Network link performance detection system and method based on signal monitoring

Abstract

The invention discloses a network link performance detection system and method based on signal monitoring, and the system comprises a logic performance monitoring module which is used for carrying out the inspection of a logic performance index, and transmitting a transient capture trigger signal when a transient abnormal preset condition is satisfied; the physical electric signal sampling module is used for collecting original physical electric signals in real time; the cross-layer real-time triggering and capturing module is provided with an annular cache region for continuously caching the original physical electric signal and solidifying cache data to generate a transient electric signal snapshot after receiving a transient capturing triggering signal; and the associated diagnosis engine is used for performing weight-level physical electric signal characteristic analysis on the transient electric signal snapshot and outputting a diagnosis result in combination with the logic performance index. According to the method, backtracking capture of the original physical electric signals is triggered in real time through the logic indexes, accurate evidence obtaining of transient physical faults is achieved, logic congestion can be effectively distinguished, and the diagnosis accuracy and the utilization efficiency of measurement resources are greatly improved.

Description

A network link performance testing system and method based on signal monitoring Technical Field

[0001] This invention belongs to the field of network technology and relates to network performance testing, specifically to a network link performance testing system and method based on signal monitoring. Background Technology

[0002] In current network operation and maintenance, monitoring network link performance is crucial. Existing technologies, such as tools like Ping or Traceroute, primarily operate at the logical layers of the network, such as the network or transport layers. These tools assess link connectivity by sending and receiving probe packets to statistically analyze logical performance metrics such as packet loss rate, latency, and jitter.

[0003] However, the aforementioned existing technologies have a significant drawback: they cannot distinguish the root cause of performance degradation. For example, when a system detects a high packet loss rate, existing technologies cannot determine whether the problem is caused by logical layer congestion (such as switch buffer overflow or bandwidth saturation) or by signal distortion caused by physical layer degradation (such as cable aging, poor interface contact, or electromagnetic interference). The solutions for these two causes are drastically different, but existing technologies cannot provide a clear diagnosis.

[0004] Building on this, another challenging issue arises: how to efficiently capture intermittent or transient physical layer faults. Many severe physical layer interferences, such as the strong electromagnetic pulses generated when large equipment starts up, may only last for a few hundred milliseconds. Such interference can cause a large number of instantaneous packet losses in the network, but by the time maintenance personnel or detection systems intervene afterward, the interference has disappeared, and the electrical signal measurements of the physical links have returned to normal, making it impossible to reproduce the fault at the scene.

[0005] If we attempt to continuously monitor the raw electrical signals of all network ports 24 / 7 using high-precision electrical signal measurement equipment (such as oscilloscopes or spectrum analyzers), it would result in extremely high hardware costs and massive data processing pressure, which is not feasible in practice. It is impossible to automatically trigger deep capture of the raw electrical signals of the physical layer at the moment when the logic performance indicators change abruptly, which would easily lead to the loss of a large amount of evidence of occasional and transient physical layer faults (such as instantaneous interference waveforms), making it difficult to eradicate the problem of occasional network jitter in the long term. Therefore, there is an urgent need for a network link performance detection system and detection method based on signal monitoring. Summary of the Invention

[0006] The purpose of this invention is to address the deficiencies in the existing technology by proposing a network link performance detection system and method based on signal monitoring.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a network link performance detection system based on signal monitoring, comprising: a logic performance monitoring module, used to perform lightweight logic performance inspection on the network link to obtain the logic performance indicators of the network link, and to generate and send a transient capture trigger signal when the logic performance indicators meet the transient anomaly preset conditions; a physical electrical signal sampling module, used to collect in real time the original physical electrical signals corresponding to the data signals transmitted on the network link; and a cross-layer real-time triggering and capture module, which connects the physical electrical signal sampling module and the logic performance monitoring module, and has an internal ring buffer. The circular buffer is used to continuously buffer the original physical electrical signals. The cross-layer real-time triggering and capture module is used to, after receiving the transient capture trigger signal, freeze the original physical electrical signals in the circular buffer before and after the trigger time corresponding to the transient capture trigger signal to generate a transient electrical signal snapshot. The associated diagnostic engine, which connects the cross-layer real-time triggering and capture module and the logic performance monitoring module, is used to perform heavyweight physical electrical signal characteristic analysis on the transient electrical signal snapshot after the cross-layer real-time triggering and capture module generates the transient electrical signal snapshot, and output the diagnostic results of the network link in combination with the logic performance indicators.

[0008] Furthermore, the logic performance indicators obtained by the logic performance monitoring module include at least one of the following: packet loss rate, latency, jitter, or CRC error count.

[0009] The transient anomaly preset conditions include: the rate of change or absolute value of the packet loss rate or CRC error count in the logical performance indicators within the preset inspection time window exceeds the preset trigger threshold.

[0010] The physical electrical signal sampling module is specifically used to measure the analog characteristics of the original physical electrical signal, and the analog characteristics are used for the physical electrical signal characteristic analysis.

[0011] The cross-layer real-time triggering and capture module is specifically used to: solidify the original physical electrical signals in the annular buffer for a preset first duration before the triggering time and a preset second duration after the triggering time, so as to generate the transient electrical signal snapshot.

[0012] The physical electrical signal characteristic analysis performed by the associated diagnostic engine includes at least one of the following: spectral analysis, eye diagram analysis, or time-domain reflectometry analysis of the transient electrical signal snapshot.

[0013] The correlation diagnostic engine is also used to: build an interference fingerprint database; compare the results of the physical electrical signal characteristic analysis with the interference fingerprint database to identify whether transient electromagnetic interference features exist in the transient electrical signal snapshot.

[0014] The diagnostic results output by the associated diagnostic engine include: physical layer transient fault diagnosis: output when the logic performance index meets the transient anomaly preset condition and the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is abnormal; logic layer congestion diagnosis: output when the logic performance index meets the transient anomaly preset condition, but the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is normal.

[0015] A network link performance detection method based on signal monitoring, applied to the aforementioned system, includes the following steps: Logical performance inspection step: A lightweight logical performance inspection of the network link is performed by a logical performance monitoring module to obtain the logical performance indicators of the network link; Physical signal caching step: The original physical electrical signals on the network link are collected in real time by a physical electrical signal sampling module, and the original physical electrical signals are continuously cached in a circular buffer of a cross-layer real-time triggering and capture module; Transient triggering and capture step: When the logical performance indicators meet the preset conditions for transient anomalies, the logical performance monitoring module generates and sends a transient capture trigger signal to the cross-layer real-time triggering and capture module. After receiving the transient capture trigger signal, the cross-layer real-time triggering and capture module freezes the original physical electrical signals in the circular buffer to generate a transient electrical signal snapshot; Correlation diagnosis step: A heavyweight physical electrical signal characteristic analysis is performed on the transient electrical signal snapshot by a correlation diagnosis engine, and the diagnostic results of the network link are output in conjunction with the logical performance indicators.

[0016] Furthermore, the associated diagnostic step specifically includes: when the logic performance indicator meets the transient anomaly preset condition and the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is abnormal, the diagnostic result is determined to be a physical layer transient fault diagnosis; when the logic performance indicator meets the transient anomaly preset condition, but the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is normal, the diagnostic result is determined to be a logic layer congestion diagnosis.

[0017] Compared to existing technologies, the advantages of this invention are as follows: by setting up a cross-layer real-time triggering and capture module and its circular buffer, and designing a linkage mechanism that triggers the solidification of physical signals by transient anomalies in logical performance indicators, "backtracking" fault scene evidence collection for network links is achieved. This solves the pain point of existing technologies that cannot capture sporadic, transient physical layer interference (such as EMI pulses) due to the disappearance of interference, and greatly improves the accuracy of fault diagnosis.

[0018] By employing a combination of lightweight inspection and heavyweight analysis, the system normally requires only low-overhead logic performance monitoring, activating the correlated diagnostic engine only when an anomaly is detected to perform in-depth analysis of the stored transient electrical signal snapshots. This avoids the enormous overhead of performing 24 / 7 high-precision electrical signal measurements on all ports, greatly improving the utilization efficiency of measurement resources while ensuring diagnostic depth.

[0019] The correlation diagnostic engine, by comparing the analysis results of logical performance metrics with transient electrical signal snapshots, can clearly distinguish between physical layer transient fault diagnosis and logical layer congestion diagnosis. This solves the fundamental problem of not being able to determine whether performance degradation is caused by physical degradation or logical congestion, providing clear guidance for network operations and maintenance. Attached Figure Description

[0020] Figure 1 is a structural block diagram of a network link performance detection system based on signal monitoring provided in an embodiment of the present invention; Figure 2 is a flowchart of a network link performance detection method based on signal monitoring provided in an embodiment of the present invention. Detailed Implementation

[0021] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0022] As shown in Figure 1-2, a network link performance detection system based on signal monitoring includes: a logical performance monitoring module, a physical electrical signal sampling module, a cross-layer real-time triggering and capture module, and an associated diagnostic engine.

[0023] The logical performance monitoring module is used to perform continuous, lightweight logical performance inspections of network links. In this embodiment, the logical performance monitoring module can obtain real-time logical performance indicators of the network links by periodically sending ICMPPing probe packets or reading the CRC error counter of the switch port via the SNMP protocol. These logical performance indicators include, but are not limited to, packet loss rate, average latency, jitter, or CRC error count values.

[0024] The logical performance monitoring module also has preset transient anomaly conditions. These conditions are used to identify occasional, sudden network performance degradation, rather than continuous performance degradation. For example, a transient anomaly condition can be set as follows: within a 1-second time window, the packet loss rate suddenly jumps from 0% to over 10% or the instantaneous growth rate of the CRC error count exceeds 100 per second.

[0025] When the logic performance monitoring module detects that the logic performance indicators meet the above-mentioned transient anomaly preset conditions, the logic performance monitoring module will immediately generate and send a transient capture trigger signal to the cross-layer real-time trigger and capture module.

[0026] A physical electrical signal sampling module is a high-precision electrical signal acquisition device, for example, physically connected to the MDI interface of a network link (such as an RJ45 interface signal pair) via a high-impedance differential probe. The physical electrical signal sampling module does not parse data packet content; its sole function is to acquire, in real-time at high frequency, the original physical electrical signal corresponding to the data signal transmitted on the network link, i.e., the analog voltage waveform of the signal.

[0027] The cross-layer real-time triggering and capture module is connected to the physical electrical signal sampling module and the logic performance monitoring module, respectively.

[0028] The cross-layer real-time triggering and capture module internally features a large-capacity circular buffer. This circular buffer is a high-speed FIFO memory that continuously receives and buffers raw physical electrical signals from the physical electrical signal sampling module, constantly overwriting the oldest signals with the newest ones. For example, the circular buffer can be configured to always retain the raw physical electrical signals from the most recent 5 seconds.

[0029] When the cross-layer real-time triggering and capture module receives a transient capture trigger signal from the logic performance monitoring module, it immediately performs a hardening action. Hardening means that the cross-layer real-time triggering and capture module immediately stops the write pointer of its circular buffer, preventing it from overwriting data, thereby locking the original physical electrical signals before and after the trigger moment in the buffer.

[0030] In this embodiment, the specific action of solidification is as follows: the original physical electrical signals for a preset first duration (e.g., 2 seconds) before the trigger moment and a preset second duration (e.g., 3 seconds) after the trigger moment are extracted from the circular buffer and packaged to generate a static transient electrical signal snapshot. The transient electrical signal snapshot is like the video clip locked by a dashcam when a collision occurs; the transient electrical signal snapshot completely preserves the physical crime scene before and after the malfunction.

[0031] The correlation diagnostic engine is used to perform heavyweight deep analysis. It connects the cross-layer real-time triggering and capture module and the logic performance monitoring module. During normal inspections, the correlation diagnostic engine can be in standby or low-power state to conserve resources. The correlation diagnostic engine is activated when the cross-layer real-time triggering and capture module generates a snapshot of transient electrical signals.

[0032] The correlation diagnostic engine performs heavyweight physical electrical signal characteristic analysis on this static transient electrical signal snapshot. Since the analysis is of static data, it is no longer subject to the stringent constraints of real-time performance. Therefore, the correlation diagnostic engine can perform very complex measurement algorithms, such as: frequency spectrum analysis (FFT) to analyze whether there are abnormal harmonic interferences in the signal, such as 50Hz power frequency interference or EMI pulses at specific frequencies; eye diagram plotting analysis, which draws eye diagrams by superimposing waveforms to quantify and analyze key electrical variables such as signal amplitude, signal-to-noise ratio, and jitter; and time-domain reflectometry (TDR) analysis, which determines whether there is impedance mismatch in the link by analyzing the reflected waveform of the signal.

[0033] In addition, the correlation diagnostic engine has a built-in interference fingerprint database. The interference fingerprint database pre-stores the unique spectral characteristics of common electromagnetic interference sources (such as motor starting, welding machines, etc.). The correlation diagnostic engine compares the results of physical electrical signal characteristic analysis (such as FFT spectrum) with the interference fingerprint database to attempt to automatically identify the interference source.

[0034] Finally, the correlation diagnostic engine will combine two input sources to make a final decision: one is the logic performance metric from the logic performance monitoring module (e.g., packet loss rate of 15%); the other is the result of the correlation diagnostic engine's own analysis of the physical electrical signal characteristics of transient electrical signal snapshots.

[0035] As shown in Figure 2, this embodiment of the invention also provides a network link performance detection method based on signal monitoring. Its workflow is as follows: After the system is powered on, the logic performance monitoring module continuously performs lightweight inspections of the network link (e.g., pinging once per second); this is the logic performance inspection step. Simultaneously, the circular buffer of the cross-layer real-time triggering and capture module continuously buffers the original physical electrical signals collected by the physical electrical signal sampling module; this is the physical signal buffering step. The system is in a low-overhead sentinel state.

[0036] At a certain moment, for example, when a large motor starts, it generates a strong EMI pulse. The logic performance monitoring module immediately detects that the logic performance indicators meet the preset conditions for transient anomalies (e.g., a surge in CRC errors). The logic performance monitoring module immediately sends a transient capture trigger signal. Upon receiving the signal, the cross-layer real-time triggering and capture module immediately saves the data in its circular buffer, generating a snapshot of the transient electrical signal containing the EMI pulse waveform. This is the transient triggering and capture step.

[0037] Subsequently, the correlation diagnostic engine is activated and begins to perform heavyweight physical electrical signal characteristic analysis on the transient electrical signal snapshot. Combined with logical performance indicators, it outputs the diagnostic results of the network link. This is the correlation diagnostic step.

[0038] In the correlation diagnosis step, there are two scenarios: First, the correlation diagnosis engine performs FFT analysis on the transient electrical signal snapshot and finds a significant abnormal spectral peak, which matches the motor starting characteristic in the interference fingerprint database. Simultaneously, logic performance indicators show high packet loss. In this case, the system outputs the diagnostic result: Physical layer transient fault diagnosis. Cause: Transient electromagnetic interference detected, with characteristics resembling motor starting. It is recommended to investigate strong interference sources around the link.

[0039] In the second scenario, the diagnostic engine performed all analyses (FFT, eye diagram) on the transient electrical signal snapshots and found that all physical electrical signal characteristics were good (high signal-to-noise ratio, low jitter, no interference). However, the logical performance metrics still showed high packet loss and high latency. In this case, the system output the diagnostic result: Logical layer congestion diagnosis. Reason: The physical link quality is good; the performance degradation is caused by network congestion. It is recommended to check the switch cache or network bandwidth.

[0040] In summary, by introducing a cross-layer real-time triggering and capture module, and designing a linked workflow of logical inspection triggering, physical cache persistence, and post-event in-depth analysis, the technical challenge of accurately capturing transient physical layer interference and distinguishing it from logical layer congestion is solved without significantly increasing the system's regular operating overhead.

[0041] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A network link performance testing system based on signal monitoring, characterized in that, include: The logic performance monitoring module is used to perform lightweight logic performance inspection on the network link to obtain the logic performance indicators of the network link, and generate and send a transient capture trigger signal when the logic performance indicators meet the transient anomaly preset conditions. The physical electrical signal sampling module is used to collect the original physical electrical signals corresponding to the data signals transmitted on the network link in real time; A cross-layer real-time triggering and capture module is provided, which is connected to the physical electrical signal sampling module and the logic performance monitoring module. The module has a ring buffer inside, which is used to continuously buffer the original physical electrical signal. After receiving the transient capture trigger signal, the cross-layer real-time triggering and capture module is used to save the original physical electrical signal in the ring buffer before and after the trigger time corresponding to the transient capture trigger signal to generate a transient electrical signal snapshot. The associated diagnostic engine connects the cross-layer real-time triggering and capture module and the logic performance monitoring module. It is used to perform heavyweight physical electrical signal characteristic analysis on the transient electrical signal snapshot after the cross-layer real-time triggering and capture module generates a transient electrical signal snapshot, and output the diagnostic results of the network link in combination with the logic performance indicators.

2. The system according to claim 1, characterized in that, The logic performance indicators obtained by the logic performance monitoring module include at least one of the following: packet loss rate, latency, jitter, or CRC error count.

3. The system according to claim 1, characterized in that, The transient anomaly preset conditions include: the rate of change or absolute value of the packet loss rate or CRC error count in the logical performance indicators within the preset inspection time window exceeds the preset trigger threshold.

4. The system according to claim 1, characterized in that, The physical electrical signal sampling module is specifically used to measure the analog characteristics of the original physical electrical signal, and the analog characteristics are used for the physical electrical signal characteristic analysis.

5. The system according to claim 1, characterized in that, The cross-layer real-time triggering and capture module is specifically used to: solidify the original physical electrical signals in the ring buffer for a preset first duration before the triggering time and a preset second duration after the triggering time, so as to generate the transient electrical signal snapshot.

6. The system according to claim 1, characterized in that, The physical electrical signal characteristic analysis performed by the associated diagnostic engine includes at least one of the following: spectral analysis, eye diagram analysis, or time-domain reflectometry analysis of the transient electrical signal snapshot.

7. The system according to claim 6, characterized in that, The correlation diagnostic engine is also used to: build an interference fingerprint database; compare the results of the physical electrical signal characteristic analysis with the interference fingerprint database to identify whether transient electromagnetic interference features exist in the transient electrical signal snapshot.

8. The system according to claim 1, characterized in that, The diagnostic results output by the associated diagnostic engine include: physical layer transient fault diagnosis: output when the logic performance index meets the transient anomaly preset condition and the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is abnormal; logic layer congestion diagnosis: output when the logic performance index meets the transient anomaly preset condition, but the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is normal.

9. A method for detecting network link performance based on signal monitoring, characterized in that, The method is applied to the system as described in any one of claims 1 to 8, and the method includes the following steps: a logic performance inspection step: performing a lightweight logic performance inspection on the network link through a logic performance monitoring module to obtain the logic performance indicators of the network link; a physical signal caching step: acquiring the original physical electrical signals on the network link in real time through a physical electrical signal sampling module, and continuously caching the original physical electrical signals through a ring buffer of a cross-layer real-time triggering and capture module; a transient triggering and capture step: when the logic performance indicator meets the preset condition for transient anomaly, the logic performance monitoring module generates and sends a transient capture trigger signal to the cross-layer real-time triggering and capture module, and after receiving the transient capture trigger signal, the cross-layer real-time triggering and capture module solidifies the original physical electrical signals in the ring buffer to generate a transient electrical signal snapshot; and a correlation diagnosis step: performing a heavyweight physical electrical signal characteristic analysis on the transient electrical signal snapshot through a correlation diagnosis engine, and outputting the diagnostic results of the network link in conjunction with the logic performance indicators.

10. The method according to claim 9, characterized in that, The associated diagnostic steps specifically include: when the logic performance index meets the transient anomaly preset condition and the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is abnormal, the diagnostic result is determined to be a physical layer transient fault diagnosis; when the logic performance index meets the transient anomaly preset condition, but the result of the physical electrical signal characteristic analysis performed on the transient electrical signal snapshot is normal, the diagnostic result is determined to be a logic layer congestion diagnosis.

Images