A method for shortwave autonomous frequency-selective radio station networking architecture and network control equipment

Abstract

This invention discloses a shortwave autonomous frequency-selective radio station networking architecture method and network control device, belonging to the field of wireless communication technology. The method includes: constructing M radio stations into a multi-level tree-like network architecture consisting of a primary network and multiple subnets, where the subnet topology can be a star network and / or a mesh network; using a non-uniform partitioning method to determine the node affiliation of each level of subnet based on the geographical distribution or organizational structure of the radio stations; limiting the subnet size, with the number of nodes in the primary network and intermediate subnets not exceeding a first preset threshold N1, and the number of nodes in the final subnet not exceeding a second preset threshold N2; for handover nodes that have both upper and lower level roles, using time-division multiple access (TDMA) or frequency-division multiple access (FDMA) to maintain their simultaneous online status. This invention, through its hierarchical and divide-and-conquer architecture design, effectively shortens network construction time, balances overall network transmission latency, and does not rely on external facilities, thus improving the practicality and reliability of the shortwave network.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, specifically relating to a shortwave autonomous frequency-selective radio station networking architecture method and network control equipment. Background Technology

[0002] Shortwave autonomous frequency selection technology significantly improves the reliability and connection rate of point-to-point communication links by automatically selecting the optimal communication frequency at both ends of the link. However, it faces serious challenges when using this technology to build communication networks with multiple radio stations, especially over large areas beyond line of sight.

[0003] The core issue is that shortwave autonomous frequency selection technology is essentially optimized for single-link communication. The optimal receiving frequencies for radios at both ends of the link are highly likely to differ. This asymmetry in transmission and reception frequencies makes it nearly impossible for multiple radios distributed over a wide area to find a common optimal frequency for co-channel networking. While co-channel networks can be formed within line-of-sight, uneven interference distribution means that some radios will inevitably fail to join the network.

[0004] like Figure 1 This is a node distribution map of a shortwave communication network in a certain area. The map shows the problems that exist when using traditional shortwave radios to form a network on the same frequency: the distance between node SB and nodes CM and JS exceeds 50 kilometers, making it difficult for these two remote nodes to join the network, and network connectivity cannot be guaranteed.

[0005] To address the challenges of beyond-line-of-sight (BLS) networking, existing technologies typically employ two approaches: The first abandons the advantage of autonomous frequency selection, reverting to the traditional method of assigning fixed frequencies to the entire network. This approach sacrifices the high reliability of autonomous frequency selection and cannot adapt to the rapid changes in shortwave channel interference. The second approach relies on wired infrastructure such as fiber optic communication networks for auxiliary networking, connecting long-distance shortwave nodes via wired connections. For example... Figure 2 As shown, this hybrid networking approach increases the complexity of the system and its dependence on wired facilities, limiting the application of shortwave communication in environments without infrastructure.

[0006] The fundamental reason lies in the lack of a dedicated network architecture and method that can adapt to the asymmetric frequency characteristics of shortwave autonomous frequency selection technology and effectively organize networks and balance performance over a wide area. Therefore, there is an urgent need to design a new network architecture to solve the problems of how to effectively organize widely distributed radio stations and how to optimize network setup time and information transmission latency under this architecture. Summary of the Invention

[0007] Purpose of the invention: The purpose of this invention is to address the technical problems of difficulty in networking shortwave autonomous frequency selective radio stations and large transmission delays in the prior art, and to provide a shortwave autonomous frequency selective radio station networking architecture method and network control equipment.

[0008] Technical Solution: The present invention provides a shortwave autonomous frequency-selective radio station networking architecture method, applied to a communication network composed of M shortwave autonomous frequency-selective radio stations. The method includes the following steps:

[0009] S1: Construct the M radio stations into a multi-level tree network architecture, which consists of a primary network and multiple subnets, wherein the subnet topology includes star network and / or mesh network.

[0010] S2: The M radio stations are divided using a non-uniform partitioning method, and the node affiliation and topology of each subnet are determined according to the geographical distribution characteristics of the radio stations and / or the organizational structure of the network user departments.

[0011] S3: The number of nodes in each level of subnet is limited, wherein the number of nodes in the first-level subnet and the intermediate subnets does not exceed a first preset threshold N1; the number of nodes in the last-level subnet does not exceed a second preset threshold N2, and N2≥N1;

[0012] S4: For the handover node that serves as a slave node of the upper-level network and a master node of the lower-level network, time division multiple access or frequency division multiple access is adopted to keep it online in two or more networks at the same time.

[0013] To further improve the above technical solution, the first preset threshold N1 is 5, and the second preset threshold N2 is 10.

[0014] Furthermore, when the last-level subnet adopts a star network structure, its number of nodes does not exceed 5; when the last-level subnet adopts a mesh network topology, its number of nodes does not exceed 10.

[0015] Furthermore, the specific steps for constructing the non-uniform partitioning include: determining the total number of nodes M and the initial network level L1; dividing the first-level network into a star network with 2 to 4 slave nodes; progressively dividing the subnets into subnets with 2 to 4 slave nodes, up to the final subnet.

[0016] Furthermore, the network is constructed using the non-uniform segmentation method to balance network setup time and information transmission delay; the method also includes an estimation step for information transmission delay between any nodes, which includes:

[0017] (1) Determine the transmission path of a message in the network, and the number of network levels traversed by this path is: ;

[0018] (2) Estimate the maximum transmission delay of the path. and minimum delay ,in:

[0019] If the end of the path is located in the last subnet of a star topology, the maximum transmission delay is... ,in Let be the number of slave nodes in the i-th level subnet along the path; the minimum latency is . ;

[0020] If the end of the path is located in the last subnet of a mesh topology, the maximum transmission delay is... The minimum latency is ;

[0021] (3) According to the formula The average transmission delay was calculated.

[0022] Furthermore, it also includes a dynamic refactoring step:

[0023] Real-time monitoring of the maximum transmission latency of each path in the network With minimum delay The ratio;

[0024] When detected When the network topology is reconfigured, it automatically triggers a re-partitioning operation for subnets with too many nodes.

[0025] The present invention also provides a network control device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor is configured to implement the above-described shortwave autonomous frequency-selective radio station networking architecture method when executing the program.

[0026] Furthermore, the memory also stores the basic data required to support the construction of the network using the non-uniform partitioning method. This basic data includes: node geographic information, distance data between nodes, and organizational structure information of the network-using departments.

[0027] Beneficial effects: Compared with the prior art, the advantages of the present invention are as follows:

[0028] Reduce network setup time: By constructing in a hierarchical manner, a large-scale network is decomposed into multiple smaller subnets, which reduces the networking complexity and scale of individual subnets, making subnet setup faster and thus significantly shortening the overall network setup time.

[0029] Improved networking efficiency: Subnets at each level can be established independently and in parallel, further improving the overall networking efficiency; the multi-level architecture provides multiple redundant paths, so a single point of failure will not cause the entire network to crash, thus improving network reliability.

[0030] Balancing and reducing transmission latency: By employing a multi-level tree architecture, especially one optimized for non-uniform partitioning, the path length and number of relays for message transmission within the network can be effectively balanced. Compared to traditional architectures divided by administrative levels, this approach, through reasonable network hierarchy and node number limitations, achieves an effective balance between network establishment time and information transmission latency, significantly reducing the average end-to-end transmission latency.

[0031] Enhancing network usability: This invention provides a pure shortwave networking solution that does not rely on wired infrastructure such as fiber optics, greatly expanding the application scenarios of shortwave autonomous frequency-selective radios in environments without infrastructure, such as emergency communication and field operations.

[0032] High adaptability: The non-uniform segmentation method can be flexibly configured according to different geographical shapes and organizational structure characteristics, and is suitable for various regional shapes such as parallelograms, semicircles, and hexagons; the tree structure facilitates network expansion, and when adding a new node, it only needs to be added to the corresponding subnet without affecting the overall network architecture, and has good scalability.

[0033] Fully leverage the advantages of autonomous frequency selection: Unlike existing technologies, this invention fully utilizes the asymmetric frequency configuration characteristics of shortwave autonomous frequency selection technology, without having to abandon autonomous frequency selection and revert to traditional methods. Attached Figure Description

[0034] Figure 1 This is a schematic diagram of the node distribution when a shortwave communication network in a certain area adopts traditional co-frequency networking.

[0035] Figure 2 This is a schematic diagram of an existing large-area shortwave radio network architecture that relies on fiber optic networks and fixed shortwave stations.

[0036] Figure 3 These are example diagrams of two single-network architectures, in which Figure 3 In this context, 'a' represents a star-shaped mesh. Figure 3 In this context, 'b' represents a mesh network.

[0037] Figure 4 This is an example diagram of a binary tree network architecture;

[0038] Figure 5 This is an example diagram of a ternary tree network architecture;

[0039] Figure 6 This is an example diagram of a non-uniformly partitioned tree network architecture;

[0040] Figure 7 This is an example diagram illustrating the equipment configuration method for intermediate nodes of a shortwave autonomous frequency-selective radio network, which employs frequency division multiple access.

[0041] Figure 8This is an example diagram illustrating the equipment configuration method for intermediate nodes of a shortwave autonomous frequency-selective radio network, which employs time-division multiple access.

[0042] Figure 9 This is an example diagram of a star network used in the final-level network.

[0043] Figure 10 This is an example diagram of a mesh network used in the final-level network.

[0044] Figure 11 It is a spatial distribution relationship of a three-level regional network (first-level and second-level networks).

[0045] Figure 12 It is a three-level network topology diagram of regions based on natural organizational structure.

[0046] Figure 13 This is a three-level network topology diagram for regions with balanced latency.

[0047] Figure 14 This is a schematic diagram of the regional node distribution and the division of primary and secondary networks.

[0048] Figure 15 This is a topology diagram of a three-tier regional network.

[0049] Figure 16 This is an example diagram of network partitioning for a secondary shortwave autonomous frequency selection radio station using traditional methods.

[0050] Figure 17 This is an example diagram of network partitioning for a secondary shortwave autonomous frequency-selective radio station using the method of this invention;

[0051] Figure 18 This is the network architecture diagram after the partitioning method of this invention. Detailed Implementation

[0052] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings, but the scope of protection of the present invention is not limited to the embodiments described.

[0053] Example 1: The overall network architecture proposed in this invention is a multi-level tree network. Due to the asymmetric frequency configuration characteristic of shortwave autonomous frequency-selective radio communication—that is, the optimal transmit and receive frequencies at both ends of the link are likely to be different—building a large-scale fully connected mesh network requires [further steps] during the link establishment phase. The time cost of building a chain is huge, making it impractical; while Star reduces the number of chain building times to N-1, the scale of a single network is limited.

[0054] Therefore, this invention employs a hierarchical interconnection method for multiple subnets to form a tree-like architecture. This architecture consists of a primary network (root network) and several levels of subnets. The basic topology of the subnets can be a star network or a mesh network. Figure 3As shown in Figure a, in a star network, a radio station is designated as the central node, and it communicates with each slave node through a polling method; for example... Figure 3 As shown in b, mesh networks are suitable for nodes that are geographically close. For example, in the final subnet, the distance between nodes is within the ground wave transmission range (usually no more than 50 kilometers). In this case, there is a high probability that co-frequency networking can be achieved, resulting in shorter transmission delays.

[0055] For a communication network with a total of M nodes, how to hierarchically divide and subnet the network is crucial for balancing performance. Compared to uniform partitioning, such as... Figure 4 The binary dendritic network shown is uniformly divided. Figure 5 The ternary tree network (uniformly divided) shown in this invention represents the hierarchical network architecture proposed in this invention. Figure 6 The non-uniform segmentation construction method shown is better suited to situations where node distribution is uneven and organizational structures vary in practical applications.

[0056] The specific steps for constructing a non-uniform partition are as follows:

[0057] Determine the overall parameters: Determine the total number of nodes M based on communication requirements, and preliminarily determine the network level L1 based on the organizational structure within the region.

[0058] First-level network division: Using a star network architecture, the root node (central node) and slave nodes of the first-level network are determined based on the node distribution characteristics within the region. The number of slave nodes is usually controlled between 2 and 4. At this point, the total number of nodes in the first-level network does not exceed 5 (1 central node and a maximum of 4 slave nodes). The rationale for this limitation is that communication between the central node and each slave node in a star network mainly uses a round-robin method, and the number of slave nodes should not be too large to ensure that the polling cycle is not too long.

[0059] Subnetting is performed hierarchically: Each slave node in the first-level subnet becomes the central node of the next-level (second-level) subnet. Slave nodes are assigned to it in a star topology, and the total number of nodes in each second-level subnet is limited to no more than 5. This process continues hierarchically downwards until all nodes are assigned to a subnet.

[0060] Designing the final-level subnet: The node count limit for the final-level subnet (the subnet located at the end of the tree structure) can be appropriately relaxed, with its total number of nodes limited to no more than 10. This is because the coverage area of ​​the final-level subnet is usually small, and there is a high probability that it is within the ground wave transmission range, enabling same-frequency or near-same-frequency networking, greatly simplifying the network construction process. If the final-level subnet adopts a star network structure, the total number of nodes should preferably not exceed 5; if a mesh network topology is adopted, the total number can be relaxed to no more than 10.

[0061] Through the above steps, a large-scale network is constructed into a multi-level tree network with clear hierarchy and controllable scale at each level. Figure 6An example of a non-uniformly segmented tree network is shown, based on Figure 6 Node B1 transmits a message to node B6 to indicate the transmission delay. , , The calculation method and balancing effect of ().

[0062] In a multi-level network architecture, a key node (called a handover node) is both a slave node in its upper-level network and a master node in its lower-level network. Since the two networks it participates in (the upper-level network and the lower-level network) are likely operating at different frequencies, this invention proposes two implementation mechanisms to ensure that the node can remain online in both networks simultaneously:

[0063] Frequency Division Multiple Access (FDMA): such as Figure 7 As shown, this handover node is configured with two or more independent transceiver units (shortwave radios). One transceiver unit operates on the frequency (F1) of the upper-level network, while the other operates on the frequency (F2) of the local network. Information routing and forwarding are achieved through the network management platform and switches, thereby realizing physical simultaneous online operation.

[0064] Time Division Multiple Access (TDMA): such as Figure 8 As shown, this handover node is equipped with only one transceiver unit. During operation, it rapidly switches between the working states of the upper-level network and the local network according to a preset time slot. Through time-sharing operation, it achieves logical simultaneous online operation.

[0065] The hierarchical network architecture proposed in this invention inherently has the function of balancing transmission latency. By hierarchically dividing the network and limiting the size of each subnet, situations where some paths are too long or there are too many relay nodes can be avoided.

[0066] Once the network is operating stably, the information transmission delay between any two nodes can be estimated. Taking the downlink transmission from the primary master station to the target node B as an example, assume that the transmission time for a single link is 1 unit of time.

[0067] The message transmission path passed through There are several levels, each level (the first...) The number of slave nodes in the (level) subnet is .

[0068] Maximum delay Within each subnet, due to the use of a round-robin mechanism, a message may need to wait for a maximum of [number] seconds. It takes a certain number of units of time for a message to be forwarded to the correct slave node. The total maximum latency is the sum of the number of slave nodes in each subnet: .

[0069] Shortest latency In the ideal scenario, messages are forwarded directly at each level without waiting, with each level taking one unit of time. The total shortest latency equals the number of network levels. .

[0070] Average latency Average latency can be estimated as the arithmetic mean of the maximum and minimum latency. .

[0071] Specifically, when the message destination is located in a final-level subnet using a mesh topology, the transmission delay after entering that subnet is 1 unit of time. In this case, the formula for calculating the total transmission delay is adjusted as follows:

[0072] Maximum latency: ;

[0073] Shortest latency: ;

[0074] Uplink transmission is the symmetrical inverse process of downlink transmission, and its delay estimation result is consistent with that of downlink. Transmission between any two non-directly connected nodes can be decomposed into an uplink segment and a downlink segment, and the total delay can be obtained by calculating them separately.

[0075] by Figure 6 For example, the path and time delay for node B1 to transmit a message to node B6 are as follows:

[0076] In the uplink path: the maximum latency for transmission from node B1 to node B2 is 4 seconds, and the minimum latency is 1 second; in the transmission from node B2 to node B3: the maximum latency is 4 seconds, and the minimum latency is 1 second.

[0077] In the downlink path: the maximum latency for transmission from node B3 to node B4 is 4, and the minimum latency is 1; the maximum latency for transmission from node B4 to node B5 is 2, and the minimum latency is 1; the maximum latency for transmission from node B5 to node B6 is 4, and the minimum latency is 1.

[0078] Path from node B1 to B6: ;

[0079] ;

[0080] ;

[0081] Refactoring trigger illustration: When At that time, subnet splitting is triggered.

[0082] Figure 9 and Figure 10 These are two examples of hybrid network architectures. Figure 9 This is an example of a star network used in the final-level network; Figure 10This is an example of a terminal network containing a mesh network, demonstrating the specific form of a multi-level network formed by the mixed expansion of star networks and mesh subnets.

[0083] Example 2: Figure 11 As shown, for a parallelogram region with an area of ​​over 100,000 square kilometers, there are multiple nodes, and the original co-frequency network has poor communication performance.

[0084] like Figure 12 As shown, using the traditional approach: if divided according to the administrative organizational structure, the primary network (centered on SB) has as many as 13 slave nodes. The estimated average message transmission delay from remote node X1 to X3 is... It is 18 units of time.

[0085] like Figure 13 As shown, the proposed solution uses a non-uniform partitioning method to break down administrative affiliations and optimize the division based on geographical proximity. The specific implementation steps are as follows:

[0086] S1: Construct a multi-level tree-like network architecture

[0087] SB is the primary network center node;

[0088] XZ, YH, SZ, and NJ are selected as the four slave nodes of the first-level network to form a star network structure;

[0089] Each slave node serves as the central node of the secondary network and continues to expand downwards.

[0090] S2: Non-uniform partitioning method

[0091] Based on the geographical distribution, the entire region is divided into 4 sub-regions;

[0092] Each sub-region forms a secondary network, with its central node being the corresponding primary network slave node;

[0093] The network is further divided into three levels.

[0094] S3: Node Quantity Limit

[0095] Level 1 network: 1 central node plus 4 slave nodes, a total of 5 nodes, satisfying the restriction of no more than 5 nodes;

[0096] The number of nodes in each subnet is controlled within a preset threshold.

[0097] The number of nodes in the final-level network does not exceed 10.

[0098] S4: Handover Node Processing

[0099] Each primary network node is both a slave node of the upper-level network and a master node of the lower-level network;

[0100] It adopts a time-division multiple access method to switch working states between different networks in a time-division manner;

[0101] After optimization, the average delay for the same path (X1 to X3) is reduced. Reduced to 12 units of time, resulting in a 33% performance improvement.

[0102] Table 1 provides Figure 11 The distance data between nodes (unit: km) in this set of maps, combined with specific geographical areas, intuitively compare the differences between the traditional administrative structure and the optimized architecture of this invention, and verify the effect of latency balancing.

[0103] Table 1: Figure 11 Distances between nodes (unit: km)

[0104] ;

[0105] Example 3: For a semi-circular region with a diameter of approximately 980 kilometers, where the nodes are more widely distributed and contain 22 sub-blocks, the method of the present invention is also applicable. For example... Figure 14 As shown, based on the radial geographical features, a primary network is constructed with ZX as the center and four slave nodes YJ, GZ, HY, and SW. The entire region is then divided into four secondary network regions: A, B, C, and D. Figure 15 The final three-level network topology is shown, which effectively controls the size of each subnet, thereby balancing and reducing overall transmission latency.

[0106] S1: Construct a tree-like network architecture

[0107] ZX is selected as the primary network central node; YJ, GZ, HY, and SW are selected as the four secondary nodes of the primary network.

[0108] S2: Radial division method

[0109] To match the semi-circular shape, a radial division is adopted; the region is divided into 4 sector regions, each corresponding to a secondary network; each secondary network continues to expand downward to form a tertiary network.

[0110] S3: Node Count Control

[0111] Strictly limit the number of nodes at each level of the network according to the preset threshold; ensure that network establishment time and transmission latency are within an acceptable range;

[0112] This architecture effectively reduces overall transmission latency and adapts to the geographical characteristics of a semi-circular region. Table 2 provides distance data (unit: km) between first- and second-level network nodes in this region, demonstrating the method of applying the architecture of this invention to regions with different geographical shapes (semi-circular, approximately 980 km in diameter).

[0113] Table 2: Distances between first- and second-level network nodes (unit: km)

[0114] ;

[0115] Example 4: For a hexagonal area of ​​about 100km north-south and about 70km east-west, although the range is relatively small, traditional co-frequency networking still has the problem of nodes being difficult to join the network.

[0116] like Figure 16 Using the traditional method of dividing the area vertically into three equal parts will result in too many subnet nodes in the middle B area (e.g., more than 5), causing excessive polling latency within the subnet and uneven network performance.

[0117] like Figure 17 As shown, the network is partitioned using the method provided by this invention, and the resulting network architecture is as follows. Figure 18 As shown, the specific steps are as follows:

[0118] S1: Network Architecture Design

[0119] SB is selected as the primary network central node; BS, QP, PD, and FX are selected as primary network slave nodes.

[0120] S2: Region Division Optimization

[0121] The hexagonal region is divided into four relatively balanced sub-regions; each sub-region forms a two-level network with no more than four nodes.

[0122] S3: Topology Selection

[0123] Choose an appropriate topology based on the geographical distribution characteristics of each subnet; the final-level network can adopt a star network or mesh network structure.

[0124] The architecture derived by the method of this invention divides the region into four secondary networks (A, B, C, and D regions), ensuring that the number of slave nodes in each subnet does not exceed 4. The corresponding network architecture is as follows: Figure 18 As shown, this architecture exhibits lower transmission latency and better uniformity. Table 3 provides distance data (unit: km) between nodes in the hexagonal region, illustrating the advantages of the non-uniform partitioning construction method and the importance of node number constraints through positive and negative case comparisons.

[0125] Table 3: Distances between nodes in the hexagonal region (unit: km)

[0126] ;

[0127] In summary, this invention, by proposing a multi-level tree-like network architecture and combining methods such as non-uniform partitioning, node number constraints, and multiple access of handover nodes, successfully solves the core technical challenges of shortwave autonomous frequency-selective radio stations in wide-area networking. The technical solution it provides has significant practical value and superior performance.

[0128] As described above, although the invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A shortwave autonomous frequency-selective radio station networking architecture method, applied to a communication network composed of M shortwave autonomous frequency-selective radio stations, characterized in that, Includes the following steps: S1: Construct the M radio stations into a multi-level tree network architecture, which consists of a primary network and multiple subnets, wherein the subnet topology includes star network and / or mesh network. S2: The M radio stations are divided using a non-uniform partitioning method. The node affiliation and topology of each level of subnet are determined based on the geographical distribution characteristics of the radio stations and / or the organizational structure of the network users. The specific steps of the non-uniform partitioning include: determining the total number of nodes M and the initial network level L1; dividing the first-level network into a star network with 2 to 4 slave nodes; dividing the subnets level by level with 2 to 4 slave nodes until the final subnet. S3: Limit the number of nodes in each level of subnet, wherein the number of nodes in the first-level subnet and intermediate subnets does not exceed the first preset threshold N1; the number of nodes in the last-level subnet does not exceed the second preset threshold N2, and N2≥N1; the first preset threshold N1 is 5, and the second preset threshold N2 is 10; S4: For the junction node that serves as a slave node of the upper-level network and a master node of the lower-level network, time division multiple access or frequency division multiple access is adopted to keep it online in two or more networks at the same time; when the last-level subnet adopts a star network structure, its number of nodes does not exceed 5; when the last-level subnet adopts a mesh network topology, its number of nodes does not exceed 10. The method further includes an estimation step for information transmission delay between any nodes, which includes: (1) Determine the transmission path of a message in the network, and the number of network levels traversed by this path is: ; (2) Estimate the maximum transmission delay of the path. and minimum delay ,in: If the end of the path is located in the last subnet of a star topology, the maximum transmission delay is... ,in Let be the number of slave nodes in the i-th level subnet along the path; the minimum latency is . ; If the end of the path is located in the last subnet of a mesh topology, the maximum transmission delay is... The minimum latency is ; (3) According to the formula The average transmission delay was calculated. It also includes a dynamic refactoring step: Real-time monitoring of the maximum transmission latency of each path in the network With minimum delay The ratio; When detected When the network topology is reconfigured, it automatically triggers a re-partitioning operation for subnets with too many nodes.

2. A network control device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor is configured to execute the program, it implements the shortwave autonomous frequency selection radio station networking architecture method of claim 1.

3. The network control device according to claim 2, characterized in that, The memory also stores the basic data required to support the construction of the network using the non-uniform segmentation method. The basic data includes: node geographic information, distance data between nodes, and organizational structure information of the network user departments.

Images