anomaly master node remain route-reachable after the succession is completed. This process organically combines the exchange of routing information summaries, shadow table maintenance, succession voting, and routing table merging. In a decentralized, variable-topology Ad-hoc network, it achieves distributed rapid succession and smooth migration of routing information when the master node fails. This avoids routing interruptions and communication blind spots caused by sudden failures of the master node, and improves the stability and communication continuity of large-scale node self-organizing networks under narrowband and harsh channel conditions.

[0045] Specifically, in this networking method, the following steps are performed before step S1.

[0046] S101: The master node maintains a routing table change log locally, recording the event type, occurrence time, and impact range of adding, deleting, and modifying routing entries in the entire network routing table.

[0047] In scenarios such as remote communication in mountainous areas and disaster emergency communication, the topology of the master station layer ad-hoc network changes frequently due to node movement, channel variations, and equipment power-on / off. This leads to continuous additions, deletions, and modifications of routing entries in the network-wide routing tables maintained by each master station node. To understand the patterns of routing table changes and optimize the timing of route updates, each master station node maintains a local routing table change log. Whenever a routing entry in the network-wide routing table changes, the master station node records the event type, occurrence time, and the range of master station nodes affected by the change in the log. The change log can be stored using a circular buffer or by appending in chronological order, retaining only changes from the most recent period to avoid excessive storage space consumption. By maintaining this change log, master station nodes can backtrack and statistically analyze the frequency, type, and impact range of routing table changes over a period of time, providing a data foundation for subsequent calculations of routing table change entropy. It also helps provide historical information on routing table changes to the succeeding master station node when the master station fails.

[0048] S102: The master station node counts the number distribution of various change events according to a preset time window, and calculates the routing table change entropy based on the number distribution. The routing table change entropy characterizes the degree of disorder in the number distribution of various change events within the preset time window.

[0049] The master node statistically analyzes various change events recorded in the routing table change log within a preset time window, obtaining the distribution of the number of change events such as route entry additions, deletions, and modifications within that time window. Based on this, the master node calculates the routing table change entropy according to the distribution of the number of various change events. This entropy value characterizes the degree of disorder in the distribution of the number of various change events within the preset time window. When routing table changes are mainly caused by a single type of event, the distribution of the number of change events is relatively concentrated, and the routing table change entropy is low, indicating that the current routing table changes are highly regular and predictable. When routing table changes are caused by multiple types of events simultaneously and dispersedly, the distribution of the number of change events is relatively dispersed, and the routing table change entropy is high, indicating that the current routing table changes are more drastic and chaotic, possibly accompanied by rapid changes in network topology. By calculating the routing table change entropy, the master node can measure the drasticness and regularity of the current routing table changes with a comprehensive indicator, rather than relying solely on the number of a single type of change event to determine the route update sending strategy.

[0050] S103: The master node compares the routing table change entropy with a preset entropy threshold. If the routing table change entropy is lower than the preset entropy threshold, it enters a routing update suppression state, merging or delaying the routing update messages to be sent to the next suppression period. If the routing table change entropy is higher than or equal to the preset entropy threshold, it immediately sends a routing update message to the adjacent master node.

[0051] After calculating the routing table change entropy within the current preset time window, the master node compares it with a preset entropy threshold and determines the routing update message sending strategy based on the comparison result. When the routing table change entropy is lower than the preset entropy threshold, it indicates that the current routing table changes are relatively gradual or have strong regularity. In this case, the master node enters a routing update suppression state, merging or delaying the routing update messages to be sent to the next suppression period to reduce the number of unnecessary routing update messages during periods of relative topology stability and reduce control message overhead in narrowband channels. When the routing table change entropy is higher than or equal to the preset entropy threshold, it indicates that the current routing table changes drastically, potentially indicating rapid topology changes or inconsistent routing information. In this case, the master node immediately sends routing update messages to neighboring master nodes so that they can update their full network routing tables and routing shadow tables in a timely manner, ensuring rapid convergence of routing information in the master-level Ad-hoc network. By using the routing table change entropy as the criterion for whether to suppress routing updates, the master node can suppress unnecessary routing updates during periods of topology stability and promptly propagate routing updates during periods of drastic topology changes, thereby achieving a balance between reducing routing overhead and ensuring routing convergence speed.

[0052] The above steps S101 to S103 maintain routing table change logs, calculate routing table change entropy, and dynamically adjust the sending strategy of routing update messages accordingly. This enables the master station node to adaptively control the routing update frequency based on the severity of routing table changes. When the topology is relatively stable, the number of routing update messages is reduced to save narrowband channel resources. When the topology changes drastically, routing updates are sent in a timely manner to ensure rapid convergence of routing information. Thus, without relying on a central control node, the efficiency and adaptability of routing control in the master station layer Ad-hoc network under complex geographical environments and highly dynamic channel conditions are improved.

[0053] Further, step S102 calculates the routing table change entropy based on the quantity distribution, including the following sub-steps.

[0054] S1021: For each type of change event, the main station node counts the number of times that change event type occurs within a preset time window, as well as the number of main station nodes affected by that change event type.

[0055] In scenarios such as remote communication in mountainous areas or disaster emergency communication, changes in the routing table of the master station layer ad-hoc network may be caused by various types of events. These include the addition of a new master station node leading to an increase in routing entries, the disconnection of a master station node causing the deletion of routing entries, and changes in link quality leading to modifications of the next hop or hop count. Within a preset time window, the master station node performs statistical analysis for each type of change event. This involves counting the number of times each type of event occurs within the time window to reflect its frequency, and also counting the number of master station nodes affected by each type of event to reflect its impact range in the network. For example, a routing entry deletion may only affect one adjacent master station node, while a critical master station node disconnection may lead to the simultaneous modification of routing entries for multiple master station nodes, resulting in significantly different impact ranges. By separately analyzing the frequency and impact range of each type of event, the master station node can provide multi-dimensional data for subsequent calculations of routing table change entropy, ensuring that the entropy value reflects not only the frequency of events but also the actual impact of events on the network topology.

[0056] S1022: The main station node calculates the impact range weight of this change event type. The impact range weight is positively correlated with the number of main station nodes affected.

[0057] To ensure that the routing table change entropy accurately reflects the varying impacts of different event types on the network, master nodes calculate an impact range weight for each event type. This weight is positively correlated with the number of master nodes affected by that event type; the more master nodes affected by a particular event type, the greater its impact range weight. For example, a route entry modification affecting only one adjacent master node has a relatively small impact range weight, while a route entry deletion or addition affecting multiple master nodes has a relatively large impact range weight. The specific calculation method for the impact range weight can be flexibly set according to network size and deployment scenario. For instance, it can directly use the number of affected master nodes as the weight, or it can normalize the number of affected master nodes before using it as the weight. By introducing the impact range weight, master nodes can ensure that events with a significant impact on the network occupy a larger proportion of the entropy value when subsequently calculating weighted probabilities and routing table change entropy, thus allowing the entropy value to more accurately reflect the actual severity of routing table changes.

[0058] S1023: The main station node calculates the weighted probability of each change event type based on the number of times and the weight of the scope of influence of each change event type.

[0059] After obtaining the occurrence count and impact range weights of each change event type, the master node further calculates the weighted probability of each change event type. Specifically, for each change event type, its occurrence count is multiplied by its corresponding impact range weight to obtain the weighted count of that type of event; then, the weighted count of that type of event is divided by the sum of the weighted counts of all types of events to obtain the weighted probability of that type of event. The weighted probability reflects the relative proportion of a certain type of event among all routing table change events, considering differences in impact range. In this way, events with fewer occurrences but larger impact ranges may have a higher weighted probability than events with more occurrences but smaller impact ranges, thus making the subsequently calculated routing table change entropy more representative of change events that have a greater impact on network stability. The calculated weighted probability is a set of normalized values, with the sum of the weighted probabilities of each type of event being one, facilitating the subsequent calculation of the routing table change entropy using the information entropy formula.

[0060] S1024: The master node calculates the routing table change entropy using the information entropy formula and weighted probability as input.

[0061] After obtaining the weighted probabilities of each type of change event, the master node calculates the routing table change entropy using the information entropy formula, with each weighted probability as input. Information entropy is an indicator in information theory used to measure the degree of uncertainty or disorder of random variables. When the weighted probability distribution of each type of event is relatively concentrated, the information entropy is low; when the weighted probability distribution of each type of event is relatively dispersed, the information entropy is high. Therefore, the calculated routing table change entropy can comprehensively reflect the degree of disorder in the quantity distribution of various routing table change events within a preset time window. If the routing table change is mainly caused by a concentrated set of events of a certain type, the weighted probability distribution is concentrated, and the routing table change entropy is low, indicating that the current routing table change has strong regularity; if the routing table change is caused by multiple types of events simultaneously and dispersedly, the weighted probability distribution is dispersed, and the routing table change entropy is high, indicating that the current routing table change is relatively drastic and chaotic. Through S1021 to S1024 above, the master station node calculates the weighted probability based on the frequency and impact range weight of each type of change event, and obtains the routing table change entropy according to the information entropy formula. This entropy value reflects both the frequency of change events and the impact range of change events on the network topology, thus providing an accurate and targeted basis for subsequent dynamic adjustment of the routing update message sending strategy based on the routing table change entropy. This improves the adaptability and efficiency of routing control in the master station layer Ad-hoc network under complex geographical environments and highly dynamic channel conditions.

[0062] Furthermore, in step S103, the route update suppression state is entered, and the route update messages to be sent are merged or delayed to the next suppression period, including the following sub-steps.

[0063] S1031: After entering the route update suppression state, the master node will group the route update messages to be sent according to the destination master node.

[0064] In scenarios such as remote communication in mountainous areas or disaster emergency communication, a master station node may generate multiple routing update messages to be sent simultaneously during a period of relatively stable topology. These messages may have the same or different destination master station nodes. When the master station node enters a routing update suppression state, it first groups the routing update messages to be sent according to their destination master station nodes. That is, routing update messages destined for the same destination master station node are grouped together, and routing update messages destined for different destination master station nodes are grouped into different groups. For example, a master station node may need to send multiple routing update messages to master station node A due to changes in multiple routing entries, and simultaneously need to send multiple routing update messages to master station node B due to changes in other routing entries. In this case, two groups of messages to be sent are formed according to the two destination master station nodes A and B respectively. By grouping according to the destination master station node, the master station node can merge multiple update messages to the same destination in subsequent processing, avoiding the repeated transmission of redundant routing information for the same destination master station during the suppression period, thereby reducing the number of control messages in the narrowband channel.

[0065] S1032: For each destination master node, the master node retains only the latest routing update message corresponding to that destination master node and discards the historical routing update messages of that destination master node.

[0066] After grouping by destination master node, the master node filters the routing update messages within each group, retaining only the latest routing update message corresponding to that destination master node and discarding historical routing update messages for that destination master node. Since routing update messages typically carry the latest state of routing entries, routing information contained in earlier generated routing update messages may have been overwritten by updates in subsequent messages. Continuing to send historical messages would not only consume additional channel resources but could also cause temporary inconsistencies in routing information at the receiver. By retaining only the latest routing update messages, the master node can transmit the latest routing change information to each destination master node with the fewest possible messages during the suppression period, allowing the receiver to obtain the complete update content at once after the suppression period ends. This reduces control overhead while ensuring the correctness and timeliness of routing updates.

[0067] S1033: The master node dynamically adjusts the duration of the suppression period based on the difference between the change entropy of the routing table and the preset entropy threshold. The larger the difference, the shorter the suppression period; the smaller the difference, the longer the suppression period.

[0068] After entering the route update suppression state, the master node does not use a fixed suppression period, but dynamically adjusts the duration of the suppression period based on the difference between the routing table change entropy and the preset entropy threshold. When the routing table change entropy is higher than the preset entropy threshold, the master node does not enter the suppression state, but immediately sends a route update message; when the routing table change entropy is lower than the preset entropy threshold, the master node enters the suppression state. At this time, the larger the difference, the lower the current routing table change entropy is relative to the threshold, that is, the smoother and more regular the routing table change. The master node can set the suppression period to be longer to further reduce the frequency of sending route update messages; the smaller the difference, the closer the current routing table change entropy is to the threshold. That is, although the routing table change is temporarily lower than the threshold, it may still rebound relatively quickly. The master node should set the suppression period to be shorter so that it can quickly exit the suppression state and send a route update message when the routing table change intensifies. By using this dynamic adjustment method based on entropy difference, the master node can extend the suppression period when the routing table changes slowly to save channel resources to the greatest extent, and shorten the suppression period when the routing table changes are close to the threshold to maintain the ability to respond to topology changes, thereby achieving a dynamic balance between suppressing route updates and ensuring route convergence.

[0069] S1034: When the suppression period arrives, the master node will send the latest routing update messages of each destination master node to the adjacent master node.

[0070] When the suppression period arrives, the master node sends the latest route update messages from all destination master nodes (previously merged) to the corresponding neighboring master nodes. Since the messages have been grouped by destination master node and only the latest messages are retained during the suppression period, the number of messages sent is reduced. Each message carries the latest route change information required by the corresponding destination master node. Upon receiving these messages, neighboring master nodes can update their entire network routing table and routing shadow table at once, without having to frequently process scattered route update messages during the suppression period. Through S1031 to S1034, the master node, under the route update suppression state, groups messages by destination master node, retains only the latest messages, dynamically adjusts the suppression period based on the entropy difference, and sends them all at once when the period arrives. This effectively reduces and delays the transmission of route update messages during the topology stability period and allows for timely recovery when topology changes intensify. Thus, without relying on a central control node, it reduces the probability of contention and collisions of route control messages in narrowband channels, improving the efficiency and stability of route control in master-level Ad-hoc networks under complex geographical environments and highly dynamic channel conditions.

[0071] Specifically, step S3 determines whether the system has the capability to take over based on its own maintained routing shadow table, including the following sub-steps.

[0072] S311: The master node determines its own takeover capability based on the coverage rate of route entries for subordinate nodes of the abnormal master node in its maintained route shadow table. The route entry coverage rate is determined based on the route entries for subordinate nodes of the abnormal master node already recorded in the route shadow table.

[0073] In scenarios such as remote communication in mountainous areas or disaster emergency communication, when a master station node malfunctions and needs to be replaced by another master station node, the ability of the replacing master station node to quickly restore routing forwarding capabilities to the malfunctioning master station's subordinate base station nodes and slave nodes depends on the amount of relevant routing information it possessed before the malfunction occurred through routing summary information exchange. When a master station node receiving a replacement vote request determines its own replacement capability, it uses the coverage rate of routing entries related to the malfunctioning master station node's subordinate nodes in its maintained routing shadow table as a basis. This coverage rate is determined based on the routing entries already recorded in the routing shadow table related to the malfunctioning master station node's subordinate nodes and is used to measure the master station node's mastery of the routing information of the malfunctioning master station's subordinate nodes. The more routing entries related to the malfunctioning master station's subordinate nodes in the routing shadow table, the faster the master station node can restore routing forwarding capabilities to the malfunctioning master station's subordinate base station nodes and slave nodes after replacement; the fewer routing entries, the longer it may take for the master station node to replenish the relevant routing information after replacement. By using routing entry coverage as the criterion for judging succession capability, the succession decision is based on the candidate master node's actual grasp of the routing information of the subordinate nodes of the abnormal master station, rather than solely on the node's own status indicators. This makes the selection of the successor master station node more in line with the actual needs of the succession task.

[0074] S312: When the coverage of the routing entries meets the preset succession capability conditions, the master node returns a vote response agreeing to the succession to the master node that initiated the vote.

[0075] When a primary node determines it has the capability to take over based on the coverage of routing entries for subordinate nodes of the abnormal primary node in its routing shadow table, it returns a vote of agreement to take over to the primary node that initiated the vote. This indicates that the primary node is willing and able to assume the task of taking over the abnormal primary node. Preset succession capability conditions can be set according to network size and actual deployment needs; for example, coverage exceeding a certain threshold may be considered as having succession capability. By returning a vote of agreement to take over, the primary node becomes one of the candidate primary nodes to take over, and the primary node that initiated the vote participates in the vote comparison with other candidate primary nodes in subsequent statistics.

[0076] S313: When the coverage of routing entries does not meet the preset succession capability conditions, the master node returns a vote response to the master node that initiated the vote, rejecting the succession. The vote response rejecting the succession carries the version information of its own routing shadow table and the information of missing routing entries, so that the master node that initiated the vote can coordinate with other master nodes to supplement the missing routing entries or select other candidate master nodes.

[0077] When a primary node determines it lacks the capability to take over based on the coverage of route entries for subordinate nodes of an abnormal primary node in its routing shadow table, it doesn't simply refuse to take over. Instead, it sends a rejection vote to the primary node that initiated the vote, along with its own routing shadow table version information and information on missing route entries. The version information indicates the timeliness of the routing information currently held by the primary node, while the missing route entry information indicates which route entries for abnormal primary node subordinate nodes the primary node has not yet acquired. Upon receiving such a rejection response with missing information, the primary node that initiated the vote can request route completion from other primary nodes based on the missing route entry information. The primary node holding the corresponding route entries will then add the missing route entries to the primary node's routing shadow table, enabling it to regain the capability to take over. If the replacement conditions cannot be met through coordinated completion, the primary node that initiated the vote can also select other candidate primary nodes. Through S311 to S313 above, the master node uses the coverage rate of routing entries for subordinate nodes of the abnormal master node in the routing shadow table as the basis for judging the succession capability. When the conditions are met, it returns a vote to agree to the succession; when the conditions are not met, it returns a vote to refuse the succession, carrying version information and information on missing routing entries. This makes the judgment of succession capability and the selection of candidate master nodes based on the actual master node's grasp of the routing information of subordinate nodes of the abnormal master node, and provides a basis for coordinating and supplementing routing information among master nodes. This improves the accuracy of the succession decision when the master node is abnormal in the decentralized Ad-hoc network and the completeness of the route recovery after the succession.

[0078] Furthermore, the routing entry coverage in step S311 is determined in the following manner.

[0079] S3111: Each time a master node receives a routing summary message from a neighboring master node, it records the timestamp of the message.

[0080] In scenarios such as remote communication in mountainous areas or disaster emergency communication, master station nodes periodically exchange route summary information to maintain their respective route shadow tables. Whenever a master station node receives route summary information from a neighboring master station node, it updates the route entries in the summary information to its local route shadow table and records the corresponding reception timestamp. The reception timestamp marks the arrival time of the summary information and serves as a basis for subsequently determining the freshness of each route entry in the route shadow table. Because the topology and channel conditions of the master station layer ad-hoc network are constantly changing, the route status reflected by route summary information received at different times may differ. Route entries contained in earlier received summary information may be outdated or updated by subsequent summary information, while later received summary information is closer to the actual route status at the current moment. By recording timestamps for each received summary information, master station nodes can distinguish the timeliness of each route entry in the route shadow table, providing basic data for subsequent calculations of freshness weights and route entry coverage.

[0081] S3112: The master node calculates the freshness weight of each route entry in the route shadow table based on the received timestamp. The closer the received timestamp is to the current time, the greater its freshness weight.

[0082] After obtaining the receive timestamps for each route entry in the routing shadow table, the master node calculates the freshness weight of each route entry based on the time interval between the receive timestamp and the current time. A shorter time interval (i.e., the closer the receive timestamp is to the current time) indicates that the routing state reflected by the route entry is newer, and its freshness weight is greater. Conversely, a longer time interval (i.e., the farther the receive timestamp is from the current time) indicates that the routing state reflected by the route entry may be outdated, and its freshness weight is smaller. The specific calculation method for the freshness weight can be flexibly set according to the network scale and deployment scenario. For example, it can use the reciprocal of the time interval or assign values ​​in segments according to the time interval, as long as the principle that the closer the receive timestamp is to the current time, the greater the freshness weight is. By introducing freshness weights, when the master node subsequently evaluates route entries in the routing shadow table related to subordinate nodes of an abnormal master node, newer route entries can occupy a larger proportion in the coverage calculation, while older route entries have a smaller impact. This avoids overestimating the master node's understanding of routing information due to the existence of a few outdated route entries.

[0083] S3113: The master node performs a weighted summation of the routing entries for subordinate nodes of the abnormal master node in the routing shadow table according to the freshness weight of each routing entry, and obtains the weighted number of routing entries.

[0084] When calculating route entry coverage, the master node does not simply count the number of route entries about subordinate nodes of the abnormal master node in the route shadow table. Instead, it performs a weighted sum based on the freshness weight of each route entry to obtain the weighted route entry count. Specifically, for each route entry in the route shadow table about a subordinate node of the abnormal master node, its freshness weight is used as the weighting value for that entry. The weighting values ​​of all such route entries are then added together to obtain the weighted route entry count. Since newer route entries have a larger freshness weight and older route entries have a smaller freshness weight, the weighted route entry count more accurately reflects the effectiveness of the routing information about subordinate nodes of the abnormal master node currently possessed, rather than merely reflecting the number of historically accumulated entries. For example, if a master node's routing shadow table contains many routing entries about abnormal master node subordinate nodes, but most of these entries were received earlier and have a lower freshness weight, then the number of its weighted routing entries is still small, indicating that the master node's currently available routing information is insufficient. Conversely, if the number of routing entries is small but the number of entries is recent and the freshness weight is high, then the number of its weighted routing entries is relatively large, indicating that the master node currently possesses more effective routing information.

[0085] S3114: The master node uses the ratio of the weighted number of routing entries to the total number of all routing entries of all subordinate nodes of the abnormal master node as the routing entry coverage rate.

[0086] After obtaining the weighted number of routing entries, the master node uses the ratio of this ratio to the total number of routing entries of all subordinate nodes of the abnormal master node as the routing entry coverage rate. The total number of routing entries of all subordinate nodes of the abnormal master node refers to the total number of routing entries corresponding to the base station nodes and slave nodes connected to the abnormal master node during normal operation. This number can be obtained by the master node based on historical routing summary information or network planning data. The routing entry coverage rate reflects the proportion of effective routing information currently held by the master node relative to the total routing information of all subordinate nodes of the abnormal master node. A higher proportion indicates that the master node can quickly restore its routing forwarding capabilities to the base station nodes and slave nodes of the abnormal master node after takeover; a lower proportion indicates that the master node may need a longer time to replenish the relevant routing information after takeover. Through the above steps S3111 to S3114, the master station node calculates the freshness weight of each route entry in the route shadow table based on the received timestamp, and uses the ratio of the weighted route entry number to the total number of all route entries as the route entry coverage rate. This ensures that the coverage rate reflects not only the number of route entries but also the timeliness of the route entries, thereby improving the accuracy of judging the master station node's succession capability and providing a more reliable basis for succession decisions when the master station node is abnormal.

[0087] Furthermore, in step S313, the master node that initiates the vote coordinates with other master nodes to supplement the missing routing entries, including the following sub-steps.

[0088] S3131: After receiving a rejection response carrying missing route entry information, the master node that initiated the vote broadcasts a route completion request to other master nodes. The route completion request includes the identifier of the abnormal master node and the missing route entry information.

[0089] In scenarios such as remote communication in mountainous areas or disaster emergency communication, when a master station node detects an anomaly in a neighboring master station node and initiates a succession vote, some master station nodes may return a vote of refusal to succeed due to insufficient coverage of routing entries for subordinate nodes of the anomaly master station node in their own routing shadow tables. In this case, the master station node that initiated the vote does not directly abandon these candidate nodes. Instead, it extracts the missing routing entry information carried in the vote of refusal and broadcasts a route completion request to other master station nodes in the master station layer ad-hoc network. This route completion request includes the identifier of the anomaly master station node and the missing routing entry information that needs to be supplemented, so that other master station nodes can clearly understand the range of routing entries they need to query and return. By broadcasting the route completion request, the master station node that initiated the vote transforms the problem of insufficient routing information from a single master station node into a problem of collaborative supplementation by multiple master station nodes. This allows master station nodes that were originally unable to succeed due to incomplete routing information to obtain the required routing entries with the assistance of other master station nodes, thereby expanding the range of available successor master station nodes.

[0090] S3132: The master node that receives the route completion request queries its own maintained route shadow table. If there are missing route entries in its own route shadow table, it returns the queried route entries in the route completion response to the master node that initiated the vote.

[0091] Upon receiving a route completion request, other master nodes check their own route shadow tables for corresponding route entries based on the abnormal master node identifier and missing route entry information carried in the request. Since each master node exchanges route summary information with different neighboring master nodes during operation, the route entries in their respective route shadow tables may differ. Some master nodes, having exchanged route information with the abnormal master node or its neighboring master nodes, may happen to know the missing route entries of other master nodes. If a missing route entry is found in its own route shadow table, the master node returns these route entries in its route completion response to the master node that initiated the vote. In this way, route information about the subordinate nodes of the abnormal master node, scattered across various master nodes, is aggregated to the master node that initiated the vote, providing a data source for subsequently completing the route shadow tables of the master node that refused to take over.

[0092] S3133: The master node that initiated the vote summarizes the route entries in each route completion response, generates a set of completed route entries, and sends the set of completed route entries to the master node that returned the response of refusing to take over the vote.

[0093] After receiving route completion responses from multiple master nodes, the master node initiating the vote summarizes the route entries carried in each response, removes duplicate entries, and retains the latest version of the route entries to generate a complete route entry set. This set contains missing route entries about subordinate nodes of the abnormal master node collected from other master nodes, which can be used to complete the route shadow table of the master node that returned the response of refusing to take over the vote. The master node initiating the vote sends this complete route entry set to the master nodes that previously returned the response of refusing to take over the vote, enabling these master nodes to update their own route shadow tables accordingly. By summarizing and distributing the complete route entry set, the master node initiating the vote achieves collaborative sharing of route information about subordinate nodes of abnormal master nodes among master nodes without changing the basic architecture of the master layer Ad-hoc network.

[0094] S3134: The master node that returns the rejection vote response writes the received set of complete route entries into its own maintained route shadow table, and re-determines whether it has the ability to take over based on the updated route entry coverage. If it does, it returns a vote response agreeing to take over to the master node that initiated the vote.

[0095] Upon receiving the complete set of route entries, the master node that returned the rejection vote response adds these entries to its own maintained route shadow table, thus supplementing the route entries for the abnormal master node's subordinate nodes. Subsequently, the master node recalculates the route entry coverage for the abnormal master node's subordinate nodes based on the updated route shadow table and again determines whether it meets the preset replacement capability conditions. If the supplemented route entry coverage meets the conditions, it indicates that the master node has the capability to replace the others with the assistance of other master nodes. In this case, it returns a vote of agreement to replace to the master node that initiated the vote, becoming one of the candidate replacement master nodes. If the supplemented coverage still does not meet the conditions, the master node continues to refuse replacement. Through the above steps S3131 to S3134, after receiving a rejection vote response carrying missing routing entry information, the master node that initiated the vote broadcasts a route completion request, collects route completion responses from other master nodes, generates a set of completed routing entries, and sends it to the master node that refused to take over. This allows the latter to reassess its takeover capability after supplementing its routing shadow table, thereby including master nodes that were originally unable to take over due to incomplete routing information into the candidate takeover range. This improves the success rate of the takeover vote and the flexibility of selecting the takeover master node when the master node is abnormal in the decentralized Ad-hoc network. At the same time, the collaborative completion among master nodes avoids the failure of the takeover process due to insufficient routing information of individual master nodes.

[0096] Specifically, in step S3, the voting response is counted. When the number of votes agreeing to take over exceeds a preset ratio, the main station node with the most votes is determined as the successor main station node, including the following sub-steps.

[0097] S321: Each master node returning a voting response carries its own device identification number priority and the coverage of routing entries for subordinate nodes of the abnormal master node in its own routing shadow table. The device identification number priority is determined according to the number segment to which the device identification number belongs; the master number segment has a higher priority than the base station number segment, and the base station number segment has a higher priority than the slave number segment.

[0098] In scenarios such as remote communication in mountainous areas or disaster emergency communication, when a master station node detects an anomaly in a neighboring master station node and initiates a succession vote, other master station nodes in the master station layer ad-hoc network, upon receiving the succession vote request, will determine whether they possess the capability to succeed the node based on their own routing shadow table and return a voting response to the master station node that initiated the vote. Each master station node returning a voting response not only indicates whether it agrees to succeed the node but also carries its own device identification number priority and the coverage rate of routing entries for subordinate nodes of the anomaly master station node in its own routing shadow table. The device identification number priority is determined by the number segment to which the device identification number belongs; the master station number segment has higher priority than the base station number segment, and the base station number segment has higher priority than the slave station number segment. Within the same number segment, the smaller the device identification number value, the higher the priority. Since all master station nodes participating in the succession vote belong to the master station number segment, their device identification number priority is determined by the numerical sorting within the master station number segment; the smaller the value, the higher the priority. Routing entry coverage reflects the master node's current understanding of routing information for nodes subordinate to the abnormal master node. A higher coverage indicates that the master node can quickly restore its routing forwarding capabilities to both base station and slave nodes after taking over. The voting response simultaneously carries the device identification number priority and routing entry coverage, providing two dimensions for calculating the weighted vote count. This allows the selection of the successor master node to comprehensively consider both the priority of candidate nodes and their understanding of routing information.

[0099] S322: The master node that initiates the vote calculates a weighted number of votes for each vote response that agrees to take over, based on the device identification number priority and the routing entry coverage.

[0100] After receiving voting responses from other master nodes, the master node initiating the vote calculates a weighted vote count for each response agreeing to take over. The weighted vote count considers both the device identification number priority and the routing entry coverage of the voting response. Higher device identification number priority and greater routing entry coverage result in a larger weighted vote count for the response agreeing to take over. This weighting method gives master nodes with more routing information on subordinate nodes of the abnormal master node and higher device identification number priority greater weight in the succession decision, thus facilitating the selection of a successor master node that possesses both high priority and sufficient routing information. For example, when two master nodes both return votes agreeing to take over, one with a lower device identification number but lower routing entry coverage, and the other with a higher device identification number but higher routing entry coverage, the weighted calculation can consider the differences in both dimensions, avoiding the selection of a master node with insufficient routing recovery capability or low priority after succession based on only a single factor. The specific calculation method for the weighted vote can be set according to the network scale and actual deployment needs, as long as the principle that the higher the priority of the device identification number and the greater the coverage of the routing entries, the greater the weighted vote will be.

[0101] S323: When the weighted number of votes in favor exceeds the preset proportion of the total weighted number of votes, the main station node that initiated the vote determines the main station node with the highest weighted number of votes as the successor main station node.

[0102] After calculating the weighted votes for each vote in favor of succession, the master node initiating the vote sums all the weighted votes to obtain the weighted approval vote count. Simultaneously, the total weighted vote count of all participating master nodes is calculated. Each participating master node determines a basic weighted vote count based on its device identification number priority and routing entry coverage. The weighted votes of master nodes agreeing to succession are included in the weighted approval vote count, while the basic weighted votes of master nodes rejecting succession are not included. When the proportion of weighted approval votes to the total weighted vote count exceeds a preset proportion, it indicates that the master nodes agreeing to succession have a sufficient majority in a weighted sense, and the succession decision has the support of enough master nodes in the master station layer. At this point, the master node initiating the vote determines the master node with the highest weighted vote count among those agreeing to succession as the successor master node. Through the above steps S321 to S323, the master station node not only counts the number of votes in favor of succession during the succession vote, but also calculates the weighted votes based on the device identification number priority and routing entry coverage. Only when the weighted number of votes in favor exceeds the preset proportion of the total weighted votes is the successor master station node determined. This makes the selection of the successor master station node comprehensively consider the priority of the candidate nodes and the degree of mastery of the routing information of the subordinate nodes of the abnormal master station, improves the accuracy of the succession decision and the completeness of the routing recovery after succession, and avoids the problem that relying solely on a simple majority vote may result in insufficient routing information or low priority of the selected successor master station node.

[0103] Furthermore, the coverage rate of routing entries for subordinate nodes of the abnormal master node in the self-routing shadow table carried in the voting response in step S321 is generated in the following manner.

[0104] S3211: When generating a voting response, the master node calculates the coverage of routing entries for subordinate nodes of the abnormal master node in its own routing shadow table, as well as the change in coverage within a preset historical time window.

[0105] In scenarios such as remote communication in mountainous areas or disaster emergency communication, when a master station node receives a succession voting request and prepares to respond, it needs to not only determine whether it currently possesses the capability to succeed the previous master station node, but also provide the voting master station node with the coverage rate of routing entries for subordinate nodes of the abnormal master station node in its routing shadow table and the change in this coverage rate within a preset historical time window. The coverage rate is calculated in the same way as described above, based on the routing entries for subordinate nodes of the abnormal master station node already recorded in the routing shadow table. The change in coverage rate reflects its recent trend; for example, within the preset historical time window, coverage may gradually increase due to continuous reception of routing summary information from neighboring master station nodes, or it may decrease due to some routing entries timeouts. By providing both coverage rate and change in coverage rate, the voting master station node can not only understand how much routing information the candidate master station node currently possesses about subordinate nodes of the abnormal master station, but also determine whether the candidate master station node's grasp of routing information is continuously improving or gradually deteriorating. For example, if a master node currently has high coverage but the change is negative and decreasing significantly, it indicates that the relevant routing entries in its routing shadow table may be rapidly becoming invalid, and it may face the risk of insufficient routing information after being replaced; if the coverage is moderate but the change is positive and increasing rapidly, it indicates that the master node is continuously supplementing relevant routing information through routing summary exchange, and its succession capability may be improved quickly.

[0106] S3212: The master node obtains the list of missing route entries in its own routing shadow table and sorts the list of missing route entries according to the priority of the device identification number of the abnormal master node corresponding to the missing route entry.

[0107] When generating a voting response, the master node also retrieves a list of missing route entries from its own routing shadow table. Missing route entries refer to route entries that should exist in the subordinate nodes of the abnormal master node but have not yet been recorded in the master node's routing shadow table. These missing route entries correspond to either base station nodes connected to the abnormal master node or slave nodes, with their device identification numbers belonging to either the base station or slave number ranges and having different priorities. The master node sorts the list of missing route entries according to the priority of the device identification numbers of the subordinate nodes corresponding to the missing route entries. The master number range has higher priority than the base station number range, and the base station number range has higher priority than the slave number range. Within the same number range, the smaller the device identification number value, the higher the priority. This sorting ensures that the list of missing route entries is arranged according to the importance of the corresponding nodes, allowing the master node initiating the vote to prioritize higher-priority missing route entries during subsequent coordination and completion. For example, if the missing routing entries include entries for both base station nodes and slave nodes, the missing entries for the base station nodes should be placed before the missing entries for the slave nodes. This is because base station nodes are responsible for coverage extension on the access side, and the lack of their routing information has a greater impact on the overall communication capabilities of the nodes under the abnormal master station.

[0108] S3213: The master node includes coverage, changes, and a sorted list of missing route entries in its voting response.

[0109] After completing coverage calculation, change rate calculation, and sorting of the missing route entry list, the primary node returns these three pieces of information to the primary node that initiated the vote in its voting response. Coverage indicates the candidate primary node's current understanding of the routing information of the abnormal primary node's subordinate nodes; change rate indicates the recent trend of this understanding; and the sorted list of missing route entries indicates the specific deficiencies in the candidate primary node's succession capabilities and the route entries that need to be prioritized for addition. Upon receiving the voting response containing this information, the primary node that initiated the vote can more comprehensively evaluate the succession capabilities of each candidate primary node and decide whether to directly appoint a successor or initiate coordination to complete the missing route entries. For example, if a candidate master node has high coverage and a positive change, the master node initiating the vote can prioritize this node as the successor master node. If a candidate master node's coverage is close to the successor capability condition but its missing route entry list contains many high-priority base station node entries, the master node initiating the vote can initiate route completion requests for these high-priority missing entries. After completion, the candidate master node will reassess its successor capability. Through S3211 to S3213, the master node carries coverage, change, and a sorted list of missing route entries in its voting response. This allows the master node initiating the vote to obtain richer and more targeted judgment criteria when counting votes and determining the successor master node, thereby improving the accuracy of successor decisions and the completeness of route recovery when a master node is abnormal in a decentralized Ad-hoc network.

[0110] Further, in step S322, a weighted number of votes is calculated for each vote response agreeing to take over, based on the device identification number priority and routing entry coverage, including the following sub-steps.

[0111] S3221: The master node that initiated the vote obtains the coverage confidence level carried in each vote response that agrees to take over. The coverage confidence level is determined based on the difference between the version number of the routing shadow table of the master node corresponding to the vote response and the latest version number of the routing table. The smaller the difference, the higher the confidence level.

[0112] In scenarios such as remote communication in mountainous areas or disaster emergency communication, after receiving multiple votes agreeing to take over, the master node initiating the vote needs to assess the credibility of the claimed takeover capabilities of these responses. To this end, the master node initiating the vote obtains the coverage confidence score carried by each vote agreeing to take over. This coverage confidence score is determined based on the difference between the routing shadow table version number of the master node corresponding to the vote and the current latest routing table version number. The routing shadow table version number reflects the age of the routing shadow table maintained by the master node, while the current latest routing table version number reflects the latest version of the entire network routing table in the master layer Ad-hoc network. The smaller the difference, the closer the routing shadow table of the master node is to the latest state, and the fresher and more reliable the routing information on which its routing entries are based, thus resulting in a higher coverage confidence score. Conversely, a larger difference indicates that the routing shadow table of the master node may be outdated, and its coverage data may be based on outdated routing information, resulting in relatively lower credibility. By introducing coverage confidence, the main station node that initiates the vote can distinguish the credibility of the coverage data provided by different candidate main station nodes, avoiding inflated coverage or misjudgment due to outdated routing shadow tables of individual main station nodes, and making the subsequent weighted vote calculation based on a more reliable data foundation.

[0113] S3222: The master node initiating the vote calculates the weighted vote count for each vote response agreeing to take over, based on the device identification number priority, route entry coverage, and coverage confidence, according to a preset weighted formula. In the preset weighted formula, the sum of the weight coefficients of the device identification number priority, route entry coverage, and coverage confidence is 1, and the weight coefficient of coverage confidence is greater than the weight coefficient of device identification number priority.

[0114] After obtaining the device identification number priority, route entry coverage, and coverage confidence level, the master node initiating the vote calculates the weighted vote count for each vote in favor of taking over, according to a preset weighted formula. In this formula, the sum of the weighting coefficients for device identification number priority, route entry coverage, and coverage confidence level is 1, and the weighting coefficient for coverage confidence level is greater than that for device identification number priority. This means that when calculating the weighted vote count, coverage confidence level has a greater impact on the result than device identification number priority. This design is based on the consideration that the core task of the succeeding master node is to restore the routing forwarding capability for nodes subordinate to the abnormal master node. Route entry coverage and its confidence level directly reflect whether the candidate master node truly possesses sufficient and up-to-date routing information, and therefore should be more important than simply managing priority. Device identification number priority serves as an auxiliary factor, providing a definite arbitration basis when multiple candidate master nodes have similar levels of routing information. Through this weighted formula, the master node initiating the vote can quantitatively score each vote response agreeing to the replacement based on three dimensions. This ensures that the final selected replacement master node possesses both a high degree of routing information mastery and credibility, while also considering device identification number priority. This improves the accuracy of the replacement decision and the completeness of route recovery after the replacement. Through S3221 to S3222, the master node initiating the vote incorporates coverage confidence when calculating the weighted vote count. It scores each vote response agreeing to the replacement based on a weighted combination of device identification number priority, routing entry coverage, and coverage confidence. This ensures that the selection of the replacement master node considers not only the candidate node's mastery and priority of routing information of the abnormal master node's subordinate nodes, but also the age and credibility of its routing shadow table. This improves the accuracy of the replacement decision and the completeness of route recovery after the replacement when the master node is abnormal in a decentralized ad-hoc network, avoiding the problem of inappropriate replacement master node selection due to outdated routing information or a single priority factor.

[0115] Specifically, in step S4, an updated network routing table is generated based on the routing information of subordinate nodes of the abnormal master node in the network routing table maintained by itself and the routing shadow table, including the following sub-steps.

[0116] S411: The taking over master node deletes the routing entries related to the abnormal master node from its own network-wide routing table.

[0117] In scenarios such as remote communication in mountainous areas or disaster emergency communication, when a master node is identified as the replacement master node, the replacement master node first needs to clear the routing entries related to the abnormal master node in its own network routing table. The abnormal master node can no longer perform routing addressing functions. If its routing entries remain in the replacement master node's network routing table, subsequent data forwarding may still attempt to route through the abnormal master node, resulting in data not being delivered or forwarding failing. Therefore, the replacement master node deletes all routing entries in its own network routing table that have the abnormal master node as the destination master node or next hop, so that its own network routing table no longer contains routing information pointing to the abnormal master node. Through this deletion operation, the replacement master node formally severs the routing relationship with the abnormal master node at the routing level, freeing up space for the subsequent inclusion of the routing information of the abnormal master node's subordinate nodes in its own network routing table, and also providing a clear basis for other master nodes to update their network routing tables.

[0118] S412: The taking over master node writes the routing entries for subordinate nodes of the abnormal master node from the routing shadow table into the full network routing table it maintains.

[0119] After deleting the routing entries for the abnormal master node, the succeeding master node further writes the routing entries for subordinate nodes of the abnormal master node from its own routing shadow table into its own maintained network-wide routing table. These routing entries in the routing shadow table, containing information about subordinate nodes of the abnormal master node, were obtained and cached by the succeeding master node before the anomaly occurred through the exchange of routing summary information with other master nodes. These entries record the reachability information of the abnormal master node's subordinate base station nodes and slave nodes. After the succeeding master node writes these routing entries into its own network-wide routing table, its network-wide routing table contains routing information to the abnormal master node's subordinate base station nodes and slave nodes. This allows service data that was originally forwarded by the abnormal master node to be forwarded by the succeeding master node instead. Through this write operation, the succeeding master node can restore its routing forwarding capability for subordinate nodes of the abnormal master node based on the routing information already in its own routing shadow table without relying on the abnormal master node to resend its complete routing table. This ensures that subordinate nodes of the abnormal master node can still be addressed and reached by other nodes in the network after the abnormal master node fails.

[0120] S413: When a routing entry in the routing shadow table for the same destination master node conflicts with the corresponding entry in the network-wide routing table maintained by itself, the routing entry with the smaller hop count is selected. If the hop counts are the same, the routing entry with the newer version number is selected.

[0121] When a succeeding master node writes routing entries from the routing shadow table for nodes subordinate to the abnormal master node into its own global routing table, it may encounter situations where the same destination master node has routing entries in both the routing shadow table and its own global routing table. Since the routing entries in the routing shadow table are indirectly obtained by the succeeding master node through adjacent master nodes, while the routing entries in its own global routing table are those originally maintained by the succeeding master node, there may be differences in next hop, hop count, or version number, leading to conflicts. For such conflicts, the succeeding master node handles them according to preset conflict resolution rules: it prioritizes the routing entry with the smaller hop count, as a smaller hop count usually means a shorter path to the destination master node, lower forwarding latency, and less channel resource usage; if two routing entries have the same hop count, it selects the routing entry with the more up-to-date version number, as an updated version number indicates that the routing state reflected by the entry is closer to the current time. Through this conflict resolution rule, the node taking over the master station can select a better route entry when merging routing information, avoiding longer routing paths or inconsistent routing information due to simple overwriting or arbitrary selection.

[0122] S414: After the merge is completed, the node that takes over from the master node will assign a new version number to the newly added routing entries and record the assigned version number in the network-wide routing table it maintains.

[0123] After completing the deletion, writing, and conflict resolution of routing entries, the succeeding master node assigns a new version number to the newly added routing entries and records the assigned version number in its own network-wide routing table. Newly added routing entries refer to those that the succeeding master node wrote from its routing shadow table to its own network-wide routing table but were not previously present in its own network-wide routing table. Assigning new version numbers to these new routing entries serves two purposes: firstly, it distinguishes the version number of the succeeding master node's network-wide routing table from that before the merge, making it easier for other master nodes to identify that the succeeding master node has generated new routing information in subsequent routing information exchanges; secondly, it provides a clear version basis for the new routing entries during subsequent conflict resolution. After the succeeding master node records the assigned version number in its own network-wide routing table, the network-wide routing table merge process is complete, and the succeeding master node can broadcast the updated routing table to other master nodes. Through the above steps S411 to S414, the replacement master node completes the update of its own network routing table by deleting the routing entries of the abnormal master node, writing the routing entries of the subordinate nodes of the abnormal master node, resolving conflicts by hop count and version number, and assigning new version numbers to the newly added routing entries. This enables the subordinate base station nodes and slave nodes of the abnormal master node to regain route reachability at the replacement master node, thereby achieving smooth migration and rapid recovery of routing information when the master node is abnormal without relying on the central control node.

[0124] Furthermore, in step S412, the node taking over from the master station writes the routing entries for the subordinate nodes of the abnormal master station node in the routing shadow table into the network-wide routing table it maintains, including the following sub-steps.

[0125] S4121: The taking over master node extracts routing entries for subordinate nodes of the abnormal master node from the routing shadow table, and calculates the coverage contribution value of each routing entry based on the receiving timestamp in the routing shadow table. The closer the receiving timestamp is to the current time, the greater the coverage contribution value.

[0126] In scenarios such as remote communication in mountainous areas or emergency communication during disasters, after the replacement master node is determined, it needs to write the routing entries for subordinate nodes of the abnormal master node from its own routing shadow table into the full network routing table to restore the routing forwarding capability for the base station nodes and slave nodes under the abnormal master node. However, the routing entries in the routing shadow table are obtained at different times by receiving routing summary information sent by neighboring master nodes, and their timeliness varies. If all relevant routing entries are directly written into the full network routing table without distinction, outdated or inaccurate routing information may be introduced, affecting the correctness of subsequent data forwarding. Therefore, the replacement master node first extracts all routing entries for subordinate nodes of the abnormal master node from the routing shadow table, and for each routing entry, calculates the coverage contribution value based on its reception timestamp in the routing shadow table. The closer the reception timestamp is to the current time, the newer the routing state reflected by the routing entry, and the greater its coverage contribution value; the farther the reception timestamp is from the current time, the more likely the routing entry is outdated, and the smaller its coverage contribution value. In this way, the successor node can quantitatively evaluate the timeliness of each relevant routing entry in the routing shadow table, providing a basis for deciding which routing entries can be directly written into the network routing table and which need to be supplemented by other master nodes.

[0127] S4122: The node taking over from the master station will directly write route entries with coverage contribution values ​​higher than the preset contribution threshold into the network-wide routing table it maintains.

[0128] After calculating the coverage contribution value of each routing entry for the subordinate nodes of the abnormal master station, the succeeding master station node compares it with a preset contribution threshold. When the coverage contribution value of a routing entry is higher than the preset contribution threshold, it indicates that the reception timestamp of the routing entry is closer to the current time, and the routing status it reflects is relatively fresh and reliable. The succeeding master station node can trust the destination master station node identifier, next hop, hop count, and other information contained in the routing entry, and therefore directly writes the routing entry into its own maintained network-wide routing table. Through this operation, the succeeding master station node can quickly incorporate timely routing entries from the routing shadow table into the network-wide routing table, enabling the more recent routing information in the subordinate nodes of the abnormal master station to quickly restore routing reachability and reduce the impact of the succession process on service communication. The preset contribution threshold can be set according to the network size and channel change rate to achieve a balance between the timeliness and availability of routing information.

[0129] S4123: For route entries whose coverage contribution value is lower than or equal to the preset contribution threshold, take over the primary node to send route entry completion requests to other primary nodes, and receive supplementary route entries returned by other primary nodes.

[0130] For route entries with coverage contribution values ​​lower than or equal to a preset contribution threshold, the succeeding master node determines that their timeliness is insufficient. Directly writing them into the full network routing table might introduce outdated or incorrect routing information, so they are not used directly. Instead, a route entry completion request is sent to other master nodes in the master layer ad-hoc network. This request includes information about the route entries that need to be supplemented, such as the identifier of the abnormal master node and the identifier of the destination master node corresponding to the specific missing or outdated route entry. Upon receiving the request, other master nodes query their own maintained routing shadow table or the full network routing table. If they possess the corresponding and relatively new route entry, they include it in the supplementary route entry and return it to the succeeding master node. In this way, the succeeding master node can obtain the latest routing information about the subordinate nodes of the abnormal master from other master nodes, compensating for the timeliness issues of some route entries in its own routing shadow table, thus ensuring that the route entries ultimately written into the full network routing table have sufficient accuracy and completeness.

[0131] S4124: The node that takes over from the master station merges the directly written routing entries with the supplementary routing entries, writes them into the network-wide routing table it maintains, and assigns a unified new version number to the merged routing entries.

[0132] After completing direct writing and supplementary collection, the succeeding master node merges the route entries directly written (those with coverage contributions exceeding a preset contribution threshold) with supplementary route entries received from other master nodes. During merging, if there are differences between the directly written and supplementary route entries for the same destination master node, the selection is based on a rule prioritizing entries with lower hop counts, and if hop counts are the same, prioritizing version number updates, to determine the final route entry to be written to the network-wide routing table. After merging, the succeeding master node assigns a unified new version number to the merged route entries and records the assigned version number in its own maintained network-wide routing table. This unified assignment of new version numbers helps distinguish the updated network-wide routing table of the succeeding master node from those of other master nodes in terms of version, facilitating subsequent routing information exchange and conflict resolution. Through steps S4121 to S4124, the successor master node calculates the coverage contribution value based on the reception timestamp of each route entry in the routing shadow table. Route entries with high timeliness are directly written into the entire network routing table, while route entries with low timeliness are supplemented by other master nodes and then merged and written. A unified new version number is assigned to the merged route entries. This ensures the accuracy and completeness of routing information and achieves a smooth migration of routing information from subordinate nodes of the abnormal master node to the successor master node, thereby improving the reliability and efficiency of route recovery when the master node is abnormal in a decentralized Ad-hoc network.

[0133] Specifically, step S4 involves broadcasting the routing table update to other master nodes, including executing the following routing table update confirmation process.

[0134] S421: After taking over as the master node, broadcast the routing table update to other master nodes, wait for the other master nodes to return the routing update confirmation message.

[0135] In scenarios such as remote communication in mountainous areas or disaster emergency communication, after the succeeding master node completes its own network-wide routing table update, it needs to broadcast the updated routing table information to other master nodes in the master-level Ad-hoc network. This allows other master nodes to synchronously update their own network-wide routing tables, thereby achieving re-convergence of routing information within the master-level layer. After broadcasting the routing table update to other master nodes, the succeeding master node does not immediately consider the update to be effective. Instead, it enters a waiting state, awaiting confirmation messages from other master nodes upon receiving the updates. These confirmation messages indicate that the receiving master node has received and processed the routing table update, and its own network-wide routing table has been adjusted accordingly. Because the master-level Ad-hoc network adopts a decentralized architecture, the succeeding master node cannot force other master nodes to accept the update. It can only understand the acceptance status of the routing table update by waiting for confirmation messages, thus providing a basis for subsequently determining whether the routing table update has been accepted by a sufficient number of master nodes in the master-level layer.

[0136] S422: When the number of master nodes returning route update confirmation messages exceeds a preset ratio, the master node takes over to confirm that the route table update has taken effect and updates the version number of the network-wide route table it maintains.

[0137] After waiting for a preset time, the succeeding master node counts the number of master nodes returning route update confirmation messages. When the number of master nodes returning route update confirmation messages exceeds a preset proportion, it indicates that a majority of master nodes in the master-level ad-hoc network have received and processed the routing table update, and the update has gained sufficient recognition from master nodes in the decentralized network. At this point, the succeeding master node confirms the routing table update is effective and updates the version number of its maintained network-wide routing table to the new version number. This version number update distinguishes the succeeding master node's current network-wide routing table from its previous version, facilitating the exchange of route summary information with other master nodes and allowing other master nodes to determine the timeliness of route entries based on the version number during subsequent conflict resolution. By setting a preset proportion and confirming the effectiveness of the routing table update accordingly, the succeeding master node avoids misjudging update failure due to a few master nodes not returning confirmation, and also avoids premature updates to the network-wide routing table version number due to a few master nodes not synchronizing.

[0138] S423: When the number of master nodes returning route update confirmation messages does not exceed the preset ratio, the successor master node re-executes steps S421 to S422 based on the network-wide routing table before merging, or requests other master nodes to assist in completing the routing table update.

[0139] When the number of master nodes returning route update confirmation messages does not exceed a preset proportion, it indicates that the number of master nodes in the master-level Ad-hoc network receiving and processing this route table update is insufficient, and the route table update has not yet gained enough recognition from master nodes. In this case, the succeeding master node does not use the currently updated network-wide route table version number as the effective version. Instead, it re-executes steps S421 to S422 based on the network-wide route table before merging, i.e., it rebroadcasts the route table update to other master nodes and waits for confirmation messages to attempt to obtain confirmation from enough master nodes again; or, the succeeding master node requests other master nodes to assist in completing the route table update, such as requesting already confirmed master nodes to assist in forwarding the route table update, or requesting other master nodes to provide supplementary route entries to complete the update content before rebroadcasting. Through this processing method, the succeeding master node can avoid directly using an ineffective route table version when the route table update has not received enough confirmation, ensuring the consistency and reliability of routing information updates in the master-level Ad-hoc network.

[0140] The above steps S421 to S423 involve broadcasting routing table updates and waiting for confirmation when the number of confirmations exceeds a preset proportion, confirming the update as effective and updating the version number, and rebroadcasting or requesting assistance when the number of confirmations is insufficient. This ensures that the routing table update receives joint confirmation from multiple master nodes in the master-level Ad-hoc network, avoiding global routing inconsistencies caused by a few master nodes not synchronizing. This improves the reliability and stability of routing table updates after abnormal succession of master nodes in a decentralized variable topology network.

[0141] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An Ad-hoc decentralized variable topology networking system based on a hierarchical management architecture, characterized in that, It includes the master station layer, the base station layer, and the slave station layer; The master station nodes in the master station layer constitute a decentralized, variable topology ad-hoc network and undertake routing and addressing functions, generating and maintaining the entire network routing table by exchanging routing information; The base station nodes in the base station layer are connected to the master station node, do not undertake routing and addressing functions, and the changes in the network entry and exit status of the base station nodes do not trigger the update of the network-wide routing table. The slave nodes in the slave layer connect to the master node or the base station node, do not undertake routing and addressing functions, and the network entry and exit status changes of the slave nodes do not trigger the update of the network-wide routing table.

2. The system according to claim 1, characterized in that, The master node, the base station node, and the slave node are all configured with a device identification number. The device identification number is divided into non-overlapping master number segments, base station number segments, and slave number segments. Each node determines itself as the master node, the base station node, or the slave node based on the number segment to which its device identification number belongs.

3. The system according to claim 1, characterized in that, In the master station layer, each master station node is configured to execute a self-organizing networking protocol, discover neighboring master station nodes by sending probe signal frames, and establish the Ad-hoc network with other master station nodes. In the Ad-hoc network, routing information is exchanged to generate and maintain the network-wide routing table.

4. The system according to claim 1, characterized in that, The network-wide routing table only contains the address information of other master nodes in the master station layer, but does not contain the address information of the base station nodes or the slave nodes; the master station node maintains the local access tables of its subordinate base station nodes and slave nodes, and the local access tables are not included in the network-wide routing table.

5. The system according to claim 1, characterized in that, The base station node is configured to send a network access request to a pre-designated master station node and establish a star connection to extend the signal coverage of the pre-designated master station node; the network access and disconnection status changes of the base station node are only processed within the star connection range and do not trigger the update of the entire network routing table; when the base station node detects an anomaly of the master station node it has accessed, the base station node scans and accesses other master station nodes.

6. The system according to claim 1, characterized in that, The slave node is configured to send a network access request to a pre-specified master node or a pre-specified base station node by default and establish a star connection; when roaming is enabled and the pre-specified master node or the pre-specified base station node is unreachable, it automatically selects an accessible master node or base station node and establishes a star connection; the network access and disconnection status changes of the slave node are only processed within the scope of the star connection and do not trigger the update of the entire network routing table.

7. The system according to claim 1, characterized in that, The master node, the base station node, and the slave node all use the same core hardware platform and start different working modes by writing device identification numbers of different number segments, thereby respectively undertaking the network roles of master station, base station, and slave station.

8. The system according to claim 1, characterized in that, The base station node and the slave node do not participate in the routing information exchange between the master station nodes; the routing information exchange between the master station nodes is only performed by the master station nodes in the master station layer.

9. The system according to claim 1, characterized in that, The master node, base station node, and slave node can all scan the neighboring nodes that they communicate directly with in one hop, generate and maintain a neighboring node table. When the destination address of the data to be sent exists in the neighboring node table, the node device can send the data directly to the corresponding neighboring node without going through the master node for routing and forwarding.

10. An Ad-hoc decentralized variable topology networking method based on a hierarchical management architecture, applied to the networking system according to any one of claims 1 to 9, executed by the master station node in the networking system, characterized in that, The method includes the following steps: S1: During operation, the master node exchanges routing information with neighboring master nodes. The routing information includes a summary of the entire network routing table. Based on the received summary information, the master node maintains a routing shadow table of neighboring master nodes locally. The routing shadow table contains at least some entries of the entire network routing table maintained by the neighboring master nodes. The routing shadow table contains routing entries, and each routing entry includes the destination master node identifier, next hop, hop count, and version number. S2: When the master node detects that a neighboring master node is abnormal, it initiates a succession voting request to other master nodes. The succession voting request includes the identifier of the abnormal master node and the version information of the routing shadow table maintained by the master node that initiated the vote. S3: The master station node that receives the replacement voting request determines whether it has the ability to take over based on its own maintained routing shadow table and returns a voting response; the master station node that initiated the vote counts the voting responses, and when the number of votes agreeing to take over exceeds a preset ratio, the master station node with the most votes is determined as the replacement master station node; S4: The taking over master node generates an updated network routing table based on its own maintained network routing table and the routing information of the subordinate nodes of the abnormal master node in the routing shadow table, and broadcasts the routing table update to other master nodes; the subordinate nodes of the abnormal master node include base station nodes and slave nodes that are connected to the abnormal master node.