SDH / PDH payload real-time bandwidth analysis method based on time synchronization

Abstract

The invention provides an SDH / PDH payload real-time bandwidth analysis method based on time synchronization, and relates to the technical field of transmission networks, and the method comprises the steps: obtaining the network state information of each network working node in an optical transmission network; preprocessing the original data frame set to obtain an average effective load corresponding to each data frame in the data frame set, and calculating an effective load dynamic estimation value based on the average effective load and a historical average effective load; obtaining a network transmission resource allocation strategy of an optical link between any two adjacent network working nodes and a network transmission resource parameter corresponding to the network transmission resource allocation strategy according to a preset constraint condition and the effective load dynamic estimation value; and obtaining a comprehensive evaluation index of the current data frame according to the network transmission resource parameter and the load bandwidth performance function, and adjusting a flow distribution path of the optical transmission network according to the comprehensive evaluation index. According to the invention, the dynamic self-adaptive adjustment of the network flow can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of transmission networks, and in particular to a time synchronization-based SDH / PDH payload real-time bandwidth analysis method. Background Art

[0002] With the rapid development of internet technology and the rapid growth of network traffic, network operators are continuously increasing the density of server and storage device deployments, and data center network nodes and links are growing exponentially. Data centers have gradually become the hub of network traffic. The continuous increase in data center traffic, as well as the varying link and service quality requirements of different types of traffic, are placing higher demands on data center networks. However, most existing routing algorithms fail to comprehensively consider the real-time status of links and the characteristics of individual traffic flows. As a result, some links in the network may be overloaded while others remain idle, causing imbalanced network load.

[0003] The Chinese patent with announcement number CN114124827B discloses a method for optical network load balancing based on time-space-frequency variation coefficient evaluation. The method obtains network load information and current service information in real time; classifies service information to obtain service classification; searches for K shortest paths based on service classification and network load information; integrates the K shortest paths in the three dimensions of time, space, and frequency and calculates the variation coefficient to obtain a load balancing adjustment strategy; and allocates the current service to the corresponding link channel according to the load balancing adjustment strategy to complete service carrying. However, the above scheme relies on fixed path search and calculation strategies, which are not flexible enough to cope with the rapid changes in network environment and service requirements. Moreover, in large-scale or emergency situations, computational bottlenecks are prone to occur, resulting in a decrease in load balancing effect. Therefore, it is very necessary to provide a real-time bandwidth analysis method for SDH / PDH payload based on time synchronization to improve the load balancing effect of optical transmission networks. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a real-time bandwidth analysis method for SDHPDH payload based on time synchronization. By acquiring the status information of each node in the optical transmission network in real time, fine-grained network status perception is achieved. At the same time, the current data frame is comprehensively evaluated using the payload bandwidth performance function. This evaluation index can quantify the network transmission performance and is directly used for traffic path decision-making, thereby realizing dynamic adaptive adjustment of network traffic.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method for analyzing real-time bandwidth of SDH / PDH payload based on time synchronization, the method comprising: Acquire network status information of each network working node in the optical transmission network, wherein the network status information includes a set of original data frames; Preprocessing the original data frame set to obtain an average payload corresponding to each data frame in the data frame set, and calculating a dynamic payload estimate based on the average payload and a historical average payload; According to preset constraints and a dynamic estimate of the payload, a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes and network transmission resource parameters corresponding to the network transmission resource allocation strategy are obtained; According to the network transmission resource parameters and the load bandwidth performance function, a comprehensive evaluation index of the current data frame is obtained, and the traffic distribution path of the optical transmission network is adjusted according to the comprehensive evaluation index.

[0006] Based on the above technical solution, preferably, the preprocessing of the data frame set to obtain the average payload corresponding to each data frame in the data frame set specifically includes: Eliminate outliers from the data frame set to obtain a transition data frame set, and align timestamps of data frames from different network working nodes in the transition data frame set based on a synchronous clock to obtain payload data corresponding to each data frame in the transition data frame set; According to the sampling period of the optical transmission network, the continuous data frames in the transition data frame set are divided into a unified time window, and the payload data corresponding to the current data frame is weightedly fused with the historical payload data in the previous time window to obtain the average payload corresponding to each data frame.

[0007] On the basis of the above technical solution, preferably, obtaining a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes according to preset constraints and a dynamic estimate of the payload specifically includes: Determining whether a bandwidth utilization rate of an optical link between any two adjacent network working nodes is less than or equal to a maximum allowable bandwidth utilization rate based on a bandwidth utilization rate constraint condition in the preset constraint condition and the dynamic estimated value of the payload; If the bandwidth utilization of the optical link between any two adjacent network working nodes is greater than the maximum allowed bandwidth utilization, the current optical link is determined to be in a fault state, and the network status information closest to the current optical link is selected to regenerate a virtual optical link, and all traffic of the current optical link is allocated to the virtual optical link, and the bandwidth utilization of the virtual optical link is less than or equal to the maximum allowed bandwidth utilization.

[0008] Preferably, the obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload further includes: If the bandwidth utilization of the optical link between any two adjacent network working nodes is less than or equal to the maximum allowed bandwidth utilization, then based on the protection capacity of the current optical link and the link protection recovery constraint, determine whether the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter; If the protection capacity of the current optical link is less than the link fault recovery performance parameter, any network working node in the current optical link is used as the initial node, and another node is used as the target node. Based on the initial node, any network working node adjacent to the current optical link is selected as the transition node to reconstruct multiple virtual optical links, and the traffic ratio between the multiple virtual optical links and the current optical link is redistributed according to the preset ratio value.

[0009] Preferably, the obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload further includes: If the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter, then judging whether the protection capacity of the current optical link is greater than or equal to the wavelength channel load parameter based on the protection capacity of the current optical link and the wavelength channel protection constraint condition; If the protection capacity of the current optical link is less than the wavelength channel load parameter, multiple wavelength channels are expanded on the current optical link, and the traffic ratios of all wavelength channels on the current optical link are redistributed.

[0010] Preferably, the expressions of the bandwidth utilization constraint condition, the link protection and restoration constraint condition, and the wavelength channel protection constraint condition are respectively: ; ; ; in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, N Represents a collection of network work nodes, C total represents the protection capacity of all optical links, β represents the synchronous correction coefficient, i max Indicates the maximum allowed bandwidth utilization, l r Indicates the rThe protection capacity of the optical link, c represents the protection capacity adjustment factor based on time synchronization quality, t represents the time synchronization factor, or n Indicates the n The candidate light path selection index corresponding to each network working node, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Represents the candidate link mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to n Network working nodes, l Indicates the l wavelength channels, m l represents the protection recovery coefficient, Indicates network status s You can use the l A collection of optical links with wavelength channels, Represents the wavelength channel mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to l wavelength channels, J ( t ) means in t Time jitter indicator at each moment, Indicates the n Network working nodes are in network state s The candidate optical routing end set under .

[0011] Preferably, the adjusting the traffic distribution path of the optical transmission network according to the comprehensive evaluation index specifically includes: Identifying resource bottleneck events on each optical link in the optical transmission network according to comprehensive evaluation indicators; The network status information corresponding to the resource bottleneck event and the historical network status information are used to recalculate the optimal path using a dynamic routing algorithm, and the traffic distribution path is scheduled in real time.

[0012] Preferably, the expression of the load-bandwidth performance function is: The expression of the load bandwidth performance function is: ; in, maxY ( t ) represents the optical link network under different constraints. tThe load-bandwidth performance function at time , K Represents the set of all data frames that need to be transmitted in the optical link network, TD ( W k ) indicates the k The transmission cost corresponding to the frame bandwidth allocation, W k Indicates the k Bandwidth allocation of frames, H k Indicates the k The sampling period of the frame, t represents the time synchronization factor, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Z Indicates the preset traffic scaling factor in the optical link network. ( t ) means in t Time r The link in n Real-time traffic on network worker nodes, Indicates the n Network working nodes are in network state s The set of candidate optical routing ends under D n ( t ) means in t Time n The estimated value of the payload of each network worker node.

[0013] Furthermore, the network transmission resource parameters include one or more of bandwidth utilization, link failure recovery performance parameters, and wavelength channel load parameters.

[0014] Furthermore, the optical transmission network includes one or more of Synchronous Digital Hierarchy (SDH) and Plesiochronous Digital Hierarchy (PDH).

[0015] The beneficial effects of the present invention are as follows: 1. By acquiring the status information of each node in the optical transmission network in real time, fine-grained network status perception is achieved. By collecting various raw data frame information, multi-dimensional parameters of network transmission performance can be monitored. By comparing the average payload of the current data frame with historical data, a dynamic load estimation model can be constructed. This model not only reflects the current network load situation but also captures the trend of load changes, enabling rapid response to burst traffic or temporary network anomalies, so as to adjust resource allocation strategies in a timely and dynamic manner to avoid resource waste or bottleneck problems. At the same time, through preset constraints and dynamic load estimates, targeted resource allocation strategies are formulated to maximize resource utilization. By comprehensively considering bandwidth utilization, link fault recovery capability, and wavelength channel load parameters, the strategy can simultaneously take into account network performance, reliability, and redundancy design. The load-bandwidth performance function is used to comprehensively evaluate the current data frame, realizing dynamic adaptive adjustment of network traffic, which can greatly reduce device link redundancy and thus improve the load balancing effect of the optical transmission network.

[0016] 2. When bandwidth utilization is within the allowable range, the protection capacity of the current optical link and the link fault recovery performance parameters are judged to accurately assess the link protection capability. This helps prioritize the fault recovery performance of key links when resources are sufficient, ensuring the overall safe and stable operation of the network. When insufficient protection capacity is detected, adjacent working nodes are selected as transition nodes, multiple virtual optical links are reconstructed, and traffic is dynamically redistributed to achieve balanced adjustment of link loads. At the same time, traffic transmission between links is finely controlled with preset ratios, effectively preventing link failures and service interruptions caused by single point failures. Reconstructing multiple virtual optical links provides multi-path disaster recovery capabilities for link fault recovery. Even if a local link failure occurs, traffic can still continue to be transmitted through other paths, improving the overall reliability and business continuity of the network and realizing flexible scheduling of physical link resources. This not only improves link disaster recovery capabilities, but also enables priority allocation of resources when necessary, maximizing the resource utilization and transmission performance of the overall network. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 A flow chart of a method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization provided by the present invention; Figure 2 This is a structural diagram of optical link switching when a network working node fails provided by the present invention. DETAILED DESCRIPTION

[0018] Example 1: A method for analyzing real-time bandwidth of SDH / PDH payload based on time synchronization, the method comprising: Acquire network status information of each network working node in the optical transmission network, wherein the network status information includes a set of original data frames; Preprocessing the original data frame set to obtain an average payload corresponding to each data frame in the data frame set, and calculating a dynamic payload estimate based on the average payload and a historical average payload; According to preset constraints and a dynamic estimate of the payload, a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes and network transmission resource parameters corresponding to the network transmission resource allocation strategy are obtained; According to the network transmission resource parameters and the load bandwidth performance function, a comprehensive evaluation index of the current data frame is obtained, and the traffic distribution path of the optical transmission network is adjusted according to the comprehensive evaluation index.

[0019] Based on the above technical solution, preferably, the preprocessing of the data frame set to obtain the average payload corresponding to each data frame in the data frame set specifically includes: Eliminate outliers from the data frame set to obtain a transition data frame set, and align timestamps of data frames from different network working nodes in the transition data frame set based on a synchronous clock to obtain payload data corresponding to each data frame in the transition data frame set; According to the sampling period of the optical transmission network, the continuous data frames in the transition data frame set are divided into a unified time window, and the payload data corresponding to the current data frame is weightedly fused with the historical payload data in the previous time window to obtain the average payload corresponding to each data frame.

[0020] On the basis of the above technical solution, preferably, obtaining a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes according to preset constraints and a dynamic estimate of the payload specifically includes: Determining whether a bandwidth utilization rate of an optical link between any two adjacent network working nodes is less than or equal to a maximum allowable bandwidth utilization rate based on a bandwidth utilization rate constraint condition in the preset constraint condition and the dynamic estimated value of the payload; If the bandwidth utilization of the optical link between any two adjacent network working nodes is greater than the maximum allowed bandwidth utilization, the current optical link is determined to be in a fault state, and the network status information closest to the current optical link is selected to regenerate a virtual optical link, and all traffic of the current optical link is allocated to the virtual optical link, and the bandwidth utilization of the virtual optical link is less than or equal to the maximum allowed bandwidth utilization.

[0021] Preferably, the obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload further includes: If the bandwidth utilization of the optical link between any two adjacent network working nodes is less than or equal to the maximum allowed bandwidth utilization, then based on the protection capacity of the current optical link and the link protection recovery constraint, determine whether the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter; If the protection capacity of the current optical link is less than the link fault recovery performance parameter, any network working node in the current optical link is used as the initial node, and another node is used as the target node. Based on the initial node, any network working node adjacent to the current optical link is selected as the transition node to reconstruct multiple virtual optical links, and the traffic ratio between the multiple virtual optical links and the current optical link is redistributed according to the preset ratio value.

[0022] Preferably, the obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload further includes: If the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter, then judging whether the protection capacity of the current optical link is greater than or equal to the wavelength channel load parameter based on the protection capacity of the current optical link and the wavelength channel protection constraint condition; If the protection capacity of the current optical link is less than the wavelength channel load parameter, multiple wavelength channels are expanded on the current optical link, and the traffic ratios of all wavelength channels on the current optical link are redistributed.

[0023] Preferably, the expressions of the bandwidth utilization constraint condition, the link protection and restoration constraint condition, and the wavelength channel protection constraint condition are respectively: ; ; ; in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, N Represents a collection of network work nodes, C total represents the protection capacity of all optical links, β represents the synchronous correction coefficient, i max Indicates the maximum allowed bandwidth utilization, l r Indicates the rThe protection capacity of the optical link, c represents the protection capacity adjustment factor based on time synchronization quality, t represents the time synchronization factor, or n Indicates the n The candidate light path selection index corresponding to each network working node, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Represents the candidate link mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to n Network working nodes, l Indicates the l wavelength channels, m l represents the protection recovery coefficient, Indicates network status s Next you can use l A collection of optical links with wavelength channels, Represents the wavelength channel mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to l wavelength channels, J ( t ) means in t Time jitter indicator at each moment, Indicates the n Network working nodes are in network state s The candidate optical routing end set under .

[0024] Preferably, the adjusting the traffic distribution path of the optical transmission network according to the comprehensive evaluation index specifically includes: Identifying resource bottleneck events on each optical link in the optical transmission network according to comprehensive evaluation indicators; The network status information corresponding to the resource bottleneck event and the historical network status information are used to recalculate the optimal path using a dynamic routing algorithm, and the traffic distribution path is scheduled in real time.

[0025] Preferably, the expression of the load-bandwidth performance function is: The expression of the load bandwidth performance function is: ; in, maxY ( t ) represents the optical link network under different constraints. tThe load-bandwidth performance function at time , K Represents the set of all data frames that need to be transmitted in the optical link network, TD ( W k ) indicates the k The transmission cost corresponding to the frame bandwidth allocation, W k Indicates the k Bandwidth allocation of frames, H k Indicates the k The sampling period of the frame, t represents the time synchronization factor, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Z Indicates the preset traffic scaling factor in the optical link network. ( t ) means in t Time r The link in n Real-time traffic on network worker nodes, Indicates the n Network working nodes are in network state s The set of candidate optical routing ends under D n ( t ) means in t Time n The estimated value of the payload of each network worker node.

[0026] Furthermore, the network transmission resource parameters include one or more of bandwidth utilization, link failure recovery performance parameters, and wavelength channel load parameters.

[0027] Furthermore, the optical transmission network includes one or more of Synchronous Digital Hierarchy (SDH) and Plesiochronous Digital Hierarchy (PDH).

[0028] Example 2: like Figure 1 As shown, the present invention discloses a real-time bandwidth analysis method for SDH / PDH payload based on time synchronization, the steps of the method are as follows: Step S1 : obtaining network status information of each network working node in an optical transmission network, wherein the network status information includes a set of original data frames; the optical transmission network includes one or more of Synchronous Digital Hierarchy (SDH) and Plesiochronous Digital Hierarchy (PDH).

[0029] In this embodiment, monitoring modules are installed at key optical nodes in the optical transmission network, such as transmission nodes or cross-connect nodes, to ensure that transmission behavior throughout the entire link is captured. Monitoring equipment, such as OTDRs and optical power meters, can be used to measure basic indicators such as optical signal strength and attenuation. Dedicated high-speed data acquisition cards or embedded processors are used to capture SDH / PDH frame data in real time after optical-to-electrical conversion. High-precision atomic clocks or GPS clocks are also introduced to ensure the internal clock accuracy of the nodes to meet network time synchronization requirements.

[0030] Each monitoring module captures optical signals passing through the node in real time and converts them into digital signals using a high-speed analog-to-digital converter (ADC). This captured data is pre-segmented to extract the start identifier, header information, and payload of the SDH / PDH frame. Internal logic parses the captured digital signal according to the SDH / PDH frame structure and reconstructs the data frame. The data frame contains key information such as payload information, error detection code, frame sequence number, and timestamp.

[0031] Furthermore, satellite positioning, such as GPS or high-precision atomic clocks, provides a unified time reference for the entire network. Clock synchronization within the optical transmission network typically utilizes the IEEE 1588 Precision Time Protocol (PTP) or other dedicated synchronization protocols to ensure that the time error at each node remains within an acceptable range. Each network node regularly receives synchronization signals from a central time server or master clock node, automatically corrects local clock deviations based on preset thresholds, and records the timestamp of each adjustment as feedback to the data acquisition system for subsequent data alignment. A centralized management terminal or network monitoring center then integrates the data frames from each network node to construct a unified set of raw data frames.

[0032] Step S2: pre-processing the original data frame set to obtain the average payload corresponding to each data frame in the data frame set, and calculating a dynamic payload estimation value based on the average payload and the historical average payload.

[0033] This step also includes steps S21 to S22.

[0034] Step S21, remove outliers from the data frame set to obtain a transition data frame set, and align the timestamps of the data frames from different network working nodes in the transition data frame set based on the synchronous clock to obtain the payload data corresponding to each data frame in the transition data frame set.

[0035] In this step, the mean, median, variance, or standard deviation is calculated for key fields in the data frame, such as payload length, bit error rate, and timestamp offset. The standard deviation (±k times the mean) or the interquartile range (IQR) method is used as the normal fluctuation range. An appropriate k value is selected to filter out data that deviates significantly from the mean. Hard upper and lower limits are set based on device characteristics and historical data. For example, if the payload falls below a certain lower limit or exceeds the maximum load capacity, it is directly identified as an anomaly. Each frame is checked for abnormal indicators such as checksum errors and discontinuous frame numbers. If present, the data is marked as an outlier. The data is partitioned into blocks, and a sliding window (e.g., 5-10 frames) is used to perform mean filtering on the local data. The current frame data is determined to determine whether it differs significantly from the mean of the window. If it exceeds the set tolerance, it is identified as an outlier and removed. A joint judgment based on multiple key indicators, such as payload, error detection code, and timestamp consistency, reduces the risk of misjudgment based on a single indicator.

[0036] Furthermore, a unified absolute time reference is provided for each node through GPS or high-precision atomic clocks. Each node uses IEEE 1588PTP or similar protocols to receive synchronization signals from the central time server and adjust the local clock error. Each network working node regularly compares the local timestamp with the central synchronization clock, records and corrects the deviation caused by device drift or network delay. The deviation information after each correction is appended to the data frame for use in the data alignment process. Within each buffer window, all data frames are sorted according to the corrected timestamp to ensure the continuity of the time series. The data frames from different nodes with aligned timestamps are integrated into a unified set of transition data frames with a complete time series. The time interval of continuous data frames is detected, and interpolation or frame skipping correction is used when necessary to further ensure the continuity and error-free of global data.

[0037] According to the SDH / PDH frame protocol standard, identify the start identifier, header field, and payload field in the frame, split each field out in turn, and focus on parsing the payload part. Extract the payload field containing the actual business data from the parsed frame structure. Format the extracted payload data, such as converting it into a readable binary or integer data array, and confirm whether the content of each frame is complete and correct based on the error detection code provided in the frame, such as the CRC checksum. If redundant or repeated information is found in multiple frames of data, deduplication is performed by comparing the frame sequence number and timestamp to ensure that the payload data is unique and accurate. The payload data of each frame is fused or compared with the historical data over a period of time, and the local average or dynamic change index is calculated. The extracted payload data is stored in the real-time analysis module or database for further bandwidth estimation and network resource scheduling.

[0038] Step S22: Divide the continuous data frames in the transition data frame set into a unified time window according to the sampling period of the optical transmission network, and perform weighted fusion using the payload data corresponding to the current data frame and the historical payload data in the previous time window to obtain the average payload corresponding to each data frame.

[0039] In this step, a fixed sampling period is set according to the design requirements of the optical transmission network. T , which determines the basic time unit for data frame division. For the transition data frame set, each data frame has a timestamp that has been synchronized. Taking the timestamp of the first data frame as the starting point, the timeline is divided into multiple equal time windows, and the length of each window is T For all data frames within the current time window, the immediate payload statistics within that window can be calculated. Combined with the historical average value of the previous window, the above formula is used to fuse historical data with the current sampling data, reducing errors caused by instantaneous fluctuations.

[0040] In one example, the payload dynamics estimation constraint is expressed as: ; in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, α represents the weight of the flow measurement value, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Z Indicates the preset traffic scaling factor in the optical link network. ( t ) means in t Time r The link in n Real-time traffic on network worker nodes, Indicates the n Network working nodes are in network state S The set of candidate optical routing ends under Represents a past time window T The average payload value within .

[0041] Step S3: Based on the preset constraints and the dynamic payload estimate, a network transmission resource allocation strategy for the optical link between any two adjacent network working nodes and network transmission resource parameters corresponding to the network transmission resource allocation strategy are obtained. The network transmission resource parameters include one or more of bandwidth utilization, link fault recovery performance parameters, and wavelength channel load parameters.

[0042] This step also includes steps S31 to S36.

[0043] Step S31 : judging whether the bandwidth utilization of the optical link between any two adjacent network working nodes is less than or equal to the maximum allowed bandwidth utilization according to the bandwidth utilization constraint in the preset constraint and the dynamic estimation value of the effective load.

[0044] In this step, the bandwidth utilization constraint is expressed as: ; in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, N Represents a collection of network work nodes, C total represents the protection capacity of all optical links, β represents the synchronous correction coefficient, i max Indicates the maximum allowed bandwidth utilization, t The bandwidth utilization constraint is expressed as the ratio of the sum of the dynamic estimated values ​​of the payload of each sub-channel or data frame on the current optical link to the total bandwidth available for carrying data on the current optical link. β It is the synchronization correction coefficient of time synchronization to the upper limit of bandwidth utilization, usually ranging from 0.3 to 0.5. i max The preset maximum allowed bandwidth utilization is generally set in the range of 0.8 to 0.95.

[0045] Step S32: If the bandwidth utilization of the optical link between any two adjacent network working nodes is greater than the maximum allowed bandwidth utilization, it is determined that the current optical link is in a fault state, and the network status information closest to the current optical link is selected to regenerate a virtual optical link, and all traffic of the current optical link is allocated to the virtual optical link, and the bandwidth utilization of the virtual optical link is less than or equal to the maximum allowed bandwidth utilization.

[0046] In this step, based on the result of bandwidth utilization exceeding the standard determined in S31, it is determined that the current optical link is in a faulty or restricted state and cannot continue to carry all traffic. The status information of each working node and its adjacent links in the current network is collected in real time from the network monitoring platform. The data content includes but is not limited to key parameters such as the current load of the link, historical traffic, signal-to-noise ratio, delay, physical distance, and link stability. In the collected information, the links adjacent to the working nodes at both ends of the current faulty link are compared, and the links with close physical locations are given priority, because short distances usually mean lower delays and better signal quality. With reference to the operating data of previous links, links with smaller fluctuations and lower failure rates over a long period of time are selected. Using the nearest neighbor algorithm or a comprehensive scoring system, for example, different weights are assigned to score according to physical distance, bandwidth utilization, signal-to-noise ratio, etc., and the most qualified candidate link or multiple link states are determined as the "closest" network status information to provide a basis for building a virtual link.

[0047] Furthermore, based on the real-time status information of candidate links, the basic parameters of the virtual link are determined, and the maximum bandwidth of the virtual link is set to ensure that it can carry all traffic on the failed link while remaining within the preset maximum bandwidth utilization range. The latency of the candidate links is comprehensively considered to ensure that the virtual link meets the requirements for time synchronization and data transmission. Based on the selected candidate links, they are logically integrated into one or more virtual links. A network topology algorithm is used to determine the node sequence and transmission path of the virtual link. Typically, virtual links can be implemented by splicing adjacent links or merging parallel channels. At the physical layer, virtual links achieve logical path reconstruction through tunneling technology and traversal protocols such as MPLS, GMPLS, or SDN segmentation, ensuring seamless data transmission on the new link.

[0048] During traffic redistribution, dynamic routing and load balancing algorithms are used to reassign traffic from the failed link to the virtual link according to pre-set policies, such as proportional distribution or priority scheduling. Buffering mechanisms and real-time monitoring ensure that data is not lost or delayed during the redistribution. After the traffic is redistributed, the real-time bandwidth utilization and other performance indicators of the virtual link are continuously monitored. If the bandwidth utilization of the new link is still close to or exceeds the upper limit, the system may need to further adjust the diversion ratio or search for other links in the alternative network state as backup.

[0049] In one example, if Figure 2 As shown in the figure, assume that two adjacent network nodes A and B communicate via an optical link. The total capacity of this optical link is 100 Gbps. After dynamic payload estimation and weighted fusion processing, the total payload data accumulated on this link at the current moment reaches 95 Gbps, resulting in a calculated bandwidth utilization of approximately 95%.

[0050] The maximum allowed bandwidth utilization in the system preset parameters i max =90%; time synchronization factor t =85%; Synchronous correction coefficient β =0.4.

[0051] The upper limit of bandwidth usage is calculated using the formula: UpperBound= bt +(1–β) i max =0.88. Since the current link utilization rate is 95%>88%, the system determines that the optical link between AB is in a faulty or overloaded state, and needs to trigger step S32 for protection recovery.

[0052] First, information on all current adjacent links is collected from the entire network status monitoring platform. Assume that there are other candidate links near A and B. For example, a link between A and C has a total capacity of 100 Gbps, a current payload of approximately 60 Gbps, and a utilization rate of 60%, and is physically close. There is also a link between C and B with a total capacity of 100 Gbps, a current payload of 65 Gbps, and a utilization rate of 65%, operating normally. By comparing geographic location, link status (such as load, latency, signal-to-noise ratio), and historical stability, links A–C and C–B are identified as candidate backup paths. After evaluation using a nearest neighbor algorithm or a comprehensive scoring method, node C is selected as the intermediate node, forming a new virtual optical link solution: a virtual link formed by connecting the two links A→C and C→B in series, whose overall status and load balancing meet the preset requirements.

[0053] Based on the selected candidate links, both A–C and C–B have a total capacity of 100 Gbps, and their current utilization rates are 60% and 65%, respectively, both well below the preset upper limit. When establishing the virtual link, the system sets the capacity of the virtual link to 100 Gbps, taking the higher link capacity as the basis, and integrates the physical parameters of the two links to ensure a logically "continuous" transmission channel. During the virtual link establishment process, tunneling technologies such as MPLS, GMPLS, or SDN control can be used for encapsulation and path planning, ensuring seamless data transmission from A to B via C. After adjusting the virtual link parameters, it is expected that the bandwidth utilization of the link when carrying traffic will be stable, for example, between 80% and 85%, well below the threshold of 88%.

[0054] Upon detecting that the original link AB is at 95% utilization, the system immediately initiates the traffic migration module. Based on the traffic scheduling algorithm, the system redistributes the 95 Gbps of traffic on the original link AB to the newly generated virtual link A→C→B in a certain proportion. This scheduling process may split the overall traffic into two parts: the primary traffic, for example, 85 Gbps, is migrated to virtual link A→C→B; and based on the remaining network bandwidth and real-time monitoring, the system may temporarily retain some traffic on other backup links or wait for further load reduction. During implementation, the system monitors the real-time bandwidth utilization of the virtual link to ensure that after adjustment, the virtual link utilization drops to, for example, 85%, below the upper limit of 88%. If monitoring detects that the utilization of the new link has risen to a critical value, a secondary adjustment is triggered or candidate links are added for load sharing to ensure overall network stability.

[0055] In this example, the original optical link A–B was identified as faulty because its utilization reached 95%, exceeding the permitted 88%. The system then selected candidate links A–C and C–B, both in good condition and physically close to each other, to create a virtual optical link. Through dynamic scheduling, the original 95 Gbps traffic was migrated to the new link. After scheduling and adjustment, the bandwidth utilization of the new link stabilized at around 85%, meeting the maximum permitted bandwidth utilization and ensuring data transmission continuity and overall network security.

[0056] Step S33: If the bandwidth utilization of the optical link between any two adjacent network working nodes is less than or equal to the maximum allowed bandwidth utilization, then determine whether the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter based on the protection capacity of the current optical link and the link protection recovery constraint condition.

[0057] In this step, the link protection restoration constraint condition is expressed as: ; ; ; ; in, l r Indicates the r The protection capacity of the optical link, c represents the protection capacity adjustment factor based on time synchronization quality, t represents the time synchronization factor, or n Indicates the n The candidate light path selection index corresponding to each network working node, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Represents the candidate link mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to n network working nodes,Ω( n , r , s , t ) represents the time-varying mapping weight function, which is used to determine the r Is the link the n The suitable candidate links of a network working node in the network state s, S Represents a network status set, including normal and various fault states. S 0 means normal working state, Ω0( n , r , s ) represents the basis mapping weight, d represents the delay attenuation coefficient, Δ t r,n Indicates the n The network working node is r The propagation delay in the optical link, F s Indicates network status s The set of failed nodes or links, p n Represents a node pair n The corresponding combination of source and destination endpoints, L Indicates the n Other related nodes besides the network working nodes are transition nodes.

[0058] In step S34, if the protection capacity of the current optical link is less than the link fault recovery performance parameter, any network working node in the current optical link is used as the initial node, and another node is used as the target node. Based on the initial node, any network working node adjacent to the current optical link is selected as the transition node to reconstruct multiple virtual optical links, and the traffic ratio between the multiple virtual optical links and the current optical link is redistributed according to the preset ratio value.

[0059] In this step, the two network working nodes connected by the current optical link constitute the candidate set. Generally, these two nodes are considered equivalent, so one can be randomly selected or determined as the "initial node" or "source node" and the other as the "target node" based on specific strategies such as traffic demand, node importance, load conditions, etc. If each node has a preset candidate optical path selection index or nBased on these indicators, the party that is more conducive to building a virtual link can be selected as the initial node; the node with lower expected traffic load or more sensitive to fault recovery can be selected as the initial node to flexibly adjust the subsequent link construction plan; if there is a partition or hierarchical structure in the network, the node that is closer to the fault recovery path and connected to more virtual link candidate nodes may be preferentially selected as the initial node.

[0060] One or more transition nodes are added between the initial node and the target node to create a more redundant and scalable transmission path using multiple virtual links. Transition nodes should be selected from network working nodes that are physically or logically "proximate" to the initial node. This means the node must be a direct neighbor of the initial node or within an acceptable transmission distance and latency range. Considering the transmission quality of the link between the transition node and the initial node, such as transmission delay and bit error rate, select a link with high performance. The transition node should have sufficient capacity to establish subsequent links with the target node, and its interface resources and network status should be sufficient to form a high-quality virtual link. The location of the transition node within the overall network should help disperse risk. For example, using only direct links could be affected by regional failures. Selecting appropriately located nodes can provide multi-path backup. The basic concept is to segment the original link, constructing at least two virtual links: a virtual link from the initial node to the transition node, and a virtual link from the transition node to the target node. Each virtual link needs to meet basic performance requirements individually, including parameters such as bandwidth, latency, and jitter. More than one virtual link alternative solution can be constructed, and recovery reliability and load balancing capabilities can be further improved through parallel links or redundant backups.

[0061] In one example, one of the two active network nodes currently at both ends of the optical link is selected as the initial node, node A, and the other as the target node, node A. While this selection is typically based on node stability, device health, or current traffic load, it can be arbitrarily selected in this step according to a pre-set strategy.

[0062] Based on the initial node A, a node that is physically close to A and in good condition is selected from all working nodes near the current optical link as the "transition node" C. The selection process can be based on network status information provided by the monitoring platform, such as current link margin, historical fluctuations, signal-to-noise ratio, and other key parameters, to ensure that the transition node has the ability to handle additional traffic and provide good transmission quality.

[0063] Using initial node A, transition node C, and target node B, multiple virtual optical links are reconstructed. Two links are constructed: one from A to C and the other from C to B. These links are combined into a complete virtual optical link. In addition to the basic path described above, multiple candidate transition nodes can be selected, such as nodes C and D, to form multiple virtual links in a preset ratio. This improves fault tolerance through parallel load balancing. After determining the virtual optical link, all traffic on the current optical link is redistributed to the virtual optical link according to a preset ratio to reduce the risk of a single optical link. Based on network planning, empirical data, or real-time monitoring results, the traffic is redistributed in a ratio of 60% to continue on the current link and 40% to the virtual link. A more refined ratio can also be used to adaptively distribute traffic across virtual links. Using the traffic scheduling module and dynamic routing algorithm, all or part of the traffic from the initial node to the target node is forwarded over the newly constructed virtual optical link. During this process, lossless data transmission and latency spikes must be ensured. If necessary, buffering and lateral monitoring technologies can be used to achieve smooth migration. After traffic redistribution, the bandwidth utilization of each virtual optical link and the remaining physical optical links is monitored to ensure that all links meet the specified maximum bandwidth utilization requirements. If the utilization of the virtual link is found to be gradually increasing, the allocation ratio can be further fine-tuned or other candidate links can be considered for parallel load sharing.

[0064] Furthermore, assuming that the current protection capacity of the A-B optical link is lower than the performance parameter, the protection capacity is insufficient to meet the fault recovery requirements. Therefore, the system operates as follows: Define A as the initial node and B as the target node; find a node C or multiple nodes such as C and D in good condition near A as transition nodes; construct a virtual link A→C→B or construct multiple links: A→C→B and A→D→B. During the construction process, set the bandwidth of each link to 100 Gbps, and control the real-time utilization of the current virtual link to, for example, below 70%; according to a preset allocation ratio, such as 60% for the main link and 40% for the virtual link, assume that the total traffic on the current A-B link is approximately 32 Gbps of the 80 Gbps and divert it to the A→C→B virtual link, thereby reducing the load on the current physical link; after the traffic switching is completed, monitor the status of all links to ensure that each link operates within the expected range, and adjust the traffic ratio between the virtual link and the physical link in real time according to future traffic changes.

[0065] Step S35: If the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter, determine whether the protection capacity of the current optical link is greater than or equal to the wavelength channel load parameter based on the protection capacity of the current optical link and the wavelength channel protection constraint condition.

[0066] In this step, the wavelength channel protection constraint condition is expressed as: ; ; ; in, l r Indicates the r The protection capacity of the optical link, t represents the time synchronization factor, or n Indicates the n The candidate light path selection index corresponding to each network working node, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under l Indicates the l wavelength channels, m b represents the protection recovery coefficient, Indicates network status s You can use the l A collection of optical links with wavelength channels, Represents the wavelength channel mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to l wavelength channels, J ( t ) means in t Time jitter indicator at each moment, Q represents the S-type function based on the time synchronization quality. Parameters a and b control the steepness and displacement of the S-type function based on the time synchronization quality, respectively.

[0067] Step S36: If the protection capacity of the current optical link is less than the wavelength channel load parameter, multiple wavelength channels are expanded in the current optical link, and the flow ratios of all wavelength channels on the current optical link are redistributed.

[0068] In this step, based on the current wavelength channel mapping decision variables and the measured protection capacity status, the current transmission capacity of a single wavelength channel is compared with the target load index to determine the number of wavelength channels that need to be added. This process can refer to the preset protection recovery coefficient m b and time jitter indicators J ( t ), thereby ensuring that the network status sUnder this scenario, the expanded set of wavelength channels can meet overall traffic and protection requirements. Using optical amplifiers, wavelength division multiplexers (WDMs), and other devices, the spectrum resources in the existing optical link are divided to enable parallel transmission of multiple wavelength channels. The wavelength channel mapping decision variables are synchronously updated to map the new wavelength channels to the existing optical link, ensuring that the network management system can identify and schedule these newly expanded wavelength channels in real time.

[0069] After adding multiple wavelength channels, a traffic redistribution process needs to be executed to ensure that the traffic on each channel is evenly distributed and to maintain the stability and efficiency of the overall network. Statistics are collected on all data traffic currently transmitted on the optical link to obtain the total traffic value to be allocated, while considering the usage status of each wavelength channel in actual operation. A preset traffic distribution ratio is designed based on the actual transmission capacity and protection capacity of each wavelength channel before and after expansion. For example, when there was only one wavelength channel originally, all traffic may be carried by this channel; after expansion, each wavelength channel is set to share the overall traffic in a certain proportion, so that the traffic utilization rate of each channel is lower than the upper limit of the overall protection capacity. The ratio can be based on the bandwidth capacity of each channel in Kbps or Gbps or the current wavelength channel protection index. d r,w and candidate link mapping.

[0070] In an example, assume that the total capacity of the physical optical link from node A to node B is 120 Gbps. Currently, only one wavelength channel is enabled on the link, carrying 95 Gbps of traffic.

[0071] The constraint is that the protection capacity of a single wavelength channel is 60 Gbps, and the current load on the physical optical link is lower than the safety load required to share the current 95 Gbps of traffic. The wavelength channel protection constraint requires that the traffic on a single channel remain within a certain safety threshold. Assuming this threshold is 50% of the rated channel capacity, the safety load cap for 60 Gbps is 60 Gbps x 0.8 = 48 Gbps. Assuming an 80% load safety margin, this is sufficient. In this scenario, a single wavelength channel cannot guarantee safe transmission, so additional wavelength channels are needed.

[0072] Currently, a single channel carries 95 Gbps, far exceeding the single-channel safety threshold of 48 Gbps. Therefore, more wavelength channels are needed to distribute the traffic. Using optical amplifiers and wavelength division multiplexing (WDM) technology, two additional wavelength channels are added to the same physical optical link, resulting in a total of three wavelength channels available for parallel transmission. The network management system updates the wavelength channel mapping decision variable, marking all three channels as available for carrying traffic between A and B, and monitors their transmission quality indicators in real time.

[0073] The entire optical link currently transmits a load of 95 Gbps. To reduce the load on each channel, the system devised a new traffic distribution ratio, evenly distributing the 95 Gbps traffic across the three wavelength channels. To ensure a safe load on each channel, the system can employ an equal distribution ratio, with each channel receiving 95 / 3 (approximately 31.7 Gbps). Alternatively, the system can employ a preset ratio based on each channel's physical parameters, such as slightly different transmission quality or latency. For example, channel 1 would receive 40% of the traffic, or 38 Gbps; channel 2 would receive 30% of the traffic, or 28.5 Gbps; and channel 3 would receive 30% of the traffic, or 28.5 Gbps. Both solutions meet the requirement of keeping each channel below the safety threshold of 48 Gbps and distribute the previously overloaded 95 Gbps traffic as evenly as possible.

[0074] The network scheduling module redirects all traffic to the three wavelength channels according to a preset ratio, ensuring smooth data migration without packet loss or significant latency. It also monitors each channel's actual bandwidth utilization and protection capacity in real time. If monitoring indicates that a channel's utilization is approaching its safety limit, the system initiates a dynamic adjustment algorithm to further fine-tune the allocation ratio. For example, it can appropriately reduce the traffic allocation to the more heavily loaded channel and increase the allocation to other channels accordingly, until all channels stabilize within the safe operating range.

[0075] In this embodiment, the expressions of the bandwidth utilization constraint condition, the link protection restoration constraint condition, and the wavelength channel protection constraint condition are respectively: ; ; ; in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, N Represents a collection of network work nodes, C total represents the protection capacity of all optical links, β represents the synchronous correction coefficient, i max Indicates the maximum allowed bandwidth utilization, l r Indicates the r The protection capacity of the optical link, c represents the protection capacity adjustment factor based on time synchronization quality, t represents the time synchronization factor, or nIndicates the n The candidate light path selection index corresponding to each network working node, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Represents the candidate link mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to n Network working nodes, l Indicates the l wavelength channels, m b represents the protection recovery coefficient, Indicates network status s Next you can use l A collection of optical links with wavelength channels, Represents the wavelength channel mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to l wavelength channels, J ( t ) means in t Time jitter indicator at the moment.

[0076] In some examples, when the communication optical link is in a normal state, the network working node's requirements for the optical link meet the following conditions: ; Among them, Δ T n Indicates the n The time synchronization deviation value of each network working node, e Indicates the maximum allowed synchronization error ratio, T frame Indicates the SDH / PDH frame period, for example, 125μs for SDH. N Represents a collection of network work sections.

[0077] The SDH / PDH capacity granularity constraint is expressed as: ; in, represents the initial calculated payload, C vc Represents the minimum imaginary container unit size, and ⌈⌉ represents the rounding-up function.

[0078] ; in, ( t ) means int Time r The link in n Real-time traffic on network working nodes, Δ t Indicates the bandwidth measurement period, V max Indicates the preset maximum bandwidth change rate.

[0079] Furthermore, when the bandwidth utilization of an optical link between any two adjacent network nodes exceeds the maximum allowable value, the system immediately identifies that the optical link is in a faulty state. By selecting the network status data closest to the currently faulty link, it automatically generates a virtual optical link and rapidly switches all traffic on the original link to the virtual link, ensuring that traffic distribution on the virtual link remains within the allowed bandwidth, thus ensuring uninterrupted network transmission. If the link's normal bandwidth utilization is less than or equal to the maximum allowable value, the solution continues to determine the protection capacity of the current link to ensure that link recovery performance requirements are met. If the protection capacity is insufficient, multiple virtual optical links are reconstructed using a collaborative approach involving the initial node, target node, and transition node. Traffic is precisely distributed according to preset ratios, achieving dynamic load balancing and improving disaster recovery capabilities. When the link protection capacity meets the recovery requirements, the system further determines whether the protection capacity of the current link matches the wavelength channel load parameters based on the wavelength channel protection constraints. If the protection capacity is insufficient to support the existing wavelength channel load, the system expands multiple wavelength channels on the current link and redistributes all traffic, thereby improving the overall transmission reliability and service quality of the optical link. By analyzing bandwidth utilization, protection capacity, and wavelength channel load parameters in real time, the system automatically selects the most appropriate scheduling strategy under varying network conditions, achieving optimal global resource allocation. This automatic adjustment mechanism, based on preset constraints and dynamic estimates, ensures efficient and stable service transmission even in the face of sudden traffic surges or localized link failures, significantly enhancing the overall network's intelligence and anti-interference capabilities.

[0080] Step S4: obtaining a comprehensive evaluation index of the current data frame according to the network transmission resource parameters and the load bandwidth performance function, and adjusting the traffic distribution path of the optical transmission network according to the comprehensive evaluation index.

[0081] This step also includes steps S41 to S42.

[0082] Step S41 : identifying resource bottleneck events on each optical link in the optical transmission network according to the comprehensive evaluation index.

[0083] In this step, a set of original data frames is obtained from each network working node. After decoding, key indicators such as bandwidth utilization, delay, delay jitter, bit error rate, signal-to-noise ratio, and protection capacity utilization of each link are extracted. The original data frames are preprocessed, the average payload of each frame is calculated, and the payload of the current data frame is dynamically estimated using historical data. A set of preset weights is set, and each single indicator such as bandwidth utilization, delay, bit error rate, and protection capacity utilization is normalized and integrated into an evaluation score. For example, the comprehensive index can be calculated according to the following formula: Comprehensive evaluation index = w 1× bandwidth utilization + w 2× Delay+ w 3×bit error rate+ w 4× protection capacity utilization +…

[0084] A safety interval or threshold can also be pre-set. When the comprehensive metric exceeds or falls below this critical value, it indicates a link anomaly or potential bottleneck. For example, if bandwidth utilization exceeds 90%, latency increases, bit error rates surge, and protection capacity utilization remains high, this could indicate a link bottleneck. Historical data can be used as a reference for comparison to determine whether the current comprehensive evaluation metric is a short-term fluctuation or a long-term, sustained violation of the safety range. Abnormal conditions over multiple consecutive sampling periods are identified as resource bottleneck events.

[0085] Furthermore, if the comprehensive evaluation indicators of a particular optical link exceed a set threshold, for example due to factors such as insufficient bandwidth utilization, latency, and abnormal bit error rates, the link is determined to have experienced a resource bottleneck. The identified bottleneck event is recorded and prompts for subsequent targeted optimization steps to ensure overall network transmission quality and fault tolerance.

[0086] Here are some common resource bottleneck events: High bandwidth utilization bottleneck event: The bandwidth utilization of an optical link between A and B is normally around 70%. However, due to a traffic burst or a surge in certain services, the link utilization suddenly reaches 95%. The comprehensive indicator shows that when bandwidth utilization exceeds a preset threshold, such as 90%, coupled with increases in latency and bit error rate, the overall evaluation score rises significantly. This long-term overload of the link not only affects transmission quality but also may lead to insufficient protection capacity, thus constituting a resource bottleneck event.

[0087] Abnormal latency and jitter bottleneck events: On certain links, latency and jitter can suddenly increase due to network congestion or device scheduling issues. Comprehensive metrics indicate that latency and jitter data exceed preset tolerances, such as latency exceeding a certain millisecond threshold or excessive jitter fluctuations. These are detected as abnormal values ​​by the overall evaluation metrics. High and unstable latency can lead to packet aggregation and frequent retransmissions. In severe cases, this can impact the real-time delivery of critical applications, creating a resource bottleneck.

[0088] Underutilized protection capacity bottleneck event: A link has a large physical capacity, but its utilization rate remains at or above the design threshold for a long period of time, for example, reaching or exceeding 0%. Comprehensive indicators indicate a risk of insufficient protection resources when real-time monitoring shows that the utilization of protection capacity is approaching the upper limit while actual link traffic continues to increase. If a failure occurs, backup resources will be unable to meet recovery requirements, resulting in a decrease in link recovery capabilities, thus identifying a resource bottleneck event.

[0089] Uneven wavelength channel load bottlenecks: In a multi-wavelength channel link, some wavelength channels may experience significantly higher loads than others due to scheduling policies or environmental parameters. This manifests as individual wavelength channel utilization significantly exceeding the overall average and exceeding the safe load threshold, accompanied by abnormal fluctuations in bit error rate and latency. This degrades the overall link's fault recovery performance, creating hotspot bottlenecks. In severe cases, these bottlenecks may trigger subsequent emergency measures such as dynamic routing and traffic scheduling.

[0090] Step S42: The network status information corresponding to the resource bottleneck event and the historical network status information are combined with a dynamic routing algorithm to recalculate the optimal path, and the traffic distribution path is scheduled in real time.

[0091] In this step, the link status information of the current resource bottleneck is extracted from step S41, including real-time bandwidth utilization, delay, bit error rate, protection capacity utilization, etc. Combined with historical network status data such as historical traffic fluctuations, fault recovery status, link stability data, etc., the cause and persistence of the current bottleneck are analyzed to assist in determining the changing trend of the actual load and transmission quality. Dynamic weights are assigned to each link based on real-time indicators and historical statistical data. For example, links with bandwidth utilization exceeding the safety threshold, abnormal delays and bit error rates will be given higher weights, which will make these links "costly" higher in path calculation, and thus be avoided by the dynamic routing algorithm. At the same time, lower weights can be set for stable links with sufficient backup to encourage traffic scheduling to these links.

[0092] Typical dynamic routing algorithms, such as those based on Dijkstra and A*, incorporate real-time metrics into the weight matrix to recalculate paths across the entire optical transmission network, generating one or more optimal paths from the source node to the destination node. Where appropriate, the system generates multiple alternative paths, utilizing load balancing or redundant routing techniques to improve the network's fault tolerance. Once the new route is determined, the network scheduling module gradually migrates traffic previously traversing the bottleneck link to the newly calculated optimal path according to pre-defined traffic switching strategies.

[0093] In this embodiment, the expression of the load-bandwidth performance function is: ; in, maxY ( t ) represents the optical link network under different constraints. t The load-bandwidth performance function at time , K Represents the set of all data frames that need to be transmitted in the optical link network, TD ( W k ) indicates the k The transmission cost corresponding to the frame bandwidth allocation, W k Indicates the k Bandwidth allocation of frames, H k Indicates the k The sampling period of the frame, t represents the time synchronization factor, d r,w Indicates the r Links under specific environmental conditions oh The transmission quality under Z Indicates the preset traffic scaling factor in the optical link network. ( t ) means in t Time r The link in n Real-time traffic on network worker nodes, Indicates the n Network working nodes are in network state s The set of candidate optical routing ends under D n ( t ) means in t Time n The estimated value of the payload of each network worker node.

Claims

1. A method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization, characterized in that: The following steps are involved: Acquire network status information of each network working node in the optical transmission network, wherein the network status information includes a set of original data frames; Preprocessing the original data frame set to obtain an average payload corresponding to each data frame in the data frame set, and calculating a dynamic payload estimate based on the average payload and a historical average payload; According to preset constraints and a dynamic estimate of the payload, a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes and network transmission resource parameters corresponding to the network transmission resource allocation strategy are obtained; According to the network transmission resource parameters and the load bandwidth performance function, a comprehensive evaluation index of the current data frame is obtained, and the traffic distribution path of the optical transmission network is adjusted according to the comprehensive evaluation index.

2. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 1, characterized in that: The data frame set is preprocessed to obtain an average payload corresponding to each data frame in the data frame set, specifically including: Eliminate outliers from the data frame set to obtain a transition data frame set, and align timestamps of data frames from different network working nodes in the transition data frame set based on a synchronous clock to obtain payload data corresponding to each data frame in the transition data frame set; According to the sampling period of the optical transmission network, the continuous data frames in the transition data frame set are divided into a unified time window, and the payload data corresponding to the current data frame is weightedly fused with the historical payload data in the previous time window to obtain the average payload corresponding to each data frame.

3. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 1, characterized in that: The obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload specifically includes: Determining whether a bandwidth utilization rate of an optical link between any two adjacent network working nodes is less than or equal to a maximum allowable bandwidth utilization rate based on a bandwidth utilization rate constraint condition in the preset constraint condition and the dynamic estimated value of the payload; If the bandwidth utilization of the optical link between any two adjacent network working nodes is greater than the maximum allowed bandwidth utilization, the current optical link is determined to be in a fault state, and the network status information closest to the current optical link is selected to regenerate a virtual optical link, and all traffic of the current optical link is allocated to the virtual optical link, and the bandwidth utilization of the virtual optical link is less than or equal to the maximum allowed bandwidth utilization.

4. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 3, characterized in that: The obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload also includes: If the bandwidth utilization of the optical link between any two adjacent network working nodes is less than or equal to the maximum allowed bandwidth utilization, then based on the protection capacity of the current optical link and the link protection recovery constraint, determine whether the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter; If the protection capacity of the current optical link is less than the link fault recovery performance parameter, any network working node in the current optical link is used as the initial node, and another node is used as the target node. Based on the initial node, any network working node adjacent to the current optical link is selected as the transition node to reconstruct multiple virtual optical links, and the traffic ratio between the multiple virtual optical links and the current optical link is redistributed according to the preset ratio value.

5. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 4, characterized in that: The obtaining of a network transmission resource allocation strategy for an optical link between any two adjacent network working nodes based on preset constraints and a dynamic estimate of the payload also includes: If the protection capacity of the current optical link is greater than or equal to the link fault recovery performance parameter, then judging whether the protection capacity of the current optical link is greater than or equal to the wavelength channel load parameter based on the protection capacity of the current optical link and the wavelength channel protection constraint condition; If the protection capacity of the current optical link is less than the wavelength channel load parameter, multiple wavelength channels are expanded on the current optical link, and the traffic ratios of all wavelength channels on the current optical link are redistributed.

6. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 5, characterized in that: The expressions of the bandwidth utilization constraint condition, link protection and restoration constraint condition, and wavelength channel protection constraint condition are respectively: ； ； ； in, D n ( t ) means in t Time n The estimated effective load of the network working nodes, N Represents a collection of network work nodes, C total represents the protection capacity of all optical links, β represents the synchronous correction coefficient, θ max Indicates the maximum allowed bandwidth utilization, λ r Indicates the r The protection capacity of the optical link, γ represents the protection capacity adjustment factor based on time synchronization quality, τ represents the time synchronization factor, η n Indicates the n The candidate light path selection index corresponding to each network working node, δ r,w Indicates the r Links under specific environmental conditions ω The transmission quality under Represents the candidate link mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to n Network working nodes, l Indicates the l wavelength channels, μ l represents the protection recovery coefficient, Indicates network status s Next you can use l A collection of optical links with wavelength channels, Represents the wavelength channel mapping decision variable, that is, in the network state s Next, select whether to r The optical link is mapped to l wavelength channels, J ( t ) means in t Time jitter indicator at each moment, Indicates the n Network working nodes are in network state s The candidate optical routing end set under .

7. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 1, characterized in that: Adjusting the traffic distribution path of the optical transmission network according to the comprehensive evaluation index specifically includes: Identifying resource bottleneck events on each optical link in the optical transmission network according to comprehensive evaluation indicators; The network status information corresponding to the resource bottleneck event and the historical network status information are used to recalculate the optimal path using a dynamic routing algorithm, and the traffic distribution path is scheduled in real time.

8. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 1, characterized in that: The expression of the load bandwidth performance function is: ； in, maxY ( t ) represents the optical link network under different constraints. t The load-bandwidth performance function at time t, K Represents the set of all data frames that need to be transmitted in the optical link network, TD ( W k ) indicates the k The transmission cost corresponding to the frame bandwidth allocation, W k Indicates the k Bandwidth allocation of frames, H k Indicates the k The sampling period of the frame, τ represents the time synchronization factor, δ r,w Indicates the r Links under specific environmental conditions ω The transmission quality under Z Indicates the preset traffic scaling factor in the optical link network. ( t ) means in t Time r The link in n Real-time traffic on network worker nodes, Indicates the n Network working nodes are in network state s The set of candidate optical routing ends under D n ( t ) means in t Time n The estimated payload of each network worker node.

9. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 6, characterized in that: The network transmission resource parameters include one or more of bandwidth utilization, link failure recovery performance parameters, and wavelength channel load parameters.

10. The method for real-time bandwidth analysis of SDH / PDH payload based on time synchronization according to claim 1, characterized in that: The optical transmission network includes one or more of Synchronous Digital Hierarchy (SDH) and Plesiochronous Digital Hierarchy (PDH).

Images