A signal analysis-based communication equipment state monitoring system and method

Abstract

The application discloses a communication equipment state monitoring system and method based on signal analysis, and belongs to the technical field of equipment state monitoring. The application collects cross-cluster communication link signals, establishes a normal baseline with historical data, compares real-time signals to screen abnormal signals beyond the service range, extracts information to form an abnormal data set; obtains a cluster topology, locates abnormal switch nodes according to the abnormal data set, traces back physical and logical associated equipment in both directions to form an association set; splits the association set into subgroups according to service attributes and other rules, constructs a sub-topology for each subgroup, integrates the sub-topology into an associated topology set, calculates network distance and extracts a simplified equipment subset; plans connection relationship with the maximum communication efficiency and signal standard as the target, adjusts port configuration, enables a standby link to optimize parameters if the main link is substandard, integrates and optimizes a sub-topology into a new topology set, and updates a cluster equipment list.

Description

Technical Field

[0001] This invention relates to the field of equipment condition monitoring technology, specifically a communication equipment condition monitoring system and method based on signal analysis. Background Technology

[0002] With the integrated development of 5G and IoT technologies, communication networks are exhibiting a trend of large-scale deployment across multiple clusters. This leads to intertwined links and complex device relationships between clusters, resulting in a surge in demands for operational stability and communication efficiency. Monitoring the status of communication equipment is crucial for ensuring reliable network operation. This requires signal analysis to identify anomalies, locate faults, and optimize the topology. However, traditional operation and maintenance models are no longer adequate for complex network scenarios, necessitating intelligent monitoring and optimization solutions.

[0003] In existing technologies, only a single abnormal device is located, without tracing related devices bidirectionally along the physical link and logical topology, which easily leads to the omission of fault propagation sources or related hidden dangers; the overall processing of related devices does not involve grouping and refining according to coupling degree, and does not streamline invalid devices, resulting in low optimization efficiency and waste of resources; the response to switching to the backup link after the main link fails is slow, and there is a lack of coordinated optimization of signal parameters and connection relationships, making it difficult to balance efficiency and signal quality. Summary of the Invention

[0004] The purpose of this invention is to provide a communication device status monitoring system and method based on signal analysis to solve the problems raised in the prior art.

[0005] To achieve the above objectives, the present invention provides the following technical solution: In a first aspect, this application provides a communication device status monitoring method based on signal analysis, comprising the following steps: Collect signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; compare the collected signals with the baseline to filter out abnormal signals that deviate from the baseline and exceed the business allowable range; extract the information corresponding to the abnormal signals to form an abnormal dataset; Obtain the topology of each cluster, locate the specific node location of each abnormal switch based on the abnormal dataset; take the abnormal switch node as the core, trace the associated devices bidirectionally along the physical link and logical topology to form a set of associated communication devices; Set grouping rules to split each set of associated communication devices into multiple subgroups; for each subgroup, construct an independent subtopology and integrate them to form an associated topology structure set; for each subtopology, calculate the network distance between device nodes, extract the device nodes that are closest to each other and have actual communication needs, and form a simplified device subset; With the goal of maximizing communication efficiency and meeting signal strength standards, the connection relationships were replanned to streamline the device subset; the device port configuration was adjusted, and when the main link signal strength did not meet the threshold required by the service, the backup link was activated and the signal parameters were optimized; all optimized sub-topologies were integrated to form a new topology set, and the cluster device list was updated.

[0006] In conjunction with the first aspect, in a first embodiment of the first aspect of this application, the step of collecting signals from communication links between different clusters and establishing a normal signal baseline based on historical signal data includes: Obtain a list of all cross-cluster communication links in the data center, label each link with a unique identifier, associate the service attributes carried by the link, and collect the signals of communication links between different clusters; deploy signal acquisition nodes to the switches at both ends of the cross-cluster links, and collect the native signal data of the links through the device's native interface; Historical signal data of cross-cluster communication links in the data center during normal operation cycles are retrieved, and the data is cleaned. Statistical feature values ​​are calculated for each link, each signal dimension, and each business scenario. The normal signal baseline includes the steady-state signal baseline and the time-series fluctuation signal baseline. The normal fluctuation range is defined by using several times the standard deviation of the mean as the baseline interval to form the steady-state signal baseline. The time-series baseline curve is fitted to mark the normal fluctuation patterns in different time periods to form the time-series fluctuation signal baseline.

[0007] In conjunction with the first aspect, in a second embodiment of the first aspect of this application, the step of comparing the collected signals with a baseline and filtering out abnormal signals that deviate from the baseline and exceed the permitted range for the service includes: For steady-state signal baselines, the comparison rule is whether the real-time value exceeds the baseline range, while also adding the business-permissible range threshold. For time-series fluctuating signal baselines, the comparison rule is whether the real-time value deviates from the fluctuation range of the time-series baseline curve, and whether the deviation exceeds the business-permissible deviation. The baseline parameters corresponding to the current link, current signal dimension, and current business scenario are retrieved, the deviation between the real-time signal value and the baseline is calculated, and it is determined whether the deviation exceeds the business-permissible range threshold. The comparison results are recorded. Based on the comparison results, abnormal signals that deviate from the baseline and exceed the business-permissible range are filtered out.

[0008] In conjunction with the first aspect, in the third embodiment of the first aspect of this application, the step of obtaining the topology of each cluster and locating the specific node location of each abnormal switch based on the abnormal dataset includes: Topology data of all clusters is retrieved from both physical and logical topology dimensions, and the topology data is standardized to obtain standardized topology data. From the abnormal dataset, extract all relevant switch identification fields to form an abnormal switch identification list. Match the abnormal switch identification list with the cluster labels in the standardized topology data to determine the target cluster to which the abnormal switch belongs. In the physical topology of the target cluster, locate the specific physical node location of the abnormal switch based on its globally unique identifier. From the hierarchical labels in the topology data, clarify the network layer of the abnormal switch in the cluster and label the core functions of this layer. In the physical topology, find devices and links that have a direct physical connection to the abnormal switch and whose data flow is input to the abnormal switch, and record the upstream device identifier, connection port, link type, and bandwidth. Find devices and links that have a direct physical connection to the abnormal switch and whose data flow is output from the abnormal switch, and record the downstream device identifier, connection port, link type, and bandwidth. When the abnormal switch involves cross-cluster links, separately label the peer switch identifier, the cluster to which it belongs, and the logical attributes of the link. Based on the logical configuration in the topology data, supplement the logical upstream and downstream relationships of the abnormal switch, associate the communication service flow corresponding to the abnormal signal, locate the forwarding path of the service flow on the abnormal switch, and clarify the upstream and downstream devices from the perspective of the service flow.

[0009] In conjunction with the first aspect, in the fourth embodiment of the first aspect of this application, the step of tracing related devices bidirectionally along physical links and logical topologies, with the abnormal switch node as the core, to form a set of related communication devices, includes: Forward tracing is performed on the physical link, starting from all service ports of the abnormal switch, matching the physical connection relationship of the port in the topology data, and extracting the directly connected downstream devices; for the directly traced downstream devices, their port connection relationship is further retrieved, and the downstream devices of the device are traced until the set tracing depth is reached; reverse tracing is performed on the physical link, based on the uplink port connection relationship of the abnormal switch, tracing the upstream devices that directly provide signal or data input to it, and for the directly upstream devices, continuing to trace their upstream devices, until the cluster core node or the peer cluster core device across the cluster link is traced. In the logical topology, perform forward tracing to extract the VLAN configuration, broadcast domain, and ACL rules of the abnormal switch, and trace all devices within the same VLAN or broadcast domain; based on the abnormal switch's routing table, ARP table, and traffic logs, locate the service flow forwarding path corresponding to the abnormal signal, and trace the downstream forwarding devices on that path; in the logical topology, perform reverse tracing to retrieve the abnormal switch's dynamic routing protocol neighbor table, link aggregation group member information, and SNMP management node information, and trace protocol neighbor devices, aggregation group member devices, and management node devices.

[0010] In conjunction with the first aspect, in the fifth embodiment of the first aspect of this application, the step of setting grouping rules to split each set of associated communication devices into multiple subgroups includes: Establish first-level grouping rules: devices with the same service type and priority, belonging to the same cluster, are grouped into the same initial group; devices with the same service type but different priorities, or devices with similar service types within the same cluster, are grouped into the same initial group. Establish second-level grouping rules: within the same initial group, the direct upstream and downstream devices of the first-level associated devices and the abnormal switch are grouped into the core subgroup; second-level associated devices are split into secondary subgroups based on logical dependencies and the same forwarding path. Within the same subgroup, when devices are physically deployed across multiple racks and have no direct communication requirements, they are split into different subgroups based on proximity. Calculate business coupling ,in, For the first subgroup With the second subgroup Average daily business interaction traffic between them for The average daily total communication traffic, for Average daily total communication traffic; calculation logic configuration correlation. Where p is the total number of logical configuration types. Configure associated weights for the m-th type of logic. for and The m-th logical configuration association identifier, This indicates the existence of the m-th type of association. Indicates no correlation; calculate the overall coupling degree. , This is a weighting coefficient for the degree of business coupling. Configure the weight coefficients for logical relevance; When the combined coupling of two subgroups exceeds a set threshold, highly interactive devices between the two groups are extracted. If the interaction is cross-group, the highly interactive devices are split to form a new subgroup.

[0011] In conjunction with the first aspect, in the sixth embodiment of the first aspect of this application, the step of constructing an independent sub-topology for each subgroup and integrating them to form an associated topology structure set includes: Sub-topology includes sub-physical topology graph and sub-logical topology graph; a sub-physical topology graph is constructed with device nodes as vertices and physical links as edges, and logical configuration information is superimposed on the physical topology layer to form a sub-logical topology graph; information of abnormal switches in subgroups of abnormal data is associated with sub-topology.

[0012] In conjunction with the first aspect, in the seventh embodiment of the first aspect of this application, the re-planning of connection relationships for a simplified subset of devices with the goal of maximizing communication efficiency and achieving signal strength targets includes: Based on historical communication logs, bidirectional service flow data for each device is extracted, and the core parameters of each service flow are labeled. Based on a simplified device subset, a binary communication relationship matrix is ​​constructed, and the matrix elements are labeled with weights according to the service flow priority. According to the principle of matching high-priority service flows with high-specification links first, service flows are initially associated with reusable link resources, and the matching results are recorded. Based on the physical deployment coordinates of the devices, the correlation between physical distance and link attenuation between devices is calculated. The closer the distance and the smaller the attenuation, the higher the priority. Devices with a weight greater than a set threshold in the communication relationship matrix are given priority for planning direct connection links. The shortest hop path is planned to avoid multi-node forwarding. Link aggregation groups are planned to ensure sufficient bandwidth. Multiple candidate schemes are generated. For each candidate scheme, the signal transmission process of each planned link is simulated. The actual signal parameters are calculated based on the link attenuation coefficient and the device port transmit power to determine whether the baseline threshold is met. When the signal parameters of a link are lower than the threshold, it is marked as a non-compliant link. For compliant links, the average transmission delay, average bandwidth utilization, and forwarding success rate are calculated, and a weighted calculation is used to obtain a comprehensive efficiency score. The scheme with a 100% signal compliance rate and the highest comprehensive efficiency score is selected as the optimal scheme, and the connection relationship is replanned.

[0013] In conjunction with the first aspect, in the eighth embodiment of the first aspect of this application, the adjustment of the device port configuration, when the primary link signal strength does not reach the threshold required by the service, to activate the backup link and optimize the signal parameters, includes: For links that fail to meet the standards, prioritize adjusting the port parameters of both ends of the equipment, including rate negotiation, transmit power, and noise reduction configuration. If the main link still fails to meet the standards after port adjustment, check the backup links within the sub-topology and trigger link switching through the equipment configuration interface to switch the service flow to the backup link. Adjust the port gain and equalizer parameters at both ends of the backup link to increase signal redundancy. Traverse all communication links between devices within the sub-topology to confirm that the signal values ​​all meet the threshold requirements.

[0014] Secondly, this application provides a communication device status monitoring system based on signal analysis, comprising: Anomaly signal acquisition module: includes: a cross-cluster communication signal acquisition unit to acquire signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; a normal signal baseline establishment unit to compare the acquired signals with the baseline and filter out abnormal signals that deviate from the baseline and exceed the business allowable range; and anomaly signal filtering unit to extract information corresponding to the abnormal signals and form an anomaly dataset. Cluster topology parsing module: includes: an abnormal switch location unit to obtain the topology of each cluster and locate the specific node location of each abnormal switch based on the abnormal dataset; and an associated device set tracing unit that uses the abnormal switch node as the core to trace associated devices bidirectionally along the physical link and logical topology to form an associated communication device set. The simplified device subset generation module includes: an associated device grouping unit that sets grouping rules and splits each associated communication device set into multiple subgroups; an associated topology set integration unit that constructs an independent subtopology for each subgroup and integrates them to form an associated topology structure set; and a simplified device subset extraction unit that calculates the network distance between device nodes for each subtopology, extracts the device nodes that are closest to each other and have actual communication needs, and forms a simplified device subset. Topology optimization and cluster reconstruction module: includes: a simplified subset connection relationship planning unit that re-plans the connection relationship for simplified device subsets with the goal of maximizing communication efficiency and meeting signal strength standards; a link signal optimization unit that adjusts the device port configuration and enables the backup link and optimizes the signal parameters when the main link signal strength does not meet the threshold required by the service; and a new topology integration and list update unit that integrates all optimized sub-topologies to form a new topology structure set and updates the cluster device list.

[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. After locating abnormal switches by combining cluster topology data, this invention traces related devices bidirectionally along physical links and logical topology, fully covering fault sources, transmission links, and hidden dangers; for example, when locating switches with abnormal latency, it can simultaneously lock the root cause of upstream optical module attenuation and downstream convergence layer switches with signal distortion, completely avoiding the omission of hidden dangers.

[0016] 2. This invention splits related devices into subgroups according to business attributes and coupling degree, ensuring strong communication dependence within the group and low coupling between groups; then, it calculates network distance to extract core communication nodes to form a simplified subset, eliminating redundant edge nodes, which can improve resource utilization.

[0017] 3. This invention presets primary and backup link thresholds. When the primary link signal fails to meet the standard, it automatically switches and simultaneously optimizes the backup link port power and noise reduction parameters, and adjusts the connection relationship to match the needs of simplified subset services. When the primary link breaks, the switching can improve forwarding efficiency, balancing efficiency and signal quality. Attached Figure Description

[0018] Figure 1 This is a schematic diagram illustrating the steps of a communication device status monitoring method based on signal analysis according to the present invention. Figure 2 This is a system structure diagram of a communication equipment status monitoring system based on signal analysis according to the present invention. Detailed Implementation

[0019] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0020] Example: Figures 1-2 As shown, the present invention provides a technical solution: like Figure 1 As shown, this application provides a communication device status monitoring method based on signal analysis, including the following steps: Step S100: Collect signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; compare the collected signals with the baseline and filter out abnormal signals that deviate from the baseline and exceed the permitted range for business; extract the information corresponding to the abnormal signals to form an abnormal dataset; Specifically, obtain a list of all cross-cluster communication links in the data center, mark each link with a unique identifier, associate the service attributes carried by the link, and collect the signals of communication links between different clusters; deploy signal acquisition nodes to the switches at both ends of the cross-cluster links, and collect the native signal data of the links through the device's native interface; Historical signal data of cross-cluster communication links in the data center during normal operation cycles are retrieved, and the data is cleaned. Statistical feature values ​​are calculated for each link, each signal dimension, and each business scenario. The normal signal baseline includes the steady-state signal baseline and the time-series fluctuation signal baseline. The normal fluctuation range is defined by using several times the standard deviation of the mean as the baseline interval to form the steady-state signal baseline. The time-series baseline curve is fitted to mark the normal fluctuation patterns in different time periods to form the time-series fluctuation signal baseline.

[0021] Furthermore, for the steady-state signal baseline, the comparison rule is whether the real-time value exceeds the baseline range, while also adding the service-permitted range threshold; for the time-series fluctuating signal baseline, the comparison rule is whether the real-time value deviates from the fluctuation range of the time-series baseline curve, and whether the deviation exceeds the service-permitted deviation; retrieve the baseline parameters corresponding to the current link, current signal dimension, and current service scenario, calculate the deviation between the real-time signal value and the baseline, determine whether the deviation exceeds the service-permitted range threshold, and record the comparison results; based on the comparison results, filter out abnormal signals that deviate from the baseline and exceed the service-permitted range.

[0022] In one specific embodiment, the cross-cluster communication link between the data center "core transaction cluster A - disaster recovery cluster B" is selected as the object. The link is uniquely identified by L001 and carries core securities trading services (the service allows optical power threshold of -6~1dBm and latency threshold of ≤15ms). The devices at both ends of the link are the core switch S1 of cluster A (port Gi1 / 0 / 24) and the core switch S2 of cluster B (port Gi2 / 0 / 18).

[0023] Signal acquisition: Deploy acquisition nodes on the corresponding ports of S1 and S2, and acquire two types of core signals through the SNMP interface: ① Steady-state signal - optical power (acquisition frequency 100ms / time); ② Timing fluctuation signal - transmission delay (50ms / time during peak 8:30-11:30, and 200ms / time during off-peak 11:30-15:00).

[0024] Baseline establishment: Retrieve 7 days of normal operation history data (no faults, stable business), clean the data, and calculate feature values: Steady-state optical power: mean -2dBm, standard deviation 0.5dBm, baseline range [-3dBm, -1dBm]; Timing delay: The average peak value of the fitted curve is 10ms, the average flat peak value is 8ms, and the allowable deviation of fluctuation is ±20%.

[0025] Real-time data acquisition (peak 9:00): Optical power -3.5dBm, latency 13.5ms.

[0026] Compare and determine: Optical power: -3.5dBm < baseline range lower limit -3dBm and < service threshold -6dBm upper limit, is considered abnormal; Latency: 13.5ms deviates from the peak average of 10ms by 35% > 20%, and is greater than the business threshold of 15ms upper limit, which is considered abnormal.

[0027] Abnormal dataset: Record L001, optical power / delay anomaly, occurrence time 9:00, S1 / S2 port, deviation 16.7% / 35%.

[0028] Step S200: Obtain the topology of each cluster, locate the specific node location of each abnormal switch based on the abnormal dataset; take the abnormal switch node as the core, trace the associated devices bidirectionally along the physical link and logical topology to form a set of associated communication devices; Specifically, topology data from all clusters is retrieved from both physical and logical topology dimensions, and the topology data is standardized to obtain standardized topology data. From the abnormal dataset, extract all relevant switch identification fields to form an abnormal switch identification list. Match the abnormal switch identification list with the cluster labels in the standardized topology data to determine the target cluster to which the abnormal switch belongs. In the physical topology of the target cluster, locate the specific physical node location of the abnormal switch based on its globally unique identifier. From the hierarchical labels in the topology data, clarify the network layer of the abnormal switch in the cluster and label the core functions of this layer. In the physical topology, find devices and links that have a direct physical connection to the abnormal switch and whose data flow is input to the abnormal switch, and record the upstream device identifier, connection port, link type, and bandwidth. Find devices and links that have a direct physical connection to the abnormal switch and whose data flow is output from the abnormal switch, and record the downstream device identifier, connection port, link type, and bandwidth. When the abnormal switch involves cross-cluster links, separately label the peer switch identifier, the cluster to which it belongs, and the logical attributes of the link. Based on the logical configuration in the topology data, supplement the logical upstream and downstream relationships of the abnormal switch, associate the communication service flow corresponding to the abnormal signal, locate the forwarding path of the service flow on the abnormal switch, and clarify the upstream and downstream devices from the perspective of the service flow.

[0029] Furthermore, forward tracing is performed in the physical link, starting from all service ports of the abnormal switch, matching the physical connection relationship of the port in the topology data, and extracting the directly connected downstream devices; for the directly traced downstream devices, their port connection relationship is retrieved, and the downstream devices of the device are traced until the set tracing depth is reached; reverse tracing is performed in the physical link, based on the uplink port connection relationship of the abnormal switch, tracing the upstream devices that directly provide signal or data input to it, and for the directly upstream devices, continuing to trace their upstream devices, until the cluster core node or the peer cluster core device across the cluster link is traced. In the logical topology, perform forward tracing to extract the VLAN configuration, broadcast domain, and ACL rules of the abnormal switch, and trace all devices within the same VLAN or broadcast domain; based on the abnormal switch's routing table, ARP table, and traffic logs, locate the service flow forwarding path corresponding to the abnormal signal, and trace the downstream forwarding devices on that path; in the logical topology, perform reverse tracing to retrieve the abnormal switch's dynamic routing protocol neighbor table, link aggregation group member information, and SNMP management node information, and trace protocol neighbor devices, aggregation group member devices, and management node devices.

[0030] In one specific embodiment, complete topology data of data center clusters A and B is retrieved. The physical topology dimension includes device models, rack deployment locations, and port connection relationships (e.g., S1-Gi1 / 0 / 24 and S2-Gi2 / 0 / 18 are directly connected via 100G fiber). The logical topology dimension is labeled with VLAN configurations (S1, S2, and the core business server all belong to VLAN 100), OSPF routing protocol neighbor relationships, and link aggregation group LAG1 (including S1-Gi1 / 0 / 24 and S1-Gi1 / 0 / 25). All devices and links are assigned globally unique identifiers to form standardized topology data.

[0031] Extract the abnormal switch identifiers S1 (Cluster A) and S2 (Cluster B) from the abnormal dataset. After matching them with the standardized topology data, locate the physical location of S1 as "Cluster A-Core Layer-Rack 3-Rack 2" and its core function is cross-cluster data forwarding; locate the physical location of S2 as "Cluster B-Core Layer-Rack 5-Rack 1" and its core function is disaster recovery data synchronization.

[0032] In terms of physical link tracing, the upstream devices of S1 are S2 of cluster B (cross-cluster link, link identifier L001, optical link bandwidth 100G, logical attribute primary) and cluster A aggregation layer switch S3 (connection port Gi1 / 0 / 23, electrical link bandwidth 25G), and the downstream devices are cluster A access layer switches S4 (Gi1 / 0 / 1) and S5 (Gi1 / 0 / 2). According to the tracing depth of 2 layers, the downstream core business servers SRV1 and SRV2 of S4 are further traced. The upstream device of S2 is cluster B aggregation layer switch S6 (Gi2 / 0 / 20), and the downstream is only the cross-cluster link peer S1.

[0033] In terms of logical topology tracing, the VLAN100 configuration of S1 is extracted and traced back to S2, S3, S4, S5 and SRV1 and SRV2 in the same VLAN; based on the OSPF routing table, the forwarding path of the abnormal delay service flow is located as S2-Gi2 / 0 / 18→S1-Gi1 / 0 / 24→S4-Gi1 / 0 / 1→SRV1, and traced back to the OSPF neighbor devices S2 and S3 of S1 and the optical modules OM1 and OM2 corresponding to the member ports of the link aggregation group LAG1.

[0034] The final set of associated communication devices includes switches S1, S2, S3, S4, S5, and S6; core service servers SRV1 and SRV2; optical modules OM1 and OM2; link L001; and related components of aggregation group LAG1.

[0035] Step S300: Set grouping rules to split each set of associated communication devices into multiple subgroups; for each subgroup, construct an independent subtopology and integrate them to form an associated topology structure set; for each subtopology, calculate the network distance between device nodes, extract the device nodes that are closest to each other and have actual communication needs, and form a simplified device subset; Specifically, first-level grouping rules are established. Devices with the same service type and priority, belonging to the same cluster, are grouped into the same initial group. Second-level grouping rules are established. Within the same initial group, devices with the same service type but different priorities, or devices with similar service types within the same cluster, are grouped into the same initial group. Second-level grouping rules are established. Within the same initial group, the direct upstream and downstream devices of the first-level associated devices and the abnormal switch are grouped into the core subgroup. Second-level associated devices are split into secondary subgroups based on logical dependencies and the same forwarding path. Within the same subgroup, when devices are physically deployed across multiple racks and have no direct communication requirements, they are split into different subgroups based on proximity. Calculate business coupling ,in, For the first subgroup With the second subgroup Average daily business interaction traffic between them for The average daily total communication traffic, for Average daily total communication traffic; calculation logic configuration correlation. Where p is the total number of logical configuration types. Configure associated weights for the m-th type of logic. for and The m-th logical configuration association identifier, This indicates the existence of the m-th type of association. Indicates no correlation; calculate the overall coupling degree. , This is a weighting coefficient for the degree of business coupling. Configure the weight coefficients for logical relevance; When the combined coupling of two subgroups exceeds a set threshold, highly interactive devices between the two groups are extracted. If the interaction is cross-group, the highly interactive devices are split to form a new subgroup.

[0036] Furthermore, the sub-topology includes a sub-physical topology graph and a sub-logical topology graph; a sub-physical topology graph is constructed with device nodes as vertices and physical links as edges, and logical configuration information is superimposed on the physical topology layer to form a sub-logical topology graph; information of abnormal switches in the subgroup of the abnormal data is associated with the sub-topology.

[0037] In one specific embodiment, grouping is performed based on the associated communication device set (including switches S1 / S2 / S3 / S4 / S5 / S6, servers SRV1 / SRV2, and optical modules OM1 / OM2) from step S200. According to the first-level rule, devices belonging to cluster A and in core transaction services (S1, S3, S4, SRV1, SRV2, OM1, OM2) are assigned to the initial group G_initial, and S2 and S6 of cluster B are assigned to the initial group G_initial2. According to the second-level rule, the direct upstream and downstream of S1 in G_initial (S3, S4, OM1, OM2) form the core subgroup G1, the remaining SRV1, SRV2, and S5 form the secondary subgroup G2, and G_initial2 directly becomes subgroup G3.

[0038] Calculate the coupling degree between G1 and G2: Tij (two sets of interactive traffic) = 150Mbps, Ti (total traffic of G1) = 500Mbps, Tj (total traffic of G2) = 300Mbps, Ctraff = 150 / (500+300-150)×100%≈23.08%; Logical association type p = 2 (VLAN sharing, routing neighbor), weight w1=w2=0.5, L1=L2=1, Clogic = 0.5×1+0.5×1=1; Take α=0.6, β=0.4, Ctotal = 0.6×23.08%+0.4×1×100%≈53.85%, exceeding the threshold of 30%, extract the highly interactive SRV1, SRV2 and S4 to split into a new subgroup G4.

[0039] When constructing sub-topologies, the physical topology of G1 uses S1 / S3 / S4 / OM1 / OM2 as vertices and corresponding optical / electrical links as edges, labeling the physical location of S1 and "latency anomalies"; VLAN 100 and OSPF neighbor information are overlaid to form a logical topology. Network distances are calculated: S1 and S3 / S4 / OM1 are both 1-hop, and SRV1 and S4 are both 1-hop. Devices that are close to each other and have communication needs are extracted to form a simplified subset of G1 (S1, S3, S4, OM1) and a simplified subset of G4 (S4, SRV1, SRV2).

[0040] Step S400: With the goal of maximizing communication efficiency and meeting signal strength requirements, re-plan the connection relationship for streamlining the device subset; adjust the device port configuration, and when the main link signal strength does not meet the threshold required by the service, enable the backup link and optimize the signal parameters; integrate all optimized sub-topologies to form a new topology structure set, and update the cluster device list.

[0041] Specifically, based on historical communication logs, bidirectional service flow data for each device is extracted, and the core parameters of each service flow are labeled; based on a simplified device subset, a binary communication relationship matrix is ​​constructed, and weights are labeled for matrix elements in combination with service flow priority; according to the principle of matching high-priority service flows with high-specification links first, service flows are initially associated with reusable link resources, and the matching results are recorded. Based on the physical deployment coordinates of the devices, the correlation between physical distance and link attenuation between devices is calculated. The closer the distance and the smaller the attenuation, the higher the priority. Devices with a weight greater than a set threshold in the communication relationship matrix are given priority for planning direct connection links. The shortest hop path is planned to avoid multi-node forwarding. Link aggregation groups are planned to ensure sufficient bandwidth. Multiple candidate schemes are generated. For each candidate scheme, the signal transmission process of each planned link is simulated. The actual signal parameters are calculated based on the link attenuation coefficient and the device port transmit power to determine whether the baseline threshold is met. When the signal parameters of a link are lower than the threshold, it is marked as a non-compliant link. For compliant links, the average transmission delay, average bandwidth utilization, and forwarding success rate are calculated, and a weighted calculation is used to obtain a comprehensive efficiency score. The scheme with a 100% signal compliance rate and the highest comprehensive efficiency score is selected as the optimal scheme, and the connection relationship is replanned.

[0042] Furthermore, for links that fail to meet the standards, the port parameters of both ends of the equipment are adjusted first, including rate negotiation, transmit power, and noise reduction configuration. If the main link still fails to meet the standards after port adjustment, the backup link within the sub-topology is checked, and link switching is triggered through the device configuration interface to switch the service flow to the backup link. The port gain and equalizer parameters at both ends of the backup link are adjusted to increase signal redundancy. All communication links between devices within the sub-topology are traversed to confirm that the signal values ​​meet the threshold requirements.

[0043] In one specific embodiment, connection relationships are planned for the simplified subsets of G1 (S1, S3, S4, OM1) and G4 (S4, SRV1, SRV2). Historical service flow data is extracted: the core transaction flow between S1 and S3 within G1 (peak 80Mbps, latency requirement ≤10ms, weight 1.0), and the transaction data flow between S4 and SRV1 within G4 (peak 50Mbps, latency requirement ≤12ms, weight 0.8). A communication matrix is ​​constructed and weights are labeled. Reusable links include the main link L001 (100G optical link, attenuation 0.3dB / km), the backup link L002 (100G optical link, attenuation 0.25dB / km), and the idle electrical link L003 (25G, attenuation 0.5dB / km).

[0044] Based on physical deployment, the distance between S1 and S3 is 10m (attenuation 3dB), and the distance between S1 and S4 is 15m (attenuation 4.5dB). Direct connections are prioritized for devices with a weight ≥ 0.8: S1-S3 uses L001, S1-S4 uses L003, and S4-SRV1 / SRV2 uses the original direct connection link. Two candidate schemes are generated: Scheme 1 primarily uses L001, and Scheme 2 primarily uses L002. Simulated signal transmission: In Scheme 1, the optical power of L001 is -3.5dBm (below the threshold of -3dBm), marking it as substandard; in Scheme 2, the optical power of L002 is -2.8dBm (meets the standard), with an average transmission latency of 8ms, bandwidth utilization of 60%, forwarding success rate of 99.95%, and a comprehensive efficiency score of 92 (higher than Scheme 1's 78). Scheme 2 is selected as the optimal scheme.

[0045] For the L002 link in Scheme 2, the measured optical power of -2.8dBm is close to the lower threshold. Adjusting the transmit power of the S1 port from -2dBm to -1.5dBm optimizes the optical power to -2.3dBm (increased redundancy). A review of all links shows that the signal parameters for S1-S3 (L002), S1-S4 (L003), and S4-SRV1 / SRV2 all meet the standards.

[0046] The optimized sub-topologies of G1, G4, and G3 are integrated: S1 is directly connected to S3 via L002, S1 is connected to S4 via L003, S4 is directly connected to SRV1 / SRV2, and S2 and S6 retain their original optimized links. The cluster device list is updated, the original faulty optical module OM2 is removed, and a backup link L002 and optimized port configuration records are added.

[0047] like Figure 2 As shown, this application provides a communication device status monitoring system based on signal analysis, comprising: Anomaly signal acquisition module: includes: a cross-cluster communication signal acquisition unit to acquire signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; a normal signal baseline establishment unit to compare the acquired signals with the baseline and filter out abnormal signals that deviate from the baseline and exceed the business allowable range; and anomaly signal filtering unit to extract information corresponding to the abnormal signals and form an anomaly dataset. Cluster topology parsing module: includes: an abnormal switch location unit to obtain the topology of each cluster and locate the specific node location of each abnormal switch based on the abnormal dataset; and an associated device set tracing unit that uses the abnormal switch node as the core to trace associated devices bidirectionally along the physical link and logical topology to form an associated communication device set. The simplified device subset generation module includes: an associated device grouping unit that sets grouping rules and splits each associated communication device set into multiple subgroups; an associated topology set integration unit that constructs an independent subtopology for each subgroup and integrates them to form an associated topology structure set; and a simplified device subset extraction unit that calculates the network distance between device nodes for each subtopology, extracts the device nodes that are closest to each other and have actual communication needs, and forms a simplified device subset. Topology optimization and cluster reconstruction module: includes: a simplified subset connection relationship planning unit that re-plans the connection relationship for simplified device subsets with the goal of maximizing communication efficiency and meeting signal strength standards; a link signal optimization unit that adjusts the device port configuration and enables the backup link and optimizes the signal parameters when the main link signal strength does not meet the threshold required by the service; and a new topology integration and list update unit that integrates all optimized sub-topologies to form a new topology structure set and updates the cluster device list.

[0048] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

Claims

1. A method for monitoring the status of communication equipment based on signal analysis, characterized in that, Includes the following steps: Collect signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; compare the collected signals with the normal signal baseline to filter out abnormal signals that deviate from the normal signal baseline and exceed the allowable range for business operations; extract the information corresponding to the abnormal signals to form an abnormal dataset; Obtain the topology of each cluster, locate the specific node location of each abnormal switch based on the abnormal dataset; take the abnormal switch as the core, trace the related devices bidirectionally along the physical link and logical topology to form a set of related communication devices; The method, centered on the abnormal switch, traces related devices bidirectionally along physical links and logical topologies to form a set of related communication devices, including: Forward tracing is performed on the physical link, starting from all service ports of the abnormal switch, matching the physical connection relationship of the service ports in the topology data, and extracting the downstream devices directly connected to the abnormal switch as direct downstream devices; for the traced direct downstream devices, their port connection relationship is further retrieved, and the downstream devices of the device are traced until the set tracing depth is reached; Reverse tracing is performed on the physical link, based on the uplink port connection relationship of the abnormal switch, tracing the upstream devices that directly provide signal or data input to the abnormal switch as direct upstream devices, and for the direct upstream devices, continuing to trace their upstream devices until the cluster core node or the peer cluster core device across the cluster link is traced. In the logical topology, forward tracing is performed to extract the VLAN configuration, broadcast domain, and ACL rules of the abnormal switch, and to trace all devices within the same VLAN or broadcast domain; based on the abnormal switch's routing table, ARP table, and traffic logs, the forwarding path of the communication service flow corresponding to the abnormal signal is located, and the downstream forwarding devices on that path are traced; in the logical topology, reverse tracing is performed to retrieve the abnormal switch's dynamic routing protocol neighbor table, link aggregation group member information, and SNMP management node information, and to trace protocol neighbor devices, aggregation group member devices, and SNMP management node devices. Set grouping rules to split each set of associated communication devices into multiple subgroups; for each subgroup, construct an independent subtopology and integrate them to form an associated topology structure set; for each subtopology, calculate the network distance between devices, extract the devices with the closest network distance and actual communication needs, and form a simplified device subset; With the goal of maximizing communication efficiency and achieving signal strength standards, the connection relationships of the streamlined device subset are replanned; the device port configurations of each device in the streamlined device subset are adjusted; when the signal strength of the main link in the streamlined device subset does not meet the threshold required by the service, the backup link is activated and the signal parameters are optimized; all sub-topologies with optimized signal parameters are integrated to form a new topology set, and the cluster device list is updated.

2. The communication equipment status monitoring method based on signal analysis according to claim 1, characterized in that, The process of collecting signals from communication links between different clusters and establishing a normal signal baseline based on historical signal data includes: Obtain a list of all cross-cluster communication links in the data center, label each link with a unique identifier, associate the service attributes carried by the link, and collect the signals of communication links between different clusters; deploy signal acquisition nodes to the switches at both ends of the cross-cluster links, and collect the native signal data of the links through the device's native interface; Historical signal data of cross-cluster communication links in the data center during normal operation cycles are retrieved, and the data is cleaned. Statistical feature values ​​are calculated for each link, each signal dimension, and each business scenario. The normal signal baseline includes the steady-state signal baseline and the time-series fluctuation signal baseline. The normal fluctuation range is defined by using several times the standard deviation of the mean as the baseline interval to form the steady-state signal baseline. The time-series baseline curve is fitted to mark the normal fluctuation patterns in different time periods to form the time-series fluctuation signal baseline.

3. The communication equipment status monitoring method based on signal analysis according to claim 2, characterized in that, The process of comparing the collected signals with a normal signal baseline and filtering out abnormal signals that deviate from the normal signal baseline and exceed the permitted range for the service includes: For steady-state signal baselines, the comparison rule is whether the real-time value exceeds the steady-state signal baseline range, while also considering the business-permissible range threshold. For time-series fluctuating signal baselines, the comparison rule is whether the real-time value deviates from the fluctuation range of the time-series fluctuating signal baseline curve, and whether the deviation exceeds the business-permissible deviation. The normal signal baseline parameters corresponding to the current link, current signal dimension, and current business scenario are retrieved, the deviation between the real-time signal parameters and the normal signal baseline is calculated, and it is determined whether the deviation exceeds the business-permissible range threshold. The comparison results are recorded. Based on the comparison results, abnormal signals that deviate from the normal signal baseline and whose real-time signal parameters deviate from the normal signal baseline exceed the business-permissible range are selected.

4. The communication equipment status monitoring method based on signal analysis according to claim 1, characterized in that, The process of obtaining the topology of each cluster and locating the specific node location of each abnormal switch based on the abnormal dataset includes: Topology data of all clusters is retrieved from both physical and logical topology dimensions, and the topology data is standardized to obtain standardized topology data. From the abnormal dataset, extract all relevant switch identification fields to form an abnormal switch identification list. Match the abnormal switch identification list with the cluster labels in the standardized topology data to determine the target cluster to which the abnormal switch belongs. In the physical topology of the target cluster, locate the specific physical node location of the abnormal switch based on its globally unique identifier. From the hierarchical labels in the topology data, clarify the network layer of the abnormal switch in the cluster and label the core functions of this layer. In the physical topology, find devices and links that have a direct physical connection to the abnormal switch and whose data flow is input to the abnormal switch, and record the upstream device identifier, connection port, link type, and bandwidth. Find devices and links that have a direct physical connection to the abnormal switch and whose data flow is output from the abnormal switch, and record the downstream device identifier, connection port, link type, and bandwidth. When the abnormal switch involves cross-cluster links, separately label the peer switch identifier, the cluster to which it belongs, and the logical attributes of the link. Based on the logical configuration in the topology data, supplement the logical upstream and downstream relationships of the abnormal switch, associate the communication service flow corresponding to the abnormal signal, locate the forwarding path of the communication service flow on the abnormal switch, and clarify the upstream and downstream devices from the perspective of the communication service flow.

5. The communication equipment status monitoring method based on signal analysis according to claim 1, characterized in that, The defined grouping rules divide each set of associated communication devices into multiple subgroups, including: Establish first-level grouping rules: devices with the same service type and priority, belonging to the same cluster, are grouped into the same initial group; devices with the same service type but different priorities, or devices with similar service types within the same cluster, are grouped into the same initial group. Establish second-level grouping rules: within the same initial group, further subdivide into core subgroups and secondary subgroups; within the same subgroup, when devices are physically deployed across multiple racks and have no direct communication requirements, they are split into different subgroups based on proximity. Calculate business coupling ,in, For the first subgroup With the second subgroup Average daily business interaction traffic between them for The average daily total communication traffic, for Average daily total communication traffic; calculation logic configuration correlation. Where p is the total number of logical configuration types. Configure associated weights for the m-th type of logic. for and The m-th logical configuration association identifier, This indicates the existence of the m-th type of association. Indicates no correlation; calculate the overall coupling degree. , This is a weighting coefficient for the degree of business coupling. Configure the weight coefficients for logical relevance; When the overall coupling degree of two subgroups is greater than a set threshold, highly interactive devices between the two subgroups are extracted. If the interaction is cross-group, the highly interactive devices are split to form a new subgroup.

6. The communication equipment status monitoring method based on signal analysis according to claim 5, characterized in that, For each subgroup, an independent sub-topology is constructed, which is then integrated to form a set of related topological structures, including: Sub-topology includes sub-physical topology graph and sub-logical topology graph; a sub-physical topology graph is constructed with devices as vertices and physical links as edges, and logical configuration information is superimposed on the physical topology layer to form a sub-logical topology graph; information of abnormal switches in the abnormal data corresponding to abnormal switches in the subgroup is associated with the sub-topology.

7. The communication equipment status monitoring method based on signal analysis according to claim 1, characterized in that, The goal of maximizing communication efficiency and achieving signal strength standards involves reorganizing connection relationships for a streamlined subset of devices, including: Based on historical communication logs, bidirectional communication service flow data for each device is extracted, and the core parameters of each communication service flow are labeled. Based on a simplified device subset, a binary communication relationship matrix is ​​constructed, and the matrix elements are labeled with weights according to the priority of communication service flows. According to the principle of matching high-priority communication service flows with high-specification links first, communication service flows and reusable link resources are initially associated, and the matching results are recorded. Based on the physical deployment coordinates of the devices, the correlation between physical distance and link attenuation between devices is calculated. The closer the distance and the smaller the attenuation, the higher the priority. Devices with a weight greater than a set threshold in the communication relationship matrix are given priority for planning direct connection links. The shortest hop path is planned to avoid multi-node forwarding. Link aggregation groups are planned to ensure sufficient bandwidth. Multiple candidate schemes are generated. For each candidate scheme, the signal transmission process of each planned link is simulated. The actual signal parameters are calculated based on the link attenuation coefficient and the device port transmit power to determine whether they meet the threshold corresponding to the normal signal baseline. When the signal parameters of a link are lower than the threshold corresponding to the normal signal baseline, it is marked as a non-compliant link. For compliant links, the average transmission delay, average bandwidth utilization, and forwarding success rate are calculated, and a weighted calculation is used to obtain a comprehensive efficiency score. The scheme with a 100% signal compliance rate and the highest comprehensive efficiency score is selected as the optimal scheme, and the connection relationship is replanned.

8. A communication equipment status monitoring system based on signal analysis, using the communication equipment status monitoring method based on signal analysis according to any one of claims 1-7, characterized in that, include: Anomaly signal acquisition module: includes: a cross-cluster communication signal acquisition unit to acquire signals from communication links between different clusters and establish a normal signal baseline based on historical signal data; a normal signal baseline establishment unit to compare the acquired signals with the normal signal baseline and filter out abnormal signals that deviate from the normal signal baseline and exceed the business allowable range; and anomaly signal filtering unit to extract information corresponding to the abnormal signals and form an anomaly dataset. Cluster topology parsing module: includes: an abnormal switch location unit to obtain the topology of each cluster and locate the specific node location of each abnormal switch based on the abnormal dataset; and an associated device set tracing unit that uses the abnormal switch as the core to trace associated devices bidirectionally along the physical link and logical topology to form an associated communication device set. The simplified device subset generation module includes: an associated device grouping unit that sets grouping rules and splits each associated communication device set into multiple subgroups; an associated topology set integration unit that constructs an independent subtopology for each subgroup and integrates them to form an associated topology structure set; and a simplified device subset extraction unit that calculates the network distance between devices for each subtopology, extracts the device with the closest network distance and actual communication needs, and forms a simplified device subset. Topology optimization and cluster reconstruction module: includes: a simplified subset connection relationship planning unit that re-plans the connection relationship for the simplified device subset with the goal of maximizing communication efficiency and meeting signal strength requirements; a link signal optimization unit that adjusts the device port configuration of each device in the simplified device subset, and when the main link signal strength in the simplified device subset does not meet the threshold required by the service, it activates the backup link and optimizes the signal parameters; and a new topology integration and list update unit that integrates all sub-topologies with optimized signal parameters to form a new topology structure set and updates the cluster device list.

Images