A reinforcement learning-based method for optimizing optical transmission service paths in power communication

Abstract

This invention discloses a reinforcement learning-based method for optimizing optical transmission service paths in power communication, belonging to the field of power communication optical transmission service path optimization technology. Specifically, it includes the following steps: real-time acquisition of link quality information generated by each target link during periodic quality fluctuations; generation of link fluctuation amplitude index and link fluctuation regularity index for each target link; construction of a link quality evaluation model based on the generated link fluctuation amplitude index and link fluctuation regularity index for each target link; and generation of quality fluctuation coefficients for each target link; and evaluation and analysis based on the generated quality fluctuation coefficients for each target link to assess the degree of periodic fluctuation in link quality during periodic quality fluctuations. This invention solves the problem of untimely adaptation to periodic link quality fluctuations in power communication optical transmission networks, ensuring stable transmission and efficient execution of critical tasks through real-time monitoring and dynamic path optimization.

Description

Technical Field

[0001] This invention relates to the field of power communication optical transmission service path optimization technology, and specifically to a power communication optical transmission service path optimization method based on reinforcement learning. Background Technology

[0002] Optical transmission service paths in power communication refer to the specific paths through which data transmission is achieved via optical fiber communication networks within a power system. These paths connect various devices and control centers within the power system, ensuring the smooth operation of various communication tasks such as power dispatching, monitoring, and fault diagnosis. In power communication systems, optical transmission is commonly used to support large-scale data transmission and efficient remote control due to its high bandwidth, low latency, and interference resistance. However, with the continuous expansion of power system scale and the increasing communication demands, the load and complexity of communication networks are also increasing. Therefore, how to select and optimize optical transmission service paths to ensure efficient, stable, and low-latency data transmission has become a crucial issue in the field of power communication.

[0003] Optimizing optical transmission service paths in power communication using reinforcement learning is beneficial because traditional path optimization methods often rely on static rules and pre-defined algorithms, which struggle to cope with the dynamic environment and real-time network load changes in power communication networks. Reinforcement learning, on the other hand, simulates real-time changes in the environment and learns and adjusts its decision-making strategies through interaction with the environment to find the optimal path. Specifically, reinforcement learning can continuously optimize path selection based on changes in network conditions (such as traffic fluctuations, link failures, and bandwidth changes), improving data transmission efficiency and reliability. For example, reinforcement learning models can dynamically adjust communication paths to minimize latency, avoid network congestion, and improve bandwidth utilization, thereby better meeting the real-time and reliability requirements of power systems. Therefore, reinforcement learning-based path optimization methods can more intelligently and efficiently address the dynamic changes in complex power communication networks, providing more stable and faster communication services.

[0004] Existing reinforcement learning-based optical transmission service path optimization techniques for power communication networks address dynamic changes and complex environments by introducing intelligent decision-making mechanisms. In this approach, reinforcement learning algorithms typically take network topology, bandwidth requirements, link status, latency, and load as environmental inputs, and use a reward mechanism to evaluate the quality of path selection. In each path selection, the algorithm makes a decision based on the current network status (e.g., link load, fiber transmission quality), choosing a communication path and receiving rewards or penalties based on actual performance during transmission (e.g., reduced latency, increased bandwidth), progressively optimizing the decision-making strategy. Specifically, through multiple interactions and explorations, the algorithm optimizes the path selection strategy, enabling the system to adaptively select the optimal optical transmission path, avoiding network bottlenecks, reducing transmission latency, and improving resource utilization, thereby enhancing the overall operational efficiency and stability of the power communication network. This optimization process not only considers static factors but also responds in real-time to dynamic network load changes, maintaining high-quality communication even in complex communication environments.

[0005] The existing technology has the following shortcomings:

[0006] In power communication optical transmission networks, the quality of certain links may fluctuate periodically when affected by external factors (such as weather changes, aging power equipment, load fluctuations, etc.). For example, signal strength and bit error rate may show regular changes over different time periods. This fluctuation is typically periodic, with certain time intervals and patterns of change. However, existing reinforcement learning-based power communication optical transmission service path optimization techniques mainly rely on historical data for path selection, often ignoring the periodic changes in link quality. Because the decision-making process of reinforcement learning algorithms is usually based on long-term learning and environmental feedback, it fails to perceive the periodic characteristics of link quality fluctuations in a timely manner, resulting in path selection failing to accurately adapt to these periodic fluctuations. When link quality fluctuates significantly over certain time periods, existing algorithms may fail to update path selection strategies in a timely manner, leading to frequent changes in path selection or misjudging the actual link condition, thus affecting network stability and reliability. Due to the instability of path selection, data transmission for critical tasks with extremely high latency requirements in power communication networks (such as power dispatch instructions, real-time fault diagnosis, etc.) may experience delays, packet loss, or interruptions, thereby affecting the system's real-time response capability and control accuracy, ultimately leading to a decrease in the reliability of the power system and potentially even triggering system security risks.

[0007] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0008] The purpose of this invention is to provide a reinforcement learning-based method for optimizing optical transmission service paths in power communication, in order to solve the problems mentioned in the background art.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a method for optimizing optical transmission service paths in power communication based on reinforcement learning, specifically including the following steps:

[0010] In power communication optical transmission networks, each link used for transmitting communication tasks is identified and monitored in real time. Based on the fluctuation pattern of link quality, all links exhibiting periodic quality fluctuations are selected and designated as target links.

[0011] The link quality information generated by each target link during periodic quality fluctuations is acquired in real time, and analyzed after acquisition to generate the link fluctuation amplitude index and link fluctuation pattern index for each target link.

[0012] A link quality assessment model is constructed based on the link fluctuation amplitude index and link fluctuation regularity index of each generated target link, and the quality fluctuation coefficient of each target link is generated.

[0013] An evaluation analysis is conducted based on the quality fluctuation coefficients of each target link to assess the degree of periodic fluctuation in link quality during periodic quality fluctuations. Based on the evaluation results, each target link is classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links.

[0014] Based on the classification of each target link, different path optimization measures are implemented for each type of link;

[0015] Based on the implemented path optimization measures, the quality fluctuations of each link are continuously monitored, and the path optimization effect is tracked in real time. The path selection strategy is automatically adjusted to ensure the continuous, stable and efficient execution of power communication tasks.

[0016] Preferably, the link quality information generated by each target link during periodic quality fluctuations is acquired in real time, and analyzed after acquisition to generate a link fluctuation amplitude index and a link fluctuation pattern index for each target link. Specifically, this includes the following steps:

[0017] The link quality information generated by each target link during periodic quality fluctuations is acquired in real time and preprocessed after acquisition.

[0018] Extract link quality fluctuation information and link fluctuation stability information from the link quality information generated by each target link during periodic quality fluctuations after preprocessing;

[0019] The extracted link quality fluctuation information and link fluctuation stability information are analyzed to generate the link fluctuation amplitude index and link fluctuation regularity index for each target link.

[0020] Preferably, the logic for obtaining the link fluctuation amplitude index of each target link is as follows:

[0021] Link quality fluctuation information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the bit error rate, signal strength, and successfully transmitted data volume at different times within a certain period of time for each target link during periodic quality fluctuations, and these are categorized as follows: , and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Bit error rate at time step Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal strength at any given time Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time. , , and All are positive integers;

[0022] The link fluctuation amplitude index for each target link is calculated using the following formula:

[0023]

[0024] In the formula, For the first Link fluctuation amplitude index of the target link.

[0025] Preferably, the logic for obtaining the link fluctuation pattern index of each target link is as follows:

[0026] Link fluctuation stability information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the signal-to-noise ratio and data traffic carried by each target link at different times within a certain period during the periodic quality fluctuations, and these are calibrated as follows: and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal-to-noise ratio at any given time. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data traffic carried at any given moment , , and All are positive integers;

[0027] Calculate the mean signal-to-noise ratio (SNR) of each target link at different times during a period of periodic quality fluctuations. ,in accordance with: ;

[0028] The link fluctuation pattern index for each target link is calculated using the following formula:

[0029]

[0030] In the formula, For the first Link fluctuation pattern index of the target link. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time.

[0031] Preferably, the link fluctuation amplitude index for each generated target link Link fluctuation pattern index A link quality assessment model is constructed, and the quality fluctuation coefficient of each target link is generated by weighted summation. The specific calculation formula is as follows:

[0032]

[0033] In the formula, For the first The quality fluctuation coefficient of the target link, and The link fluctuation amplitude index for each target link Link fluctuation pattern index The non-zero weight coefficients, and .

[0034] Preferably, a pre-defined threshold range for the quality fluctuation coefficient is determined. And after determination, it is compared with the quality fluctuation coefficients of each generated target link. A comparison was conducted, and the degree of periodic fluctuation in link quality during the periodic quality fluctuation period was evaluated based on the comparison results. Based on the evaluation results, each target link was classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links. The specific comparison analysis and classification are as follows:

[0035] like The target link exhibits low volatility in link quality during the periodic quality fluctuation period, thus classifying it as a low-volatility link.

[0036] like The target link exhibits moderate fluctuations in link quality during the periodic quality fluctuation period, thus classifying it as a moderate fluctuation link.

[0037] like The target link exhibits high volatility in its link quality during periods of periodic quality fluctuations, thus classifying it as a high-volatility link.

[0038] Preferably, based on the division of each target link, different path optimization measures are implemented for each type of link, specifically:

[0039] For highly volatile links, the specific path optimization measures adopted are as follows: immediately adjust the path selection strategy, including: prioritizing links with high signal strength and low load to avoid frequent path switching and reduce the impact of link fluctuations; enabling redundant links to distribute the load and ensure that highly volatile links do not cause overload and communication interruption; dynamically adjusting the network load balancing strategy to optimize traffic distribution between links and reduce the load pressure on highly volatile links; sending high-priority scheduling warnings to the network control center and activating automated control mechanisms to ensure that the real-time performance and reliability of communication tasks are not affected.

[0040] For links with moderate fluctuations, the specific path optimization measures are as follows: optimize the path selection strategy and dynamically adjust the path according to changes in network traffic; adjust the link scheduling priority according to changes in the link fluctuation index and throughput to ensure that the performance of links with moderate fluctuations does not degrade under high load; and regularly check the network load balancing strategy and make corresponding adjustments to ensure that the link load is not concentrated on links with moderate fluctuations.

[0041] For low-fluctuation links, the specific path optimization measures are as follows: maintain the current path selection strategy unchanged, prioritize the use of low-fluctuation links to transmit latency-sensitive tasks, and ensure their transmission stability; regularly monitor low-fluctuation links and optimize and adjust them according to the actual needs of network traffic; and ensure that low-fluctuation links remain efficient and stable in long-term operation by continuously monitoring link quality fluctuations.

[0042] The technical effects and advantages provided by the present invention in the above technical solution are as follows:

[0043] 1. This invention effectively identifies the regularity and intensity of link quality fluctuations by conducting real-time quality monitoring and periodic fluctuation analysis of links in power communication optical transmission networks. By generating link fluctuation amplitude and link fluctuation regularity indices, the degree of fluctuation of each link during periodic quality fluctuations can be accurately assessed, thereby achieving real-time classification of link status. This process overcomes the problem that existing reinforcement learning-based path optimization methods typically rely on historical data for path selection, and can dynamically adapt to the periodic changes in link quality. In this way, not only can the link status be comprehensively monitored, but more accurate data support can also be provided for path selection, thus avoiding the shortcomings of traditional path selection methods that cannot respond to link quality fluctuations in a timely manner.

[0044] 2. This invention accurately classifies different types of links (high-fluctuation links, medium-fluctuation links, and low-fluctuation links) based on the evaluation results of link quality fluctuations, and implements customized path optimization measures for each type of link. For high-fluctuation links, the system prioritizes links with high signal strength and low load to avoid frequent path switching, and uses redundant links to share the load, ensuring network stability even during peak load periods. For medium-fluctuation links, the system dynamically adjusts path priorities and optimizes load distribution to reduce the burden on medium-fluctuation links, ensuring no performance degradation under high load. For low-fluctuation links, the system maintains path stability and reduces unnecessary path adjustments, ensuring stable transmission of latency-sensitive tasks. This optimization strategy based on link fluctuation characteristics allows path selection to dynamically adapt to the fluctuation characteristics of different links, ensuring efficient and stable execution of communication tasks.

[0045] 3. This invention also emphasizes the real-time monitoring and automatic adjustment functions of path optimization. By continuously monitoring link quality fluctuations, the system can track the effect of path optimization in real time and automatically adjust the path selection strategy based on changes in link quality. This means that even if abnormal fluctuations occur in the link during certain periods, the system can respond quickly and adjust the path selection in a timely manner, avoiding delays or interruptions in communication tasks caused by link fluctuations. Through this dynamic adjustment mechanism, the system can ensure that power communication tasks, especially those with extremely high latency requirements (such as power dispatch instructions and real-time fault diagnosis), can be executed continuously and stably in a network environment with frequent fluctuations. This improves the reliability, real-time response capability, and task execution efficiency of the entire power communication network, avoiding communication interruptions and performance degradation caused by unstable path selection. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0047] Figure 1 This is a flowchart illustrating a reinforcement learning-based method for optimizing optical transmission service paths in power communication according to the present invention. Detailed Implementation

[0048] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0049] This invention provides, for example Figure 1 The proposed method for optimizing optical transmission service paths in power communication based on reinforcement learning includes the following steps:

[0050] In power communication optical transmission networks, each link used for transmitting communication tasks is identified and monitored in real time. Based on the fluctuation pattern of link quality, all links exhibiting periodic quality fluctuations are selected and designated as target links.

[0051] In power communication optical transmission networks, identifying and monitoring the links used for communication tasks in real time first requires collecting link status data in real time through a network monitoring system. The status information of each link, including parameters such as signal strength, bit error rate, bandwidth utilization, link latency, and load, is continuously monitored and recorded. To ensure the accuracy and real-time nature of monitoring, the system typically uses protocols (such as SNMP or NETCONF) to exchange data with various devices, thereby obtaining the real-time operating status of the links. This data is transmitted to a central monitoring platform, where it is combined with network topology information to identify links primarily used for transmitting critical tasks such as power dispatching and control commands, ensuring these links receive priority and real-time monitoring. The real-time monitoring process utilizes a big data platform to analyze the link status, thereby achieving comprehensive monitoring of all links.

[0052] By identifying links exhibiting periodic quality fluctuations based on their volatility patterns, signal processing and data analysis methods can be used to conduct in-depth analysis of the real-time collected link quality data. First, time series analysis can detect periodic fluctuation patterns in link quality (such as bit error rate and signal strength). Applying techniques like Fourier transform or wavelet transform, the frequency and amplitude of periodic fluctuations in the link data can be identified, determining which links exhibit significant periodic fluctuations. Based on set thresholds, the system filters out links whose fluctuation frequency and amplitude exceed standards, designating them as target links. Machine learning techniques are then used to further analyze these links, accurately identifying their periodic fluctuation characteristics. These target links are designated as "links to be analyzed," providing data support for subsequent path optimization decisions and ensuring the system can manage link fluctuations in a targeted manner.

[0053] This process aims to overcome the shortcomings of existing reinforcement learning-based path optimization algorithms in dealing with periodic quality fluctuations. Existing algorithms typically select paths based on historical data, but they cannot identify and respond to periodic fluctuations in link quality in real time. This results in path selection failing to adapt to link fluctuations promptly, leading to path instability. This instability is particularly significant for latency-critical tasks in power communication networks (such as power dispatch instructions and real-time fault diagnosis), potentially causing delays, packet loss, or communication interruptions, thus affecting the real-time response capability and control accuracy of the power system. By monitoring link quality fluctuations in real time and accurately calibrating target links, the system can effectively identify the fluctuation characteristics of these links and implement personalized path optimization strategies, thereby ensuring stable transmission of network communication tasks and improving the reliability and security of the power system.

[0054] The link quality information generated by each target link during periodic quality fluctuations is acquired in real time, and analyzed after acquisition to generate the link fluctuation amplitude index and link fluctuation pattern index for each target link.

[0055] In this embodiment, the link quality information generated by each target link during periodic quality fluctuations is acquired in real time, and then analyzed to generate the link fluctuation amplitude index and link fluctuation pattern index for each target link. Specifically, the steps include:

[0056] The link quality information generated by each target link during periodic quality fluctuations is acquired in real time and preprocessed after acquisition.

[0057] Real-time acquisition of link quality information generated during periodic quality fluctuations in each target link can be achieved by deploying a network monitoring system. This system establishes communication with the monitoring interfaces of the link devices (such as SNMP, NETCONF, gRPC, etc.) and periodically collects key link quality data. Specifically, the system can acquire link information such as bit error rate (BER), signal strength (RSSI), throughput, and link load in real time. All data is transmitted to the data acquisition platform in real time via network transmission protocols. The data acquisition platform can receive data streams from different link devices through data stream interfaces (such as REST API, WebSocket, etc.) and mark and store the data according to timestamps and link identifiers for subsequent processing and analysis. In addition, the system can set the monitoring frequency and data acquisition window to dynamically acquire link quality information during periodic fluctuations, ensuring real-time performance and data accuracy.

[0058] The purpose of preprocessing is to ensure that the acquired link quality information is of high quality and consistency, and suitable for subsequent analysis. Preprocessing includes steps such as data cleaning, denoising, and standardization to eliminate invalid or misleading data and ensure that the data reflects the true link status. Data cleaning aims to remove null values, duplicate data, or data with incorrect formatting, ensuring data integrity. Denoising can remove random fluctuations in the data using smoothing algorithms (such as moving averages and exponential smoothing), especially when there are transient anomalies in signal strength or bit error rate, preventing these anomalous data from affecting subsequent analysis results. Standardization transforms the data into a unified standard scale so that quality information from different links can be compared uniformly; Z-score standardization or Min-Max standardization is commonly used. These preprocessing steps can be automated using data processing frameworks (such as Apache Spark and Pandas) to ensure that the data has been normalized and irrelevant noise removed before entering the subsequent analysis stage, improving the accuracy and reliability of the analysis results.

[0059] Extract link quality fluctuation information and link fluctuation stability information from the link quality information generated by each target link during periodic quality fluctuations after preprocessing;

[0060] Extracting link quality fluctuation information and link fluctuation stability information can be achieved through data analysis and signal processing techniques. First, for the preprocessed data of each target link, the system can apply statistical analysis methods to extract link quality fluctuation information. Specifically, fluctuation rate calculations (such as standard deviation and root mean square error) can be used to measure the fluctuation amplitude of the link's bit error rate, signal strength, and throughput, reflecting the link's quality fluctuations. Then, link fluctuation stability information can be extracted using methods such as spectrum analysis and autocorrelation analysis. Spectrum analysis (e.g., using Fourier transform or wavelet transform) can help identify the periodic components of link quality fluctuations and extract the frequency characteristics of the fluctuations, while autocorrelation analysis can reveal the regularity and stability of the fluctuations. Through these methods, fluctuation information representing the amplitude of fluctuations and stability information representing regularity can be extracted from the link's bit error rate, signal strength, and throughput data, respectively. All these operations can be automated using data analysis platforms (such as Apache Spark, Pandas in Python, or NumPy), ensuring efficient data extraction and processing.

[0061] The extracted link quality fluctuation information and link fluctuation stability information are analyzed to generate the link fluctuation amplitude index and link fluctuation regularity index for each target link.

[0062] In this embodiment, the logic for obtaining the link fluctuation amplitude index of each target link is as follows:

[0063] Link quality fluctuation information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the bit error rate, signal strength, and successfully transmitted data volume at different times within a certain period of time for each target link during periodic quality fluctuations, and these are categorized as follows: , and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Bit error rate at time step Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal strength at any given time Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time. , , and All are positive integers;

[0064] Real-time acquisition of data such as bit error rate, signal strength, and successfully transmitted data at different times during a period of periodic quality fluctuations for each target link can be achieved by deploying a dedicated network monitoring system in conjunction with existing network management protocols and real-time data streams. Specifically, this involves establishing a connection with the link devices using network management protocols (such as SNMP, NETCONF, gRPC, etc.) and periodically extracting real-time operational status data from the link devices. Firstly, the bit error rate can be obtained in real-time through the diagnostic functions of the link devices. Bit error rate (BER) is a crucial indicator for measuring link transmission quality, reflecting the percentage of erroneous data transmitted per unit time. A high BER typically indicates poor link quality or interference. Secondly, signal strength can also be obtained through the device's monitoring interface. Signal strength represents the signal strength received by the receiver; lower signal strength usually indicates link instability, potentially leading to greater fluctuations or signal loss. Finally, the amount of successfully transmitted data can be obtained by monitoring link traffic. Throughput data reflects the amount of data successfully transmitted per unit time and is a key indicator for measuring link load and transmission efficiency. During this data acquisition process, the system continuously receives real-time data streams from link devices through data flow interfaces (such as REST APIs and WebSockets), stores them in the monitoring platform, and tags the data with timestamps to ensure timeliness and accuracy. Through a data storage and analysis system, this real-time acquired data is stored in a database (such as MySQL, MongoDB, or InfluxDB time-series databases) and used in subsequent processing to provide a basis for link quality analysis. The system can periodically update data according to defined time windows (such as every minute or every second) to ensure real-time monitoring and dynamic capture of link changes.

[0065] The link fluctuation amplitude index for each target link is calculated using the following formula:

[0066]

[0067] In the formula, For the first Link fluctuation amplitude index of the target link.

[0068] Link fluctuation index The calculation formula quantifies the intensity and amplitude of link quality fluctuations through a comprehensive analysis of the bit error rate, signal strength, and throughput during periodic quality fluctuations. The formula contains... The term is used to reflect the impact of signal strength on link fluctuations. The lower the signal strength, the greater the fluctuation in link quality. The exponential operation can aggravate the impact of link fluctuations when the signal is weak, ensuring that the fluctuations of poor signals are fully considered. The term represents throughput, which is the amount of data successfully transmitted by the link at that moment. When the throughput is high, the link load is heavy, and the fluctuation intensity is more significant. Therefore, the throughput is directly multiplied into the formula, reflecting the contribution of link load fluctuations to the overall fluctuation amplitude. The algorithm addresses the bit error rate (BER). A higher BER indicates greater quality fluctuations in the link, and a stronger impact on the fluctuation amplitude. Logarithmic operations smooth out the impact of the BER, preventing excessively high BER values ​​from leading to overly large calculated values ​​and ensuring that fluctuation intensity is not excessively amplified by extreme values. By combining these factors and averaging them, a fluctuation amplitude index for each link is calculated. This index accurately quantifies the fluctuation intensity of the link during periodic quality fluctuations, providing a valid basis for subsequent path optimization and scheduling decisions.

[0069] No. Link fluctuation index of target link The magnitude of the link fluctuation amplitude index directly reflects the intensity and amplitude of the link's fluctuations during periodic quality fluctuations. Specifically, a higher link fluctuation amplitude index indicates that the link experiences larger fluctuations during periodic quality fluctuations, suggesting frequent and severe quality fluctuations, potentially due to external interference or network load fluctuations. Conversely, a lower link fluctuation amplitude index indicates relatively smaller quality fluctuations, more stable fluctuations, and stronger link stability. Therefore, there is a direct positive correlation between the magnitude of the link fluctuation amplitude index and "assessing the degree of periodic fluctuations in link quality for each target link during periodic quality fluctuations." This is achieved by calculating and analyzing the index for each target link. The system can quantify the intensity and periodicity of link fluctuations, thereby identifying links with larger fluctuations and prioritizing adjustments during path optimization to ensure the transmission stability of critical tasks.

[0070] In this embodiment, the logic for obtaining the link fluctuation pattern index of each target link is as follows:

[0071] Link fluctuation stability information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the signal-to-noise ratio and data traffic carried by each target link at different times within a certain period during the periodic quality fluctuations, and these are calibrated as follows: and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal-to-noise ratio at any given time. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data traffic carried at any given moment , , and All are positive integers;

[0072] Real-time acquisition of the signal-to-noise ratio (SNR) and data traffic carried by each target link at different times during a period of periodic quality fluctuations can be achieved by deploying a network monitoring system and a real-time data acquisition platform. First, the SNR is a key indicator of link signal quality, typically obtained through the link's receiver or the wireless module of network devices. The system can communicate with devices such as routers, switches, or base stations via protocols like SNMP (Simple Network Management Protocol), NETCONF, or gRPC to acquire the SNR of each target link in real time. Data reflects the signal quality of the link. In environments with poor signal, A lower value indicates that the link may face significant noise interference and instability. Secondly, data traffic represents the amount of data transmitted by the link per unit time, reflecting the link load. Network monitoring systems can monitor and record the throughput data of each link in real time using traffic monitoring tools (such as sFlow, NetFlow) or traffic analyzers (such as Wireshark, tcpdump). Specifically, data traffic can be calculated by analyzing the size and rate of data packets transmitted at network interfaces, as well as the success and loss of packets. Real-time acquired data traffic is transmitted to the monitoring platform through data flow interfaces (such as REST API, WebSocket) to ensure that changes in link load are reflected promptly. In practice, this data is stored in databases (such as InfluxDB, TimescaleDB) and timestamped to ensure the timeliness of link quality information. In this way, the monitoring system can continuously update the link data at different points in time. This allows for real-time monitoring of the link's quality status, providing essential information for subsequent analysis and optimization.

[0073] Calculate the mean signal-to-noise ratio (SNR) of each target link at different times during a period of periodic quality fluctuations. ,in accordance with: ;

[0074] The link fluctuation pattern index for each target link is calculated using the following formula:

[0075]

[0076] In the formula, For the first Link fluctuation pattern index of the target link. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time.

[0077] Link Fluctuation Pattern Index The calculation formula, through multi-dimensional quantification, comprehensively considers link throughput, signal-to-noise ratio, and link load, aiming to evaluate the regularity and stability of the link during periodic quality fluctuations. Firstly, The throughput was subjected to exponential decay processing. Throughput reflects the load on the link; higher throughput indicates a heavier link load and potentially more severe fluctuations. Through exponential calculation, the impact of throughput fluctuations on stability is attenuated, thus preventing excessive amplification of the stability index by large throughput fluctuations. Next, The item standardizes the signal-to-noise ratio. It is the first Signal-to-noise ratio at any given time. This is the average signal-to-noise ratio (SNR) of the link. The standardized SNR eliminates differences in fluctuation amplitude between different links, making comparisons between them fairer and more consistent. Finally, The link load was logarithmically smoothed, as link load is a crucial factor affecting link stability; both excessively high and low loads can disrupt the regularity of fluctuations. Smoothing load fluctuations with a logarithmic function effectively mitigates the impact of extreme load changes on the regularity of fluctuations, thereby improving the accuracy of link stability assessment. In summary, these computational steps enable the link fluctuation regularity index to accurately reflect the stability and regularity of the link during periodic fluctuations, providing a scientific basis for subsequent path optimization.

[0078] No. Link fluctuation pattern index of target link The magnitude of the link volatility index is closely related to the regularity of link quality fluctuations during periodic fluctuations. A higher link volatility index indicates that the signal and noise of the link are relatively stable during periodic quality fluctuations, and the fluctuations in link load and throughput are relatively regular, indicating that the link quality fluctuations exhibit a certain degree of periodicity and predictability. Therefore, links with a higher volatility index have more stable and predictable quality fluctuations. Conversely, a lower link volatility index means that the link quality fluctuations are more chaotic, and there may be irregular or random fluctuation patterns. This indicates that the link is less stable during periodic fluctuations, with higher volatility and greater unpredictability. By calculating the link volatility index, the system can quantify the regularity of quality fluctuations for each target link, thus providing a basis for subsequent path optimization. This allows for more effective optimization strategies to be adopted on links with large fluctuations, ensuring the stability and reliability of network communication.

[0079] A link quality assessment model is constructed based on the link fluctuation amplitude index and link fluctuation regularity index of each generated target link, and the quality fluctuation coefficient of each target link is generated.

[0080] In this embodiment, the link fluctuation amplitude index of each generated target link is... Link fluctuation pattern index A link quality assessment model is constructed, and the quality fluctuation coefficient of each target link is generated by weighted summation. The specific calculation formula is as follows:

[0081]

[0082] In the formula, For the first The quality fluctuation coefficient of the target link, and The link fluctuation amplitude index for each target link Link fluctuation pattern index The non-zero weight coefficients, and .

[0083] In the link quality assessment model, a quality fluctuation coefficient is generated for each target link. This is achieved through a weighted summation method. First, the link fluctuation amplitude index... Link fluctuation pattern index The intensity and regularity of link fluctuations were quantified separately. By weighted summing of these two indices, the overall situation of link quality fluctuations can be comprehensively assessed. Specifically, In the formula, and It is a non-zero weight coefficient, and satisfies This means that the sum of the two weighting coefficients is 1, which is used to balance the contributions of the link fluctuation amplitude index and the link fluctuation regularity index to the quality fluctuation coefficient. The fluctuation index represents the weight of the fluctuation index in the calculation. Generally, when the quality fluctuation of the link is large, the fluctuation index will have a higher weight. These represent the weights of the volatility index, suitable for assessing links with relatively stable volatility patterns. By adjusting these two weighting coefficients, the impact on the assessment of different types of links can be flexibly controlled, ensuring that the link volatility coefficient can be dynamically optimized according to specific network conditions, ultimately providing accurate data support for path optimization and traffic scheduling.

[0084] An evaluation analysis is conducted based on the quality fluctuation coefficients of each target link to assess the degree of periodic fluctuation in link quality during periodic quality fluctuations. Based on the evaluation results, each target link is classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links.

[0085] In this embodiment, a pre-set threshold range for the quality fluctuation coefficient is determined. And after determination, it is compared with the quality fluctuation coefficients of each generated target link. A comparison was conducted, and the degree of periodic fluctuation in link quality during the periodic quality fluctuation period was evaluated based on the comparison results. Based on the evaluation results, each target link was classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links. The specific comparison analysis and classification are as follows:

[0086] like The target link exhibits low volatility in link quality during the periodic quality fluctuation period, thus classifying it as a low-volatility link.

[0087] This situation indicates that the link experiences relatively small and stable quality fluctuations during periods of periodic quality volatility. Low fluctuation amplitude typically indicates reliable transmission quality and minimal susceptibility to external interference. For such low-fluctuation links, the system can transmit critical task data, such as power dispatch instructions or fault diagnosis data, relatively stably, usually without requiring frequent path adjustments. The impact of this type of link is that it provides higher communication stability with minimal negative impact on overall network performance; therefore, it can be prioritized for applications requiring high reliability and low latency.

[0088] like The target link exhibits moderate fluctuations in link quality during the periodic quality fluctuation period, thus classifying it as a moderate fluctuation link.

[0089] This indicates that the link's quality fluctuation is at a moderate level. While the link can still maintain normal transmission performance, periodic quality fluctuations may cause delays or temporary instability in some data transmissions. For such links, while there's no need for excessive concern, their performance under high load or significant fluctuations should be considered when selecting paths. The impact of moderately fluctuating links is that the system should allocate traffic appropriately during path selection, avoiding placing too much critical workload on these links to reduce potential risks caused by fluctuations. Performance can be optimized through redundant paths or timely adjustments to ensure efficient network operation.

[0090] like The target link exhibits high volatility in its link quality during periods of periodic quality fluctuations, thus classifying it as a high-volatility link.

[0091] This situation indicates that the link experiences significant quality fluctuations during periods of periodic quality volatility, exhibiting strong variability and potentially being frequently affected by external interference or changes in network load. Such links often fail to provide stable communication performance, especially in high-bandwidth or latency-sensitive tasks, potentially leading to data loss, increased latency, or frequent path switching. For highly volatile links, the system should avoid transmitting high-priority, real-time-critical data traffic through these links. Instead, measures such as optimizing path selection and enabling redundant links should be implemented to ensure network stability. The impact of highly volatile links necessitates additional network management measures to mitigate their negative impact on overall network performance and prevent disruption to the stability and reliability of critical communication tasks.

[0092] Determining a pre-defined threshold range for quality fluctuation coefficients can be achieved by analyzing historical link quality data and network performance requirements. First, quality fluctuation coefficient data for each link over a specific time period can be collected and processed using statistical analysis methods (such as mean and standard deviation) to derive a reasonable fluctuation range. Specifically, cluster analysis or percentile analysis (such as the distribution of data at the 95th and 99th percentiles) can be used to determine the distribution range of the fluctuation coefficients. This allows for the identification of common fluctuation intervals from the data, which can then be used as a benchmark for setting the threshold range. Second, network administrators or system designers adjust the upper and lower limits of the thresholds based on actual application requirements, link reliability requirements, and acceptable quality fluctuation ranges. For example, in certain critical tasks, a lower threshold might be set as the minimum requirement for low-fluctuation links, while a higher threshold might be set as the maximum acceptable fluctuation range for high-fluctuation links. To ensure the accuracy and dynamic adaptability of the thresholds, the system should be regularly updated and optimized based on new link quality data. Machine learning algorithms (such as adaptive threshold adjustment) can be used to adjust the threshold range based on real-time data, thereby maintaining the accuracy and flexibility of link classification.

[0093] Based on the classification of each target link, different path optimization measures are implemented for each type of link;

[0094] In this embodiment, based on the division results of each target link, different path optimization measures are implemented for each type of link, specifically:

[0095] For highly volatile links, the specific path optimization measures adopted are as follows: immediately adjust the path selection strategy, including: prioritizing links with high signal strength and low load to avoid frequent path switching and reduce the impact of link fluctuations; enabling redundant links to distribute the load and ensure that highly volatile links do not cause overload and communication interruption; dynamically adjusting the network load balancing strategy to optimize traffic distribution between links and reduce the load pressure on highly volatile links; sending high-priority scheduling warnings to the network control center and activating automated control mechanisms to ensure that the real-time performance and reliability of communication tasks are not affected.

[0096] For highly volatile links, the core of path optimization is to address frequent link quality fluctuations through real-time monitoring and dynamic adjustment strategies. First, links with high signal strength and low load are prioritized. This is achieved by collecting real-time signal strength and load data (such as bandwidth utilization and latency) from the network monitoring platform and dynamically selecting paths with lower load and stronger signals based on this data. This can be automated by the network management system to ensure link stability. Frequent path switching is avoided because it leads to communication instability and higher latency. The system optimizes path selection using path selection algorithms (such as shortest path algorithms and load balancing algorithms) to reduce the frequency of link switching. Redundant link activation is implemented by deploying a redundant path monitoring system. When the primary link experiences excessive fluctuations, the system automatically switches to a backup link or distributes traffic to redundant links through dynamic traffic allocation to avoid overload. Finally, through automated control mechanisms and a scheduling and early warning system, the system can send real-time warnings to the network control center and promptly initiate automatic adjustments to ensure that highly volatile links do not cause task loss or delays, thus guaranteeing real-time requirements.

[0097] For links with moderate fluctuations, the specific path optimization measures are as follows: optimize the path selection strategy, dynamically adjust the path according to changes in network traffic, and avoid carrying too much critical task data on links with moderate fluctuations; adjust the link scheduling priority according to changes in the link fluctuation index and throughput to ensure that links with moderate fluctuations do not experience performance degradation under high load; and regularly check the network load balancing strategy and make corresponding adjustments to ensure that the link load is not concentrated on links with moderate fluctuations, thereby reducing potential latency or congestion problems.

[0098] Path optimization measures for moderately volatile links primarily aim to balance link fluctuations and load pressure, ensuring that the link does not experience excessive performance fluctuations when carrying traffic. Optimized path selection strategies can be implemented through a dynamic traffic monitoring system that tracks link load changes in real time and automatically adjusts paths based on network traffic variations. This can be achieved through an intelligent traffic scheduling system, prioritizing load-balanced and stable links for task allocation when network load is high. Link scheduling priority adjustments can be achieved through priority queue management and scheduling algorithms. The system dynamically adjusts the priority of each link, ensuring that critical tasks are not transmitted to moderately volatile links under high load, thus avoiding performance degradation. Regularly checking load balancing strategies relies on continuous network load monitoring and performance analysis to ensure reasonable allocation of network resources and prevent overload of any single link, which could impact overall communication performance. Through the combination of these strategies, the transmission performance of moderately volatile links is ensured to remain stable, preventing severe performance degradation under high load.

[0099] For low-fluctuation links, the specific path optimization measures are as follows: maintain the current path selection strategy unchanged, prioritize the use of low-fluctuation links to transmit latency-sensitive tasks, and ensure their transmission stability; regularly monitor low-fluctuation links and optimize and adjust them according to the actual needs of network traffic to avoid unnecessary load on the links; ensure that low-fluctuation links remain efficient and stable in long-term operation by continuously monitoring link quality fluctuations, and reserve appropriate redundant paths for unforeseen needs.

[0100] Low-fluctuation links possess strong stability, and the primary path optimization measure leverages this stability to handle latency-sensitive tasks. Maintaining the current path selection strategy is crucial because low-fluctuation links inherently offer high transmission stability, minimizing the need for frequent path adjustments and avoiding unnecessary switching overhead. By dynamically monitoring link performance, particularly throughput, latency, and bit error rate, the system can assess link status in real time, ensuring consistent performance even under high traffic. If a link experiences excessive load or slight fluctuations, the system automatically performs traffic scheduling, adjusting data flow as needed to prevent overloading. Regularly monitoring link quality, especially the long-term stability of low-fluctuation links, ensures efficient operation even with changing network environments. Redundant path reservations ensure rapid switching to backup links in case of problems with low-fluctuation links, preventing service interruptions and guaranteeing task continuity and reliability. This approach effectively utilizes the advantages of low-fluctuation links while mitigating potential risks that could threaten system stability through redundancy.

[0101] Based on the implemented path optimization measures, the quality fluctuations of each link are continuously monitored, and the path optimization effect is tracked in real time. The path selection strategy is automatically adjusted to ensure the continuous, stable and efficient execution of power communication tasks.

[0102] Continuous monitoring of link quality fluctuations can be achieved through a real-time link quality monitoring system. The system should continuously collect real-time data from the links, such as key indicators like bit error rate, signal strength, throughput, latency, and link load. This data is acquired periodically through interaction between the network monitoring platform and the link devices to ensure timeliness and accuracy. The data acquisition system can store data based on a time-series database (such as InfluxDB) to ensure that historical and real-time link quality data are updated synchronously. By comparing current link quality data with set standard thresholds or historical performance data, the system can detect link quality fluctuations in real time and compare the deviation between actual data and optimization targets, thereby understanding the effectiveness of path optimization.

[0103] To ensure the continuous, stable, and efficient execution of power communication tasks, the system should possess the ability to automatically adjust its path selection strategy. Through intelligent path scheduling algorithms, the system can automatically trigger a path switching mechanism when abnormal fluctuations in link quality or load changes are detected. For example, when link quality fluctuations exceed a set threshold, the system can automatically select a higher-quality alternative link using preset rules or a self-learning mechanism, ensuring the task is not affected. Furthermore, a load balancing mechanism can help distribute network traffic, ensuring links do not operate under overload. When the system detects that a link is overloaded or its quality fluctuations are aggravated, it will dynamically adjust traffic allocation or switch to a redundant link, thereby effectively reducing network risks and improving the reliability and efficiency of task execution. This facilitates the system's adaptive optimization, ensuring that power communication tasks can be completed efficiently and stably under different network environments and task requirements.

[0104] The above formulas are all dimensionless calculations. The formulas are derived from software simulations based on a large amount of collected data to obtain the most recent real-world results. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0105] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0106] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0107] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0108] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0109] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

Claims

1. A method for optimizing optical transmission service paths in power communication based on reinforcement learning, characterized in that, Specifically, the following steps are included: In power communication optical transmission networks, each link used for transmitting communication tasks is identified and monitored in real time. Based on the fluctuation pattern of link quality, all links exhibiting periodic quality fluctuations are selected and designated as target links. The link quality information generated by each target link during periodic quality fluctuations is acquired in real time, and analyzed after acquisition to generate the link fluctuation amplitude index and link fluctuation pattern index for each target link. A link quality assessment model is constructed based on the link fluctuation amplitude index and link fluctuation regularity index of each generated target link, and the quality fluctuation coefficient of each target link is generated. An evaluation analysis is conducted based on the quality fluctuation coefficients of each target link to assess the degree of periodic fluctuation in link quality during periodic quality fluctuations. Based on the evaluation results, each target link is classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links. Based on the classification of each target link, different path optimization measures are implemented for each type of link; Based on the implemented path optimization measures, the quality fluctuations of each link are continuously monitored, and the path optimization effect is tracked in real time. The path selection strategy is automatically adjusted to ensure the continuous, stable and efficient execution of power communication tasks.

2. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 1, characterized in that, The link quality information generated by each target link during periodic quality fluctuations is acquired in real time and analyzed to generate a link fluctuation amplitude index and a link fluctuation pattern index for each target link. The specific steps include: The link quality information generated by each target link during periodic quality fluctuations is acquired in real time and preprocessed after acquisition. Extract link quality fluctuation information and link fluctuation stability information from the link quality information generated by each target link during periodic quality fluctuations after preprocessing; The extracted link quality fluctuation information and link fluctuation stability information are analyzed to generate the link fluctuation amplitude index and link fluctuation regularity index for each target link.

3. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 2, characterized in that, The logic for obtaining the link fluctuation amplitude index for each target link is as follows: Link quality fluctuation information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the bit error rate, signal strength, and successfully transmitted data volume at different times within a certain period of time for each target link during periodic quality fluctuations, and these are categorized as follows: , and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Bit error rate at time step Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal strength at any given time Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time. , , and All are positive integers; The link fluctuation amplitude index for each target link is calculated using the following formula: ； In the formula, For the first Link fluctuation amplitude index of the target link.

4. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 3, characterized in that, The logic for obtaining the link fluctuation pattern index for each target link is as follows: Link fluctuation stability information is extracted from the link quality information generated by each target link during periodic quality fluctuations after preprocessing. Specifically, this includes the signal-to-noise ratio and data traffic carried by each target link at different times within a certain period during the periodic quality fluctuations, and these are calibrated as follows: and , Indicates the first The target link experiences a period of time during cyclical quality fluctuations. Signal-to-noise ratio at any given time. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data traffic carried at any given moment , , and All are positive integers; Calculate the mean signal-to-noise ratio (SNR) of each target link at different times during a period of periodic quality fluctuations. ,in accordance with: ; The link fluctuation pattern index for each target link is calculated using the following formula: ； In the formula, For the first Link fluctuation pattern index of the target link. Indicates the first The target link experiences a period of time during cyclical quality fluctuations. The amount of data successfully transmitted at any given time.

5. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 4, characterized in that, Link fluctuation amplitude index for each generated target link Link fluctuation pattern index A link quality assessment model is constructed, and the quality fluctuation coefficient of each target link is generated by weighted summation. The specific calculation formula is as follows: ； In the formula, For the first The quality fluctuation coefficient of the target link, and The link fluctuation amplitude index for each target link Link fluctuation pattern index The non-zero weight coefficients, and .

6. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 5, characterized in that, Determine the pre-set threshold range for the quality fluctuation coefficient. And after determination, it is compared with the quality fluctuation coefficients of each generated target link. A comparison was conducted, and the degree of periodic fluctuation in link quality during the periodic quality fluctuation period was evaluated based on the comparison results. Based on the evaluation results, each target link was classified into three categories: high-fluctuation links, medium-fluctuation links, and low-fluctuation links. The specific comparison analysis and classification are as follows: like The target link exhibits low volatility in link quality during the periodic quality fluctuation period, thus classifying it as a low-volatility link. like The target link exhibits moderate fluctuations in link quality during the periodic quality fluctuation period, thus classifying it as a moderate fluctuation link. like The target link exhibits high volatility in its link quality during periods of periodic quality fluctuations, thus classifying it as a high-volatility link.

7. The method for optimizing optical transmission service paths in power communication based on reinforcement learning according to claim 6, characterized in that, Based on the classification of each target link, different path optimization measures are implemented for each type of link, specifically: For highly volatile links, the specific path optimization measures adopted are as follows: immediately adjust the path selection strategy, including: prioritizing links with high signal strength and low load to avoid frequent path switching and reduce the impact of link fluctuations; enabling redundant links to distribute the load and ensure that highly volatile links do not cause overload and communication interruption; dynamically adjusting the network load balancing strategy to optimize traffic distribution between links and reduce the load pressure on highly volatile links; sending high-priority scheduling warnings to the network control center and activating automated control mechanisms to ensure that the real-time performance and reliability of communication tasks are not affected. For links with moderate fluctuations, the specific path optimization measures are as follows: optimize the path selection strategy and dynamically adjust the path according to changes in network traffic; adjust the link scheduling priority according to changes in the link fluctuation index and throughput to ensure that the performance of links with moderate fluctuations does not degrade under high load; and regularly check the network load balancing strategy and make corresponding adjustments to ensure that the link load is not concentrated on links with moderate fluctuations. For low-fluctuation links, the specific path optimization measures are as follows: maintain the current path selection strategy unchanged, prioritize the use of low-fluctuation links to transmit latency-sensitive tasks, and ensure their transmission stability; regularly monitor low-fluctuation links and optimize and adjust them according to the actual needs of network traffic; and ensure that low-fluctuation links remain efficient and stable in long-term operation by continuously monitoring link quality fluctuations.

Images