A low-latency self-organizing network communication method and system

By analyzing network topology information in real time and evaluating dynamic changes in communication paths using machine learning models, intelligently adjusting the routing update frequency, solving the problem that the routing table update frequency in the existing technology cannot be followed up in time, and improving the low-latency ad hoc network Real-time and reliability of communications, suitable for mission-critical communications.

CN118804205BActive Publication Date: 2024-11-22SHENZHEN YIXINSHUN ELECTRONIC CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411252306.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-09-09
Publication Date
2024-11-22
Estimated Expiration
2044-09-09

AI Technical Summary

Technical Problem

In the existing low-latency ad hoc network communication, the routing table update frequency cannot follow up on node status changes in time, resulting in network splits and data loss, or the update frequency is too high to occupy resources, affecting network performance.

Method used

By collecting and analyzing network topology information in real time, using machine learning models to evaluate the dynamic changes in communication paths, intelligently adjust the routing update frequency, and increase or decrease the update frequency according to high-dynamic or low-dynamic paths.

Benefits of technology

It effectively avoids network splits and data loss caused by lag in routing information, saves network bandwidth and computing resources, improves the real-time and reliability of communication, is suitable for mission-critical communication scenarios, and significantly improves network performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118804205B_ABST
    Figure CN118804205B_ABST
Patent Text Reader

Abstract

The present invention discloses a low-latency self-organizing network communication method and system, comprising the following steps: collecting current topology information from each node in the network, analyzing the collected topology information, and drawing a topology diagram of the current network; extracting dynamic change features of the nodes on each communication path, and using a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features. The present invention collects and analyzes network topology information in real time, and uses a machine learning model to evaluate the dynamic changes of communication path nodes. It can intelligently adjust the routing update frequency of different paths, increase the update frequency in high dynamic paths, and avoid network splitting and data loss caused by routing lag; reduce the frequency in low dynamic paths to save resources and reduce performance waste, effectively improving the real-time and reliability of communication.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of network communication technology, and in particular to a low-latency self-organizing network communication method and system. Background Art

[0002] Low-latency ad hoc communication refers to a wireless communication network architecture in which network nodes can automatically form, manage and maintain network connections without relying on fixed infrastructure. Low latency means that the time it takes to transmit data from one node to another is very short, which is critical for application scenarios that require real-time or near-real-time communication, such as drone formations, emergency rescue, or battlefield communications. The characteristics of ad hoc networks are that the network can respond quickly and flexibly in a dynamically changing environment, and can cope with frequent movement or failure of nodes. Through low-latency ad hoc communication, the network can ensure the stability and reliability of communication while ensuring efficient data transmission, and adapt to complex application requirements.

[0003] In low-latency self-organizing network communications, automatic network node formation means that each node can autonomously discover and connect to other nodes to form a network without relying on fixed infrastructure or central controllers. Each node has routing and forwarding capabilities. Through dynamic routing selection and distributed control, nodes can automatically select the best path for communication based on the real-time status of the network (such as node location, signal strength, load, etc.). In this way, the network can quickly adapt to the movement or changes of nodes, ensuring that data can be transmitted through the optimal path with the lowest latency, thereby realizing an efficient and flexible communication network.

[0004] Dynamic routing protocols need to continuously update routing tables to reflect the latest topology of the network. Existing technologies usually use a timed update method to refresh the node status on each communication path, which can meet basic needs in most cases. However, when the node status in the communication network changes rapidly and the routing table update frequency cannot keep up in time, serious consequences may occur. First, the lag in routing information may cause some nodes to lose the ability to communicate with other nodes due to unreachable paths, which in turn causes the network to split and form multiple subnets that cannot communicate with each other. This network partition will hinder the transmission of data throughout the network and seriously affect the overall function of the network, especially in mission-critical communications. Second, outdated routing information may cause data packets to be directed to invalid paths or non-existent nodes, causing data packet loss, which not only reduces the reliability of communication, but also may cause permanent loss of important data. When the node status in the communication network changes slowly and the routing table is updated frequently, frequent routing updates consume network bandwidth and computing resources. If the node status changes slowly and the update frequency is too high, unnecessary network resources and computing power will be occupied, resulting in insufficient resources for other critical tasks and affecting the overall network performance.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a low-latency self-organizing network communication method and system, which can intelligently adjust the routing update frequency of different paths by collecting and analyzing network topology information in real time and using machine learning models to evaluate the dynamic changes of communication path nodes, increase the update frequency in high-dynamic paths, and avoid network splitting and data loss caused by routing lags; reduce the frequency in low-dynamic paths to save resources and reduce performance waste, effectively improving the real-time and reliability of communication, and is particularly suitable for mission-critical communication scenarios, significantly enhancing the overall performance of the network to solve the problems in the above-mentioned background technology.

[0007] In order to achieve the above object, the present invention provides the following technical solution: a low-latency self-organizing network communication method, comprising the following steps:

[0008] Collect the current topology information from each node in the network, analyze the collected topology information, and draw a topology diagram of the current network;

[0009] Extract dynamic change features of nodes on each communication path, and use a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features;

[0010] According to the results of intelligent evaluation, all communication paths are divided into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths;

[0011] For the communication paths divided into stable changing paths, a standard node status update frequency is set, and the status of the communication nodes is updated at the standard node status update frequency;

[0012] For high-dynamic paths and low-dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency based on the standard node status update frequency based on real-time dynamic change information.

[0013] Preferably, the topology diagram of the current network is drawn, and the specific steps are as follows:

[0014] Start the state collection mechanism on each node in the network;

[0015] After the status collection mechanism is started, all nodes send their collected information to one or more central nodes for aggregation;

[0016] After data aggregation and storage are completed, the collected node status information begins to be analyzed;

[0017] Based on the analysis results, use network visualization tools or dedicated software to draw the current network topology diagram.

[0018] Preferably, dynamic change features of the nodes on each communication path are extracted, and the extracted features include the rate of change of the number of hops between the nodes on the communication path and the rate of change of the energy consumption of the nodes on the communication path. After acquisition, the rate of change of the number of hops between the nodes on the communication path and the rate of change of the energy consumption of the nodes on the communication path are analyzed under the monitoring window, and a hopping index and a node energy consumption index are generated respectively. The hopping index and the node energy consumption index are input into a pre-trained machine learning model, and a path node dynamic change coefficient is generated by the machine learning model. The path node dynamic change coefficient is used to comprehensively evaluate the node change status on each communication path.

[0019] Preferably, the hop count change rate between nodes on the communication path is analyzed to generate a hop index, and the specific steps are as follows:

[0020] In the monitoring window, the number of hops between each pair of nodes on the communication path is collected in real time, and the time interval is set. As the sampling period, Indicates at a point in time When , the number of hops between nodes on the communication path;

[0021] Calculate the hop count change rate between adjacent sampling time points , the calculation expression is: ,in, Indicates at a point in time The hop count change rate at that time, k is an adjustment factor used to control the sensitivity of the change rate, and i represents the current moment in the time series;

[0022] In order to further analyze the cumulative effect of the hop count change, we calculate the The cumulative effect of , the calculation expression is: ,in, Indicates the time point The cumulative change effect at time point j The time point index within;

[0023] Cumulative change effect Calculate the jump index between each pair of nodes. The calculation expression is: , where represents the jump index between each pair of nodes, and n represents the number of sampling times in the monitoring window;

[0024] Comprehensively analyze the communication path and calculate the jump index. The calculation expression is: , where represents the hopping index between the wth pair of nodes, m represents the total number of nodes on the communication path, Indicates the hopping index of the communication path.

[0025] Preferably, the node energy consumption index is generated after analyzing the speed change of node energy consumption on the communication path, and the specific steps are as follows:

[0026] For each node on the communication path, calculate its instantaneous energy consumption rate. Suppose each node is and The remaining energies are and , then the instantaneous energy consumption rate The calculation expression is: , where Indicates the node at a point in time The remaining energy when Indicates the node at a point in time The processing power of Indicates the node at a point in time Working voltage at

[0027] Calculate the cumulative energy consumption coefficient of the node , the cumulative energy consumption coefficient reflects the cumulative energy consumption of the node within the monitoring window, and the calculation expression is: , where represents the cumulative energy consumption coefficient of the node within the monitoring window, Indicates at a point in time The instantaneous energy consumption rate at Indicates at a point in time The transmission time is the duration of the node communication activity. Indicates the node at a point in time The current working frequency at , which indicates the complexity of the node processing task;

[0028] Calculate the sensitivity of energy consumption changes , the energy consumption change sensitivity reflects the sensitivity of the energy consumption rate to time changes, and the calculation expression is: , where represents the time rate of change of the cumulative energy consumption coefficient, and They represent the time change rates of node processing power and operating voltage respectively;

[0029] The node energy consumption coefficient of each node is calculated by comprehensively considering the energy consumption change sensitivity, cumulative energy consumption coefficient and instantaneous energy consumption rate. The calculation expression is: , where represents the node energy consumption coefficient of each node, and T represents the total duration of the monitoring window;

[0030] Calculate the node energy consumption index, the calculation expression is: , where represents the node energy consumption coefficient of the f-th node, Indicates the node energy consumption index of the communication path.

[0031] Preferably, the dynamic change coefficient of the path node generated after comprehensive analysis of the nodes on the communication path under the monitoring window is compared and analyzed with a preset reference range of the dynamic change coefficient of the path node, and the communication path is divided. The specific steps of division are as follows:

[0032] If the path node dynamic change coefficient is within the path node dynamic change coefficient reference range, the communication path is divided into a stable change path;

[0033] If the path node dynamic change coefficient is less than the minimum value of the path node dynamic change coefficient reference range, the communication path is classified as a low dynamic path;

[0034] If the path node dynamic change coefficient is greater than the maximum value of the path node dynamic change coefficient reference range, the communication path is classified as a high dynamic path.

[0035] Preferably, for high dynamic paths and low dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency as follows:

[0036] Define fuzzy sets and membership functions for jump index, node energy consumption index, and standard node status update frequency;

[0037] Based on fuzzy sets, a fuzzy rule base is constructed. The rule form of the rule base is: "If" the jump index and the node energy consumption index belong to a certain fuzzy set, the actual node status update frequency will be adjusted to a certain level;

[0038] The real-time jump index and node energy consumption index are input into the fuzzy reasoning engine, and fuzzy reasoning is performed in combination with the defined fuzzy rule base. The fuzzy reasoning process is executed through the "if-then" rule to determine the update frequency that best suits the current path conditions. The reasoning formula is: , where represents the fuzzy reasoning process, and They represent the membership functions of the jump index and node energy consumption index respectively, Indicates fuzzy output;

[0039] The fuzzy output obtained by fuzzy inference is analyzed using the centroid method Defuzzification is performed to generate the fuzzy actual node state update frequency. The defuzzification formula is: , is the membership function used in the defuzzification process, is the update frequency of the fuzzy actual node state obtained after defuzzification;

[0040] Update frequency according to the fuzzy actual node state after defuzzification , adjust the actual node status update frequency of the final application, and the adjustment formula is as follows: ,in, is the node status update frequency of the final application, Indicates the standard node status update frequency.

[0041] A low-latency self-organizing network communication system, comprising a topology information collection and analysis module, a dynamic change feature extraction and evaluation module, a communication path classification module, a standard frequency update module and a fuzzy logic control module;

[0042] The topology information collection and analysis module collects the current topology information from each node in the network, analyzes the collected topology information, and draws a topology diagram of the current network;

[0043] The dynamic change feature extraction and evaluation module extracts dynamic change features of nodes on each communication path, and uses a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features;

[0044] The communication path classification module divides all communication paths into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths, based on the results of intelligent evaluation;

[0045] A standard frequency update module, for a communication path divided into a stable change path, sets a standard node status update frequency, and updates the status of the communication node at the standard node status update frequency;

[0046] The fuzzy logic control module adopts fuzzy logic control for high dynamic paths and low dynamic paths. Based on the real-time dynamic change information, it intelligently increases or decreases the actual node status update frequency on the basis of the standard node status update frequency.

[0047] In the above technical solution, the technical effects and advantages provided by the present invention are:

[0048] The present invention collects and analyzes network topology information in real time, accurately extracts the dynamic change characteristics of nodes on the communication path, and based on these characteristics, evaluates the dynamic changes of nodes on the communication path through a machine learning model, and can effectively identify and distinguish high dynamic, stable change and low dynamic paths, thereby taking targeted routing update strategies for different types of paths. In a high dynamic path, by improving the node status update frequency, the changes in the network topology can be reflected in a timely manner, avoiding network splitting and data packet loss caused by routing information lag. In a low dynamic path, reducing the update frequency effectively reduces the consumption of network bandwidth and computing resources, and avoids the performance waste caused by frequent updates. Through this intelligent adjustment method, not only the real-time and reliability of communication are enhanced, but also the efficient operation of the network is maintained while saving resources. It is particularly suitable for mission-critical communication scenarios and plays a significant role in improving the overall performance of the network. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0050] Figure 1 The present invention is a method flow chart of a low-latency self-organizing network communication method.

[0051] Figure 2 A schematic diagram of a module of a low-latency self-organizing network communication system of the present invention. DETAILED DESCRIPTION

[0052] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be implemented in a variety of forms and should not be construed as limited to the examples set forth herein; rather, these example embodiments are provided so that the description of the present disclosure will be more comprehensive and complete, and the concept of the example embodiments will be fully conveyed to those skilled in the art.

[0053] The present invention provides Figure 1 A low-latency self-organizing network communication method shown includes the following steps:

[0054] Collect the current topology information from each node in the network, analyze the collected topology information, and draw a topology diagram of the current network;

[0055] Draw the topology diagram of the current network. The specific steps are as follows:

[0056] Start the state collection mechanism on each node in the network;

[0057] This mechanism collects the node's own status information and the connection status with other neighboring nodes periodically or based on event triggering. This status information includes the node's geographical location, signal strength, connection bandwidth, latency, node activity status, etc. The connection information between nodes is usually exchanged through link state protocols (such as OSPF) or distance vector protocols (such as RIP). Starting the status collection mechanism is the basis of the entire process, ensuring that the latest and accurate topology information is obtained from the network.

[0058] After the status collection mechanism is started, all nodes send their collected information to one or more central nodes for aggregation;

[0059] These central nodes can be specific nodes in the network or distributed database systems that are used to centrally store and process the collected status information. In this step, the integrity and timeliness of the data must be ensured so that the central node can accurately reflect the topology of the entire network. For large-scale networks, data preprocessing and compression may also be required for efficient storage and analysis.

[0060] After data aggregation and storage are completed, the collected node status information begins to be analyzed;

[0061] This analysis process includes evaluating the connection status between nodes, determining the quality of the connection, and predicting the possible movement trend of the nodes. Through these analyses, it is possible to identify which connections are stable and which connections may change in the future. Machine learning algorithms can also be used to conduct in-depth analysis of node status data to discover potential topological structure change trends and prepare for the next drawing step.

[0062] Based on the analysis results, use network visualization tools or dedicated software to draw the current network topology diagram;

[0063] A topology diagram usually displays network nodes and their connection status in a graphical way, including key parameters such as node location, connection bandwidth, and latency. The drawn topology diagram is not only an intuitive display of the current network status, but also an important reference for network management and routing optimization. This diagram can be updated in real time in the network monitoring system, helping administrators and automation systems make quick decisions and optimize the overall performance of the network.

[0064] Extract dynamic change features of nodes on each communication path, and use a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features;

[0065] The dynamic change features of the nodes on each communication path are extracted. The extracted features include the hop count change rate between the nodes on the communication path and the speed change of the energy consumption of the nodes on the communication path. After acquisition, the hop count change rate between the nodes on the communication path and the speed change of the energy consumption of the nodes on the communication path are analyzed under the monitoring window, and the hop index and the node energy consumption index are generated respectively. The hop index and the node energy consumption index are input into the pre-trained machine learning model, and the path node dynamic change coefficient is generated by the machine learning model. The path node dynamic change coefficient is used to comprehensively evaluate the node change status on each communication path.

[0066] A high rate of change in the number of hops between nodes on a communication path usually indicates that the node status of the communication path is changing rapidly. The number of hops (i.e., the number of intermediate nodes from the source node to the destination node) directly reflects the length and complexity of the path. When the number of hops changes frequently, it means that nodes on the path may constantly join or exit the communication link due to movement, failure, signal attenuation, etc. This change causes the routing protocol to frequently update the path selection, thereby increasing the computational overhead and latency of the network. In addition, frequent changes in the number of hops may also cause data packets to switch between different paths, increasing the uncertainty and packet loss rate of data transmission. Therefore, a high rate of change in the number of hops indicates that the node status on the path is in an unstable or rapidly changing state, which can have a negative impact on the overall performance and reliability of the path, especially in real-time communications or mission-critical applications.

[0067] In the monitoring window, the hop count change rate between nodes on the communication path is analyzed to generate a hop index. The specific steps are as follows:

[0068] In the monitoring window, the number of hops between each pair of nodes on the communication path is collected in real time, and the time interval is set. As the sampling period, Indicates at a point in time When , the number of hops between nodes on the communication path;

[0069] Real-time collection of the number of hops between each pair of nodes on the communication path is usually achieved by using a monitoring mechanism based on network routing protocols. Specifically, dynamic routing protocols such as OSPF (Open Shortest Path First) or BGP (Border Gateway Protocol) can be used. These protocols regularly exchange routing information between nodes, including path metadata such as hop counts. By deploying routing monitoring tools (such as SNMP monitoring, NetFlow or sFlow analysis) on the monitoring nodes, the number of hops between each pair of nodes on the path can be collected and recorded in real time. The monitoring system will continuously obtain and analyze the hop count data on each communication path based on the updated information of the routing table to ensure the real-time and accuracy of the path changes between nodes. This method can not only capture the current hop count, but also track the changes in the hop count over time, providing basic data for further analysis.

[0070] Calculate the hop count change rate between adjacent sampling time points , the calculation expression is: ,in, Indicates at a point in time The hop count change rate at that time, k is an adjustment factor used to control the sensitivity of the change rate, and i represents the current moment in the time series;

[0071] It should be noted that the exponential function part of the formula It is a smoothing function that increases the sensitivity of the rate of change to small changes while avoiding sudden changes caused by large changes.

[0072] In order to further analyze the cumulative effect of the hop count change, we calculate the The cumulative effect of , the calculation expression is: ,in, Indicates the time point The cumulative change effect at time point j The time point index within;

[0073] By changing the rate of each Multiplying by a modified term of a logarithmic function can highlight significant jump count changes while giving appropriate weight to small changes.

[0074] Cumulative change effect Calculate the jump index between each pair of nodes. The calculation expression is: , where represents the jump index between each pair of nodes, and n represents the number of sampling times in the monitoring window;

[0075] This formula combines the cumulative change effect and the nonlinear relationship of the jump number. The influence of hop count is considered to ensure that the hop index can reflect the dynamic characteristics of hop count changes in the entire monitoring window, especially highlighting the hop count fluctuation characteristics of rapidly changing paths.

[0076] Comprehensively analyze the communication path and calculate the jump index. The calculation expression is: , where represents the hopping index between the wth pair of nodes, m represents the total number of nodes on the communication path, Indicates the hopping index of the communication path.

[0077] In the monitoring window, the larger the jump index value generated by analyzing the hop count change rate between nodes on the communication path, the faster the node status of the communication path changes. In other words, a larger jump index indicates that the node status on this path has changed frequently and significantly during the monitoring period, such as the addition or exit of nodes, signal quality fluctuations, etc. On the contrary, a smaller jump index indicates that the node status on the path is relatively stable and changes less.

[0078] The rapid increase in the speed of node energy consumption on the communication path does indicate that the node state of the path is changing rapidly. Generally, the energy consumption of a node is closely related to factors such as the complexity of the task it processes, the frequency of communication, and the adjustment of signal strength. When the energy consumption rate is significantly accelerated, it often means that the node is processing more data transmission tasks or frequently participating in the routing adjustment of the communication path. In addition, the node may increase its energy consumption due to frequent movement or continuous adjustment of the transmission power to adapt to environmental changes. This drastic change in energy consumption usually reflects that the environment or workload of the node has changed rapidly, causing its state to become unstable. If not handled in time, it may lead to premature failure of the node, interruption of communication, and even cause reliability problems of the entire path. Therefore, the sudden increase in the energy consumption rate is an important indicator for judging the rapid change of the node state on the communication path.

[0079] In the monitoring window, the node energy consumption index is generated by analyzing the speed change of node energy consumption on the communication path. The specific steps are as follows:

[0080] For each node on the communication path, calculate its instantaneous energy consumption rate. Suppose each node is and The remaining energies are and , then the instantaneous energy consumption rate The calculation expression is: , where Indicates the node at a point in time The remaining energy when Indicates the node at a point in time The processing power at that moment reflects the workload of the node at that moment. Indicates the node at a point in time The working voltage at , reflects the energy consumption characteristics of the node;

[0081] The remaining energy of a node at time point t can be directly read through the node's built-in power monitoring module or sensor, which usually monitors the battery voltage and current in real time and calculates the current remaining energy. The processing power of a node at time point t can be determined by monitoring the load level of the node's CPU or other processor and the current operating mode, which are usually provided by the node's operating system or embedded control system. The operating voltage of the node at time point t can be obtained in real time through a voltage sensor embedded in the node. The acquisition of these parameters can be achieved through built-in hardware monitoring tools, operating system interfaces, and statistical functions provided by the network protocol stack.

[0082] Calculate the cumulative energy consumption coefficient of the node , the cumulative energy consumption coefficient reflects the cumulative energy consumption of the node within the monitoring window, and the calculation expression is: , where represents the cumulative energy consumption coefficient of the node within the monitoring window, Indicates at a point in time The instantaneous energy consumption rate at Indicates at a point in time The transmission time is the duration of the node communication activity. Indicates the node at a point in time The current working frequency at , which indicates the complexity of the node processing task;

[0083] Calculate the sensitivity of energy consumption changes , the energy consumption change sensitivity reflects the sensitivity of the energy consumption rate to time changes, and the calculation expression is: , where represents the time rate of change of the cumulative energy consumption coefficient, and They represent the time change rates of node processing power and operating voltage respectively;

[0084] The node energy consumption coefficient of each node is calculated by comprehensively considering the energy consumption change sensitivity, cumulative energy consumption coefficient and instantaneous energy consumption rate. The calculation expression is: , where represents the node energy consumption coefficient of each node, and T represents the total duration of the monitoring window;

[0085] The ratio term calculates the cumulative energy consumption coefficient The ratio of the historical energy consumption of a node to the total energy consumption in the time period reflects the relationship between the historical energy consumption of the node and its total energy consumption. Through the total energy consumption The final result of the node energy consumption index is adjusted by performing nonlinear transformation. When the total energy consumption is low, the node energy consumption coefficient is high, which enhances the influence of other parameters; when the total energy consumption is high, the index value decreases, which weakens the influence of other parameters. This makes the energy consumption index more sensitive to changes in energy consumption.

[0086] Calculate the node energy consumption index, the calculation expression is: , where represents the node energy consumption coefficient of the fth node, m represents the total number of nodes on the communication path, Indicates the node energy consumption index of the communication path.

[0087] It can be seen from the node energy consumption index that, under the monitoring window, the larger the performance value of the node energy consumption index generated after analyzing the speed change of node energy consumption on the communication path, the more drastic the change of node energy consumption on the communication path, which reflects that the node state of the communication path changes faster. This means that the node has experienced significant energy consumption fluctuations in a short period of time, which may be caused by frequent processing tasks, communication activities or voltage fluctuations. On the contrary, if the performance value of the node energy consumption index is small, it means that the energy consumption of the node is relatively stable, the node state on the communication path changes slowly, and the path is more stable. Through this index, the dynamics of the communication path can be effectively evaluated and monitored.

[0088] The machine learning model is not specifically limited here, and can achieve the jump index and node energy consumption index Perform comprehensive analysis to generate dynamic change coefficients of path nodes In order to realize the technical solution of the present invention, the present invention provides a specific implementation method;

[0089] Path node dynamic change coefficient The resulting calculation formula is: , where , Transition Index and node energy consumption index The preset scaling factor of , Both are greater than 0.

[0090] It can be seen from the dynamic change coefficient of the path node that, under the monitoring window, the larger the value of the jump index generated after analyzing the change rate of the number of hops between the nodes on the communication path, the larger the performance value of the node energy consumption index generated after analyzing the speed change of the energy consumption of the nodes on the communication path, that is, the larger the performance value of the dynamic change coefficient of the path node generated under the monitoring window, the faster the node state of the communication path changes, and vice versa.

[0091] According to the results of intelligent evaluation, all communication paths are divided into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths;

[0092] The dynamic change coefficient of the path node generated after comprehensive analysis of the nodes on the communication path under the monitoring window is compared and analyzed with the preset reference range of the dynamic change coefficient of the path node, and the communication path is divided. The specific steps of division are as follows:

[0093] If the path node dynamic change coefficient is within the path node dynamic change coefficient reference range, the communication path is divided into a stable change path;

[0094] If the path node dynamic change coefficient is less than the minimum value of the path node dynamic change coefficient reference range, the communication path is classified as a low dynamic path;

[0095] If the dynamic change coefficient of the path node is greater than the maximum value of the reference range of the dynamic change coefficient of the path node, the communication path is classified as a high dynamic path;

[0096] A high-dynamic path refers to a path where the node status changes frequently, a normal-change path refers to a path where the node status is relatively stable but still has certain fluctuations, and a low-dynamic path refers to a path where the node status changes very slowly.

[0097] For the communication paths divided into stable changing paths, a standard node status update frequency is set, and the status of the communication nodes is updated at the standard node status update frequency;

[0098] The standard node status update frequency is based on the traditional timed update method, which can effectively maintain the communication quality of the normally changing path. This step ensures that the path maintenance under normal circumstances is neither excessive nor ineffective, and maintains the rationality of resource utilization.

[0099] For high-dynamic paths and low-dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency based on the standard node status update frequency based on real-time dynamic change information;

[0100] For high-dynamic paths and low-dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency as follows:

[0101] Define fuzzy sets and membership functions for jump index, node energy consumption index, and standard node status update frequency;

[0102] The fuzzy set of jump index and node energy consumption index can be defined as "Low", "Medium", and "High".

[0103] The fuzzy set of standard node status update frequencies can be defined as "Low Frequency", "Medium Frequency", and "High Frequency".

[0104] The membership function can be a common triangle, trapezoid or Gaussian function, etc., which can be selected according to the actual application and data distribution. The membership function defines how the input parameters (such as the jump index and the node energy consumption index) are mapped to the fuzzy set.

[0105] Based on fuzzy sets, a fuzzy rule base is constructed. The rule form of the rule base is: "If" the jump index and the node energy consumption index belong to a certain fuzzy set, the actual node status update frequency will be adjusted to a certain level.

[0106] For example:

[0107] If the transition index is "high" and the node energy consumption index is "high", the actual node status update frequency is "high frequency".

[0108] If the hopping index is "medium" and the node energy consumption index is "medium", the actual node status update frequency is "medium frequency".

[0109] If the hopping index is "low" and the node energy consumption index is "low", the actual node status update frequency is "low frequency".

[0110] The output of these rules will affect the direction of adjusting the node status update frequency.

[0111] The real-time jump index and node energy consumption index are input into the fuzzy reasoning engine, and combined with the defined fuzzy rule base, fuzzy reasoning is performed (the result of the reasoning is a fuzzy set, which represents the possible actual node status update frequency). The fuzzy reasoning process is executed through the "if-then" rule to determine the update frequency that best suits the current path conditions. The reasoning formula is: , represents the fuzzy reasoning process, and They represent the membership functions of the jump index and node energy consumption index respectively, Indicates fuzzy output;

[0112] The fuzzy output obtained by fuzzy inference is analyzed using the centroid method Defuzzification is performed to generate the fuzzy actual node state update frequency. The defuzzification formula is: , is the membership function used in the defuzzification process, is the update frequency of the fuzzy actual node state obtained after defuzzification;

[0113] Update frequency according to the fuzzy actual node state after defuzzification , adjust the actual node status update frequency of the final application, and the adjustment formula is as follows: ,in, is the node status update frequency of the final application, Indicates the standard node status update frequency.

[0114] For highly dynamic paths, will be significantly improved to cope with frequent state changes. For low dynamic paths, Will be reduced to save resources.

[0115] The present invention collects and analyzes network topology information in real time, accurately extracts the dynamic change characteristics of nodes on the communication path, and based on these characteristics, evaluates the dynamic changes of nodes on the communication path through a machine learning model, and can effectively identify and distinguish high dynamic, stable change and low dynamic paths, thereby taking targeted routing update strategies for different types of paths. In a high dynamic path, by improving the node status update frequency, the changes in the network topology can be reflected in a timely manner, avoiding network splitting and data packet loss caused by routing information lag. In a low dynamic path, reducing the update frequency effectively reduces the consumption of network bandwidth and computing resources, and avoids the performance waste caused by frequent updates. Through this intelligent adjustment method, not only the real-time and reliability of communication are enhanced, but also the efficient operation of the network is maintained while saving resources. It is particularly suitable for mission-critical communication scenarios and plays a significant role in improving the overall performance of the network.

[0116] The present invention provides Figure 2 A low-latency self-organizing network communication system shown includes a topology information collection and analysis module, a dynamic change feature extraction and evaluation module, a communication path classification module, a standard frequency update module and a fuzzy logic control module;

[0117] The topology information collection and analysis module collects the current topology information from each node in the network, analyzes the collected topology information, and draws a topology diagram of the current network;

[0118] The dynamic change feature extraction and evaluation module extracts dynamic change features of nodes on each communication path, and uses a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features;

[0119] The communication path classification module divides all communication paths into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths, based on the results of intelligent evaluation;

[0120] A standard frequency update module, for a communication path divided into a stable change path, sets a standard node status update frequency, and updates the status of the communication node at the standard node status update frequency;

[0121] Fuzzy logic control module, for high dynamic paths and low dynamic paths, adopts fuzzy logic control, based on real-time dynamic change information, intelligently increases or decreases the actual node status update frequency on the basis of the standard node status update frequency;

[0122] A low-latency self-organizing network communication method provided in an embodiment of the present invention is implemented by the above-mentioned low-latency self-organizing network communication system. The specific method and process of a low-latency self-organizing network communication system are detailed in the embodiment of the above-mentioned low-latency self-organizing network communication method, which will not be repeated here.

[0123] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0124] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0125] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A low-latency self-organizing network communication method, It is characterized in that The following steps are involved: Collect the current topology information from each node in the network, analyze the collected topology information, and draw a topology diagram of the current network; Extract dynamic change features of nodes on each communication path, and use a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features; According to the results of intelligent evaluation, all communication paths are divided into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths; For the communication paths divided into stable changing paths, a standard node status update frequency is set, and the status of the communication nodes is updated at the standard node status update frequency; For high-dynamic paths and low-dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency based on the standard node status update frequency based on real-time dynamic change information; Extract dynamic change features of nodes on each communication path. The extracted features include the hop count change rate between nodes on the communication path and the speed change of energy consumption of nodes on the communication path. After acquisition, analyze the hop count change rate between nodes on the communication path and the speed change of energy consumption of nodes on the communication path under the monitoring window, and generate a hop index and a node energy consumption index respectively. Input the hop index and the node energy consumption index into a pre-trained machine learning model, generate a path node dynamic change coefficient through the machine learning model, and comprehensively evaluate the node change status on each communication path through the path node dynamic change coefficient. The hop count change rate between nodes on the communication path is analyzed to generate a hop index. The specific steps are as follows: Under the monitoring window, the number of hops between each pair of nodes on the communication path is collected in real time, and the time interval Δt is set as the sampling period. i ) represents the time point t i When , the number of hops between nodes on the communication path; Calculate the hop count change rate R(t i ), the expression for calculation is: Among them, R(t i ) represents the time point t i The hop count change rate at that time, k is an adjustment factor used to control the sensitivity of the change rate, and i represents the current moment in the time series; In order to further analyze the cumulative effect of the hop count change, we calculate the i The cumulative change effect C(t i ), the expression for calculation is: Among them, C(t i ) represents the time point t i The cumulative change effect at time t i The time point index within; Through the cumulative change effect C(t i ) Calculate the jump index between each pair of nodes, and the calculation expression is: In the formula, Transition inx ' represents the jump index between each pair of nodes, and n represents the number of sampling times in the monitoring window; Comprehensively analyze the communication path and calculate the jump index. The calculation expression is: In the formula, represents the transition index between the wth pair of nodes, m represents the total number of nodes on the communication path, and Transition inx Indicates the jump index of the communication path; In the monitoring window, the node energy consumption index is generated by analyzing the speed change of node energy consumption on the communication path. The specific steps are as follows: For each node on the communication path, calculate its instantaneous energy consumption rate. Suppose each node is i and t i+1 The residual energy at the time is E(t i ) and E(t i+1 ), then the instantaneous energy consumption rate R yy (t i ) is calculated as: In the formula, E(t i ) indicates that the node is at time point t i The remaining energy at i ) indicates that the node is at time point t i The processing power at the time, V(t i ) indicates that the node is at time point t i Working voltage at Calculate the cumulative energy consumption coefficient C(t) of the node. The cumulative energy consumption coefficient reflects the cumulative energy consumption of the node within the monitoring window. The calculation expression is: Where C(t) represents the cumulative energy consumption coefficient of the node in the monitoring window, R yy (t i ) represents the time point t i The instantaneous energy consumption rate at i ) represents the time point t i The transmission time, that is, the duration of the node communication activity, Q(t i ) indicates that the node is at time point t i The current working frequency at , which indicates the complexity of the node processing task; Calculate the energy consumption change sensitivity S(t). The energy consumption change sensitivity reflects the sensitivity of the energy consumption rate to time changes. The calculation expression is: In the formula, represents the time rate of change of the cumulative energy consumption coefficient, and They represent the time change rates of node processing power and operating voltage respectively; The node energy consumption coefficient of each node is calculated by comprehensively considering the energy consumption change sensitivity, cumulative energy consumption coefficient and instantaneous energy consumption rate. The calculation expression is: In the formula, Node_energy represents the node energy consumption coefficient of each node, and T represents the total duration of the monitoring window; Calculate the node energy consumption index, the calculation expression is: Where, Node_energy f Represents the node energy consumption coefficient of the fth node, Energy inx represents the node energy consumption index of the communication path; The dynamic change coefficient of the path node generated after comprehensive analysis of the nodes on the communication path under the monitoring window is compared and analyzed with the preset reference range of the dynamic change coefficient of the path node, and the communication path is divided. The specific steps of division are as follows: If the path node dynamic change coefficient is within the path node dynamic change coefficient reference range, the communication path is divided into a stable change path; If the path node dynamic change coefficient is less than the minimum value of the path node dynamic change coefficient reference range, the communication path is classified as a low dynamic path; If the dynamic change coefficient of the path node is greater than the maximum value of the reference range of the dynamic change coefficient of the path node, the communication path is classified as a high dynamic path; For high-dynamic paths and low-dynamic paths, fuzzy logic control is used to intelligently increase or decrease the actual node status update frequency as follows: Define fuzzy sets and membership functions for jump index, node energy consumption index, and standard node status update frequency; Based on fuzzy sets, a fuzzy rule base is constructed. The rule form of the rule base is: "If" the jump index and the node energy consumption index belong to a certain fuzzy set, the actual node status update frequency will be adjusted to a certain level; The real-time jump index and node energy consumption index are input into the fuzzy reasoning engine, and fuzzy reasoning is performed in combination with the defined fuzzy rule base. The fuzzy reasoning process is executed through the "if-then" rule to determine the update frequency that best suits the current path conditions. The reasoning formula is: In the formula, Fuzzy Inference represents the fuzzy inference process, μ(Transition inx (t)) and μ(Energy inx (t)) represent the membership functions of the jump index and node energy consumption index, respectively, and F fuzzy Indicates fuzzy output; The fuzzy output F obtained by fuzzy inference is analyzed using the centroid method. fuzzy Defuzzification is performed to generate the fuzzy actual node state update frequency. The defuzzification formula is: μ(F) is the membership function used in the defuzzification process, F actual is the update frequency of the fuzzy actual node state obtained after defuzzification; Update the frequency F according to the defuzzified fuzzy actual node state actual , adjust the actual node status update frequency of the final application, and the adjustment formula is as follows: F new =F std ·F actual , where F new is the node status update frequency of the final application, F std Indicates the standard node status update frequency.

2. A low-latency self-organizing network communication method according to claim 1, It is characterized in that Draw the topology diagram of the current network. The specific steps are as follows: Start the state collection mechanism on each node in the network; After the status collection mechanism is started, all nodes send their collected information to one or more central nodes for aggregation; After data aggregation and storage are completed, the collected node status information begins to be analyzed; Based on the analysis results, use network visualization tools or dedicated software to draw the current network topology diagram.

3. A low-latency self-organizing network communication system, used to implement a low-latency self-organizing network communication method as described in any one of claims 1-2 above, It is characterized in that It includes topology information collection and analysis module, dynamic change feature extraction and evaluation module, communication path classification module, standard frequency update module and fuzzy logic control module; The topology information collection and analysis module collects the current topology information from each node in the network, analyzes the collected topology information, and draws a topology diagram of the current network; The dynamic change feature extraction and evaluation module extracts dynamic change features of nodes on each communication path, and uses a pre-trained machine learning model to comprehensively evaluate the node change status on each communication path based on the extracted dynamic change features; The communication path classification module divides all communication paths into three categories, including high-dynamic paths, stable-changing paths, and low-dynamic paths, based on the results of intelligent evaluation; A standard frequency update module, for a communication path divided into a stable change path, sets a standard node status update frequency, and updates the status of the communication node at the standard node status update frequency; The fuzzy logic control module adopts fuzzy logic control for high dynamic paths and low dynamic paths. Based on the real-time dynamic change information, it intelligently increases or decreases the actual node status update frequency on the basis of the standard node status update frequency.