Voice signaling tracking service system for tracking calls

Abstract

The invention discloses a voice signaling tracking service system for tracking a call, and belongs to the technical field of mobile communication services. According to the method, the problems of long troubleshooting time and influence on the system recovery speed due to dependence on manual analysis and lack of an automatic analysis technology in the prior art are solved, and by capturing the signaling message in real time and analyzing the signaling message based on the signaling process model, abnormal and potential faults can be quickly found, and the delay of fault detection is reduced; the adjustment strategy corresponding to the fault type is obtained from the decision tree model, the network configuration or signaling process is dynamically adjusted, manual intervention is reduced, and the fault processing efficiency is improved; according to the voice signaling tracking service system, an alarm mechanism is arranged, and when an abnormality is detected, related personnel are notified in time, so that a fault can be found and processed in time, the real-time performance, the accuracy and the automation level of the voice signaling tracking service system are improved, and the quality of mobile communication service is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mobile communication services, and specifically to a voice signaling tracking service system for tracking calls. Background Art

[0002] Signaling needs to be parsed and processed by the receiving party during transmission. For example, in an IP telephone system, the service logic layer is responsible for managing call states, such as ringing, in call, hung up, etc., and triggering state changes based on user actions or other factors. The voice signaling tracking service is used to track the establishment process between calls, facilitating operation and maintenance personnel to locate and find faults generated during calls. The tracked signaling information includes: session information exchange, device capability negotiation, and other information.

[0003] However, in the prior art, the fault location of the system mainly relies on manual analysis and lacks the support of automated tools, resulting in long fault troubleshooting time and affecting the system recovery speed.

[0004] Therefore, it does not meet the existing requirements, and for this reason, we propose a voice signaling tracking service system for tracking calls. Summary of the Invention

[0005] The purpose of the present invention is to provide a voice signaling tracking service system for tracking calls. By capturing signaling messages in real time and analyzing them based on a signaling process model, it can quickly detect anomalies and potential faults, reducing the delay in fault detection; obtaining adjustment strategies corresponding to fault types from a decision tree model, dynamically adjusting network configurations or signaling processes, reducing manual intervention, and improving the efficiency of fault handling; by setting an alarm mechanism, relevant personnel can be notified in a timely manner when anomalies are detected, thereby ensuring that faults can be discovered and processed in a timely manner, avoiding the expansion of problems, and thus enhancing the real-time performance, accuracy, and automation level of the voice signaling tracking service system, improving the quality of mobile communication services, and solving the problems raised in the above background art.

[0006] To achieve the above purpose, the present invention provides the following technical solutions:

[0007] A voice signaling tracking service system for tracking calls, the system includes: a signaling tracking unit and a fault handling unit, and the fault handling unit includes a model construction module, a signaling analysis module, and a tracking service module;

[0008] The signaling tracking unit is configured to capture and record signaling messages during a call in real time;

[0009] The model construction module is configured to build a signaling process model for normal calls based on historical signaling messages, define anomaly detection rules and signaling fault types based on expert experience, use a rule engine to detect anomalies in the process, and build a decision tree model to classify fault types, historical signaling messages, and adjustment strategies;

[0010] The signaling analysis module is configured to analyze real-time signaling messages based on the signaling process model, promptly detect anomalies and potential faults; and set an alarm mechanism to trigger an alarm when an anomaly is detected;

[0011] The tracking service module is configured to locate the cause of the fault based on the anomaly detection result, and obtain the adjustment strategy of the same fault type from the decision tree model for dynamic adjustment.

[0012] Furthermore, the voice signaling tracking service system further includes:

[0013] The signaling tracking quality monitoring module is used to monitor the operating parameters of the signaling tracking unit in real time and determine anomalies in the signaling tracking quality;

[0014] Wherein, the signaling tracking quality monitoring module performs the following steps:

[0015] Monitor the operating parameters of the signaling tracking unit in real time, wherein the operating parameters include the signaling capture rate (unit: number of messages per second), the message capture delay ratio (unit: dimensionless), and the storage bandwidth occupancy ratio (unit: dimensionless);

[0016] Retrieve the storage bandwidth occupancy ratio and the message capture delay ratio corresponding to each unit time in real time;

[0017] Calculate the available storage bandwidth occupancy ratio corresponding to the storage bandwidth occupancy ratio corresponding to each unit time;

[0018] Compare the available storage bandwidth occupancy ratio corresponding to each unit time with the message capture delay ratio;

[0019] When the available storage bandwidth occupancy ratio corresponding to each unit time is lower than the message capture delay ratio, the operating quality of the signaling tracking unit is determined using the signaling capture rate, the message capture delay ratio, and the storage bandwidth occupancy ratio.

[0020] Furthermore, when the available storage bandwidth occupancy ratio corresponding to each unit time is lower than the message capture delay ratio, determining the operating quality of the signaling tracking unit using the signaling capture rate, the message capture delay ratio, and the storage bandwidth occupancy ratio includes:

[0021] When the ratio of the available storage bandwidth corresponding to each unit time is lower than the message capture delay ratio, perform a ratio process on the ratio of the available storage bandwidth to the message capture delay ratio to obtain reference ratio data;

[0022] Perform a conversion process on the reference ratio data to obtain the reference ratio corresponding to the converted reference ratio data;

[0023] Among them, the model of the conversion process is as follows:

[0024]

[0025] Among them, s represents the reference ratio; k represents the reference ratio data;

[0026] Extract the signaling capture rate change amplitude between the signaling capture rate corresponding to each unit time and the signaling capture rate corresponding to the previous unit time;

[0027] Use the signaling capture rate change amplitude between the signaling capture rate corresponding to each unit time and the signaling capture rate corresponding to the previous unit time and the signaling capture rate corresponding to the current unit time to obtain the signaling capture rate change rate;

[0028] Compare the signaling capture rate change rate with the reference ratio;

[0029] According to the signaling capture rate change rate exceeding the reference ratio, it is determined that there is a risk of abnormal operation of the signaling tracking unit, and an early warning of abnormal operation quality risk is given.

[0030] Furthermore, the model construction module includes:

[0031] The process definition module is configured to define a standard signaling process model based on the historical signaling messages of normal calls and determine the normal behavior of each step, including: time range, message format;

[0032] The rule definition module is configured to define abnormal detection rules in the signaling process model based on expert experience to form a rule engine mode, so that the signaling process model implements abnormal detection rules based on the use of the rule engine;

[0033] The model training module is configured to train the signaling process model based on the training set of historical signaling messages, so that the signaling process model can accurately describe the normal signaling process; verify the accuracy of the signaling process model based on the test set of historical signaling messages, so that the signaling process model can correctly identify the normal signaling process.

[0034] Furthermore, the model construction module also includes:

[0035] A fault definition module, configured to define signaling fault types based on expert experience, and provide a detailed description, possible causes, and corresponding adjustment strategies for each fault type;

[0036] A fault classification module, configured to classify fault types, historical signaling messages, and adjustment strategies based on the key features of historical signaling messages, enabling the decision tree model to correctly classify fault types and adjustment strategies.

[0037] Furthermore, the signaling analysis module includes:

[0038] A model integration module, configured to integrate the signaling process model into the existing system based on a message queue, and configure the topic and partition of the message queue to ensure the orderliness and reliability of signaling messages; use the message queue as the transmission and buffer layer for signaling messages, and after capturing real-time signaling messages, send the signaling messages into the message queue;

[0039] A signaling analysis module, configured to obtain signaling messages from the message queue in real time for the signaling process model and perform real-time analysis using a rule engine;

[0040] An alarm definition module, configured to define alarm rules based on the anomaly detection results of historical signaling messages, and trigger the alarm rules and synchronously feedback to the management department when the signaling process model analyzes that there are anomalies in real-time signaling messages; the alarm rules include: signaling loss alarm and delay anomaly alarm.

[0041] Furthermore, the tracking service module includes:

[0042] A fault location module, configured to locate the cause of the fault based on the anomaly detection results, combined with the network topology and signaling logs, including: network congestion, equipment failure, configuration error;

[0043] A fault matching module, configured to match the located cause of the fault with the fault types in the decision tree model to obtain adjustment strategies;

[0044] A dynamic repair module, configured to execute the adjustment strategy using the Ansible automation tool based on the obtained adjustment strategy, and dynamically adjust the network configuration or signaling process.

[0045] Furthermore, the tracking service module further includes:

[0046] A repair verification module, configured to continuously monitor the network status and signaling process, verify whether the adjustment strategy is effective, and ensure that the system returns to normal;

[0047] A repair feedback module, configured to collect the execution results of the adjustment strategy and optimize the decision tree model and adjustment strategy.

[0048] Further, the signaling tracking unit includes:

[0049] A signaling collection module, configured to collect historical signaling messages, and capture and record all signaling messages during a call using a signaling tracking tool to form a signaling log; the signaling messages include: Invite session initiation request, 100 Trying request in progress, 180 Ringing the called party is ringing, 200 OK the called party accepts the request and establishes a call, and ACK the calling party confirms the call establishment;

[0050] A process check module, configured to check whether the signaling process is complete based on the signaling log.

[0051] Further, the signaling tracking unit further includes:

[0052] A data processing module, configured to clean the signaling messages to remove noise and invalid data; perform standardization processing on the signaling messages, including: unifying the timestamp format and field naming; converting the signaling messages into a format suitable for input to the signaling process model;

[0053] A feature extraction module, configured to extract key features of the signaling messages, including: message type, timestamp, delay, and SDP information, calculate statistical features, including: average delay, maximum delay, and message loss rate, and use feature importance screening to filter out feature information that affects anomaly detection.

[0054] Compared with the prior art, the beneficial effects of the present invention are:

[0055] In the present invention, by capturing signaling messages in real time and analyzing them based on the signaling process model, anomalies and potential faults can be quickly discovered, the delay in fault detection can be reduced, and the response speed of the system can be improved; obtain adjustment strategies corresponding to the fault types from the decision tree model, dynamically adjust the network configuration or signaling process, thereby reducing manual intervention, improving the efficiency of fault handling, and avoiding affecting the call quality due to untimely fault handling; by setting an alarm mechanism, relevant personnel are notified in a timely manner when an anomaly is detected, so as to ensure that the fault can be discovered and processed in a timely manner, avoid the problem from expanding, thereby improving the real-time performance, accuracy, and automation level of the voice signaling tracking service system, optimizing the network performance, reducing the operation and maintenance costs, and improving the quality of mobile communication services. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a block diagram of the components of the voice signaling tracking service system for tracking calls according to the present invention;

[0057] Figure 2 It is a flowchart of the voice signaling tracking service according to the present invention;

[0058] Figure 3Code interface diagram of the voice signaling tracking service of the present invention;

[0059] Figure 4 Interface diagram of the voice signaling tracking of the present invention. Detailed implementation manners

[0060] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0061] To solve the technical problem in the prior art that the fault location of the system mainly relies on manual analysis and lacks the support of automated tools, resulting in a long time-consuming fault troubleshooting and affecting the system recovery speed, please refer to Figures 1-4 In this embodiment, the following technical solutions are provided:

[0062] A voice signaling tracking service system for tracking calls, the system includes: a signaling tracking unit and a fault handling unit;

[0063] The signaling tracking unit is configured to capture and record signaling messages during a call in real time; the signaling tracking unit includes:

[0064] A signaling collection module for collecting historical signaling messages, and capturing and recording all signaling messages during a call using a signaling tracking tool to form a signaling log; the signaling messages include: Invite initiate session request, 100 Trying request in progress, 180 Ringing called party is ringing, 200OK called party accepts the request and establishes a call, and ACK calling party confirms call establishment; specifically, extracting historical signaling messages from a database or log file for analysis and modeling; using a signaling tracking tool such as SIPp to capture signaling messages during a call in real time, and recording the captured signaling messages in chronological order to form a structured signaling log; for example: each signaling message should include the following fields: timestamp: record the sending or receiving time of the message; message type: such as Invite, 100 Trying, 180 Ringing, etc.; source address: the IP address and port of the sender; destination address: the IP address and port of the receiver; message content: such as SDP information, error code, etc.

[0065] The process check module is configured to check whether the signaling process is complete based on the signaling log. For example, whether the Invite-initiated session request is successfully sent and the 100 Trying request being processed response is received; whether the 180 Ringing (the callee is ringing) and 200 OK (the callee accepts the request and the call is established) are received; whether the ACK (the caller confirms the call establishment) is successfully sent and the call establishment is confirmed; thus ensuring that each step of the signaling process is executed normally and avoiding call failures caused by the absence of a certain message.

[0066] The data processing module is configured to clean the signaling messages, remove noise and invalid data, and filter out data packets irrelevant to signaling, such as RTP streams and HTTP requests; delete signaling messages with incorrect formats or missing key fields, such as the Call-ID missing in the Invite, remove duplicate signaling messages to avoid affecting the analysis results, and ensure the accuracy and integrity of the data; perform standardization processing on the signaling messages, including: unifying the timestamp format and field naming, such as the ISO8601 standard: YYYY-MM-DDTHH:MM:SSZ; unifying the From field as Caller, the To field as Callee, converting the IP address to the standard format, and unifying the port number to the integer type for subsequent analysis; convert the signaling messages into a format suitable for input to the signaling process model, such as arranging the signaling messages in chronological order to form time series data; extract the key features of the signaling messages, such as: message type, delay, SDP information, etc., and convert them into feature vectors; convert the data into a format suitable for input to the signaling process model, such as CSV, JSON, etc., for use by subsequent modules.

[0067] The feature extraction module is configured to extract the key features of the signaling messages, including: message types such as Invite, 100 Trying, 200 OK, etc., timestamps such as the sending or receiving time of the signaling messages, delays such as the delay from Invite to 100 Trying, and key fields in the SDP information, such as media type, codec format, IP address, and port; and calculate statistical features, including: average delay, that is, calculate the average delay of each stage in the signaling process, such as the average delay from Invite to 200 OK; maximum delay, that is, calculate the maximum delay in the signaling process to identify abnormal situations; and message loss rate, that is, calculate the loss rate of the signaling messages, such as the proportion of not receiving 100 Trying after Invite; then use feature importance screening to filter out the feature information that affects anomaly detection and remove redundant features to improve the efficiency and accuracy of the model.

[0068] Beneficial effects achieved by the above content: Through the above operations, key features can be efficiently extracted from signaling messages, providing data support for subsequent anomaly detection, fault analysis, and performance optimization, and improving the accuracy of signaling anomaly detection.

[0069] Specifically, the voice signaling tracking service system further includes:

[0070] A signaling tracking quality monitoring module for real-time monitoring of the operating parameters of the signaling tracking unit and making an anomaly determination on the signaling tracking quality;

[0071] Among them, the signaling tracking quality monitoring module performs the following steps:

[0072] Real-time monitoring of the operating parameters of the signaling tracking unit, where the operating parameters include the signaling capture rate (unit: number of messages per second), the message capture delay ratio (unit: dimensionless), and the storage bandwidth occupancy ratio (unit: dimensionless);

[0073] Real-time retrieval of the storage bandwidth occupancy ratio and the message capture delay ratio corresponding to each unit time;

[0074] Based on the available storage bandwidth occupancy ratio corresponding to the storage bandwidth occupancy ratio corresponding to each unit time;

[0075] Compare the available storage bandwidth occupancy ratio corresponding to each unit time with the message capture delay ratio;

[0076] When the available storage bandwidth occupancy ratio corresponding to each unit time is lower than the message capture delay ratio, the operating quality of the signaling tracking unit is determined using the signaling capture rate, the message capture delay ratio, and the storage bandwidth occupancy ratio.

[0077] The technical effects of the above technical solution are as follows: Real-time monitoring of operating parameters such as the signaling capture rate, the message capture delay ratio, and the storage bandwidth occupancy ratio of the signaling tracking unit can timely obtain the operating status information of the device. This enables operation and maintenance personnel to understand the working conditions of the signaling tracking unit at any time, whether it is the speed of signaling capture, message delay, or the usage status of storage resources, and provides a real-time and accurate data basis for subsequent analysis and decision-making. By real-time retrieving the storage bandwidth occupancy ratio corresponding to each unit time and calculating the available storage bandwidth occupancy ratio based on this, the usage of storage resources can be clearly understood. Comparing the available storage bandwidth occupancy ratio with the message capture delay ratio helps to discover the relationship between storage resources and message processing. For example, when the available storage bandwidth occupancy ratio is low, it may lead to an increase in message capture delay. Through this analysis, problems existing in the utilization of storage resources can be timely discovered, providing a basis for optimizing the storage resource configuration and improving resource utilization efficiency.

[0078] When the ratio of the available storage bandwidth per unit time is lower than the message capture delay ratio, the operation quality of the signaling tracking unit is determined by using the signaling capture rate, the message capture delay ratio, and the storage bandwidth ratio. This mechanism can issue early warnings in a timely manner when the storage resources are strained and may affect the message processing efficiency. The operation and maintenance personnel can, based on the comprehensive analysis of these parameters, discover potential problems such as possible performance degradation and faults of the signaling tracking unit in advance, and thus take corresponding measures for prevention and handling to avoid the expansion of problems and ensure the stable operation of the signaling tracking unit. Judging the operation quality of the signaling tracking unit by synthesizing multiple parameters such as the signaling capture rate, the message capture delay ratio, and the storage bandwidth ratio is more comprehensive and accurate than the evaluation with a single parameter. Different parameters reflect the operation conditions of the signaling tracking unit from different perspectives. For example, the signaling capture rate reflects the ability of the device to process signals, the message capture delay ratio reflects the timeliness of message processing, and the storage bandwidth ratio is related to the usage status of storage resources. By comprehensively analyzing these parameters, the operation quality of the signaling tracking unit can be accurately evaluated, providing strong support for the optimization and improvement of the device. By monitoring the operation parameters of the signaling tracking unit and judging its operation quality, potential problems can be discovered and solved in a timely manner to ensure the stable operation of the device. A stably operating signaling tracking unit is crucial for the entire communication system. It can ensure the accurate capture and processing of signals, thereby maintaining the normal operation of the communication system and improving the stability and reliability of the system.

[0079] Specifically, when the ratio of the available storage bandwidth per unit time is lower than the message capture delay ratio, the operation quality of the signaling tracking unit is determined by using the signaling capture rate, the message capture delay ratio, and the storage bandwidth ratio, including:

[0080] When the ratio of the available storage bandwidth per unit time is lower than the message capture delay ratio, the ratio of the available storage bandwidth ratio to the message capture delay ratio is processed to obtain reference ratio data;

[0081] The reference ratio data is processed by conversion to obtain the reference ratio corresponding to the converted reference ratio data;

[0082] Among them, the conversion processing model is as follows:

[0083]

[0084] Among them, s represents the reference ratio; k represents the reference ratio data;

[0085] Extract the change amplitude of the signaling capture rate between the signaling capture rate per unit time and the signaling capture rate corresponding to the previous unit time;

[0086] Obtain the signaling capture rate change rate by using the amplitude of the change in the signaling capture rate corresponding to each unit time and the signaling capture rate corresponding to its previous unit time, and the signaling capture rate corresponding to the current unit time;

[0087] Compare the signaling capture rate change rate with the reference ratio;

[0088] When the signaling capture rate change rate exceeds the reference ratio, it is determined that there is a risk of abnormal operation of the signaling tracking unit, and an early warning of abnormal operation quality risk is given.

[0089] The technical effects of the above technical solution are as follows: By real-time monitoring the operating parameters of the signaling tracking unit, including the signaling capture rate, message capture delay ratio, storage bandwidth occupancy ratio, etc., and conducting comprehensive analysis and comparison, it is possible to timely detect whether there is a risk of abnormal operation of the signaling tracking unit. When the signaling capture rate change rate exceeds the reference ratio, it indicates that there may be an operation problem, so an early warning can be issued in a timely manner for relevant personnel to take measures for handling to avoid the problem from deteriorating further. The ratio of the available storage bandwidth occupancy ratio to the message capture delay ratio is processed, and the change in the signaling capture rate is quantitatively analyzed to obtain the signaling capture rate change rate, and the operation quality is determined by comparing the two. This quantitative evaluation method can more accurately reflect the actual operating conditions of the signaling tracking unit. Compared with single-parameter monitoring, it more comprehensively and objectively evaluates the operation quality of the signaling tracking unit, reducing the possibility of misjudgment and missed judgment.

[0090] Meanwhile, by comprehensively considering the change rate of the signaling capture rate and the reference ratio, a multi-dimensional basis for anomaly determination is provided. The change rate of the signaling capture rate reflects the dynamic change of the signaling capture rate over time, while the reference ratio is obtained based on the storage bandwidth occupancy and the message capture delay ratio, representing the relationship between storage resources and message processing delay. By comparing these two metrics, the operating status of the signaling tracking unit can be analyzed more comprehensively, avoiding misjudgments that may occur due to a single metric judgment, thereby improving the accuracy of anomaly detection. At the same time, it can effectively improve the timeliness and sensitivity of signaling tracking anomaly determination. When the change in the signaling capture rate is abnormal and is associated with the storage resources and message delay situation, potential system failures or performance degradation issues can be detected in a timely manner, and intervention can be carried out in advance to ensure the stable operation of the signaling tracking unit, thereby enhancing the stability of the entire communication system. On the other hand, the calculation of the reference ratio involves the storage bandwidth occupancy. When an anomaly is detected by comparing it with the change rate of the signaling capture rate, it can prompt the operation and maintenance personnel to pay attention to the usage of storage resources. If the increase in message capture delay and the abnormal change in the signaling capture rate are caused by insufficient storage resources, the storage resource configuration can be adjusted specifically to optimize the resource utilization efficiency and avoid the impact of resource waste or insufficiency on system performance. At the same time, by real-time monitoring parameters such as the signaling capture rate, storage bandwidth occupancy, and message capture delay ratio, and calculating the change rate of the signaling capture rate and the reference ratio accordingly, the change in the operating status of the signaling tracking unit can be reflected in real time. This real-time and dynamic nature enables the system to quickly respond to anomalies during operation, issue early warnings in a timely manner, and improve the system's adaptability and processing efficiency to changes.

[0091] The fault handling unit includes a model construction module, a signaling analysis module, and a tracking service module;

[0092] The model construction module is configured to build a signaling process model for normal calls based on historical signaling messages, define anomaly detection rules and signaling fault types based on expert experience, detect anomalies in the process using a rule engine, and build a decision tree model to classify fault types, historical signaling messages, and adjustment strategies; The model construction module includes:

[0093] The process definition module is configured to define a standard signaling process model based on historical signaling messages of normal calls, such as: Invite→100 Trying→180 Ringing→200OK→ACK, and determine the normal behavior of each step, including: time range, such as: the delay from Invite to 100 Trying should be less than 500ms; message format, such as: the Invite message must contain fields such as Call-ID, From, and To.

[0094] The rule definition module is configured to define the anomaly detection rules in the signaling process model based on expert experience, forming a rule engine pattern, such as: not receiving 100 Trying after Invite; missing SDP information in 200 OK; excessive delay from 180 Ringing to 200 OK, enabling the signaling process model to implement anomaly detection rules based on the use of a rule engine.

[0095] The model training module is configured to train the signaling process model based on the training set of historical signaling messages, enabling the signaling process model to accurately describe the normal signaling process; verify the accuracy of the signaling process model based on the test set of historical signaling messages, enabling the signaling process model to correctly identify the normal signaling process; specifically, by extracting the training set and test set from historical signaling messages, ensuring that the data set contains the signaling messages of normal calls and annotating the normal behavior of each step; selecting a rule engine to describe the signaling process, using the training set to train the signaling process model to ensure that the model can accurately describe the normal signaling process; and during the training process, adjusting the model parameters to optimize its performance; then using the test set to verify the accuracy of the model and calculating evaluation metrics, such as: accuracy rate, recall rate, F1 score; and storing the defined signaling process model and rules as a configuration file or data structure for use by subsequent modules.

[0096] The fault definition module is configured to define the signaling fault types based on expert experience and provide a detailed description, possible causes, and corresponding adjustment strategies for each fault type; the fault types are such as: signaling loss: not receiving 100 Trying for Invite; format error: missing SDP information in 200 OK; delay anomaly: excessive delay from 180 Ringing to 200 OK.

[0097] The fault classification module is configured to classify the fault types, historical signaling messages, and adjustment strategies based on the key features of historical signaling messages, enabling the decision tree model to correctly classify the fault types and adjustment strategies; specifically, based on the signaling fault types defined by expert experience, combined with the key features of historical signaling messages, annotate the fault types and adjustment strategies in the training set and test set of its historical signaling messages; use the training set to train the decision tree model to ensure that the model can correctly classify the fault types and adjustment strategies; and use the test set to verify the accuracy of its model; deploy the trained model to the voice signaling tracking service system, enabling the decision tree model to classify the fault types and adjustment strategies in a timely manner according to signaling anomalies.

[0098] The signaling analysis module is configured to analyze real-time signaling messages based on the signaling process model, discover anomalies and potential faults in a timely manner; and set an alarm mechanism to trigger an alarm when an anomaly is detected; the signaling analysis module includes:

[0099] The model integration module is configured to integrate the signaling process model into the existing system based on the message queue. For example, select the Kafka message queue system, and configure the topics and partitions of the message queue. The topics can be divided according to the signaling type or business scenario. For example, different topics such as registration, call, and release can be created according to the signaling type; and an appropriate number of partitions can be configured for each topic to ensure the orderliness and parallel processing ability of the messages. For example, partition according to the source or target ID of the signaling to ensure that all messages of the same signaling flow enter the same partition; ensure the orderliness of the same signaling flow through the partition mechanism and the message key Key. For example, use the session ID of the signaling as the message key to ensure that messages of the same session enter the same partition and are processed in order; and configure the persistence mechanism of the message queue. For example, the ACK mechanism and retry strategy of Kafka to ensure that messages are not lost, thereby ensuring the orderliness and reliability of the signaling messages; use the message queue as the transmission and buffer layer of the signaling messages. After capturing the real-time signaling messages, serialize the signaling messages and send them to the message queue so that the signaling process can capture the signaling messages from it.

[0100] The signaling analysis module is configured to obtain the signaling messages from the message queue in real time for the signaling process model and perform real-time analysis using the rule engine to analyze whether there are any abnormalities in the voice signaling messages during the current call.

[0101] The alarm definition module is configured to define alarm rules based on the abnormal detection results of historical signaling messages. When the signaling process model analyzes that there are abnormalities in the real-time signaling messages, trigger the alarm rules and synchronously feedback to the management department; the alarm rules include: signaling loss alarm and delay exception alarm. The signaling loss alarm rule is defined as: when it is detected in the signaling process model that a certain signaling message does not arrive within the expected time, trigger the alarm; implementation method: through the time window mechanism, monitor the arrival time of the signaling messages, and trigger the alarm if it times out. The delay exception alarm rule is defined as: when the transmission delay of the signaling message exceeds the preset threshold, trigger the alarm; implementation method: calculate the time difference between the sending time and the receiving time of the signaling message, and trigger the alarm if it exceeds the threshold. The alarm method can send the alarm information to the management department through email, SMS, instant messaging tools such as DingTalk, Slack or the alarm platform. After receiving the alarm, the management department takes corresponding handling measures according to the alarm level and type, such as: troubleshooting problems, optimizing the system or notifying relevant personnel; and store the alarm information in the database or log system for subsequent analysis and traceability.

[0102] The tracking service module is configured to locate the cause of the failure based on the abnormal detection results and obtain the adjustment strategy of the same failure type from the decision tree model for dynamic adjustment; the tracking service module includes:

[0103] The fault location module is configured to identify abnormal points in the network or signaling process based on the abnormal detection results, such as high latency, packet loss, signaling loss, etc.; combine network topology analysis to analyze the network nodes, links or devices where the abnormal points are located to narrow down the fault scope; and trace the transmission path of signaling messages through signaling logs to identify possible fault links, such as equipment failures, configuration errors, and locate the fault causes based on the analysis results, including network congestion, equipment failures, configuration errors.

[0104] The fault matching module is configured to match the located fault causes with the fault types in the decision tree model to obtain adjustment strategies; and optimize the adjustment strategies according to the current network status and the severity of the fault to ensure their applicability and effectiveness.

[0105] The dynamic repair module is configured to execute the adjustment strategies using the Ansible automation tool based on the obtained adjustment strategies to dynamically adjust the network configuration or signaling process; including the following operations: adjusting the network device configuration, such as routing, bandwidth limitation; optimizing the signaling process, such as redirecting the signaling path, adjusting the signaling priority; restarting or replacing the faulty equipment; and recording the execution process and results of the adjustment strategies for subsequent analysis and optimization.

[0106] The repair verification module is configured to continuously monitor the network status and signaling process using monitoring tools, such as Prometheus, to verify whether the adjustment strategies are effective; and monitor the transmission of signaling messages to confirm whether the faults are completely resolved. If not, re-locate the fault causes and adjust the strategies to ensure that the system returns to normal.

[0107] The repair feedback module is configured to collect the execution results of the adjustment strategies, including the fault recovery time, network status changes, etc.; optimize the decision tree model and adjustment strategies based on the execution results to improve the accuracy of fault matching and strategy recommendation; and optimize the adjustment strategies according to the actual effects to enhance their applicability and execution efficiency.

[0108] As shown in the Figure 2 - Appendix Figure 4 following: In one embodiment, the voice signaling tracking service includes the following processes:

[0109] (1) The terminal contained in Server A calls the terminal contained in Server B, and Server A sends an Invite request to the gatekeeper.

[0110] (2) The gatekeeper modifies the address in the Contact field of the invite request message header initiated by Server A to its own Server B side address and forwards the Invite request to the Server B address.

[0111] Note: The invite message contains the audio and video formats supported by the terminal on the server A side and the audio and video RTP ports of the calling party used in this call;

[0112] (3) Server B sends a 100 Trying response message in the call processing to the gatekeeper;

[0113] (4) The gatekeeper sends a 100 Trying response message in the call processing to server A;

[0114] (5) Server B instructs the called user to ring. After the user rings, it sends a 180 Ringing ringing message to the gatekeeper;

[0115] (6) The gatekeeper forwards the ringing message of the called user to server A;

[0116] (7) The called terminal user on the server B side picks up the phone, and server B returns a 200 OK response indicating successful connection to the gatekeeper;

[0117] (8) The gatekeeper forwards this success indication (200 OK) to server A;

[0118] (9) After receiving the message, server A sends an ACK message to the gatekeeper for confirmation;

[0119] (10) The gatekeeper forwards the ACK confirmation message to terminal proxy B;

[0120] (11) A communication connection is established between the calling and called users, and the call starts;

[0121] It should be noted in the above process that: The terminal on the server B side detects the audio and video formats supported in the invite message of the calling party, compares them with the formats it supports, and then sends a 200 OK packet containing the finally negotiated audio and video formats and its own audio and video RTP ports to the boundary; Secondly, both parties use the negotiated audio and video coding formats and RTP ports for communication. The gatekeeper must obtain the RTP ports listened to by both servers during the SIP signaling negotiation process and open the forwarding channels for these ports.

[0122] Beneficial effects achieved by the above content: By capturing signaling messages in real time and analyzing them based on the signaling process model, anomalies and potential faults can be quickly detected, reducing the latency of fault detection and improving the system's response speed. Secondly, an adjustment strategy corresponding to the fault type is obtained from the decision tree model to dynamically adjust the network configuration or signaling process, thereby reducing manual intervention, improving the efficiency of fault handling, and avoiding affecting call quality due to untimely fault handling. Finally, by setting an alarm mechanism, relevant personnel are notified in a timely manner when an anomaly is detected, ensuring that faults can be discovered and handled in a timely manner, preventing problems from expanding, and thus enhancing the real-time performance, accuracy, and automation level of the voice signaling tracking service system, optimizing network performance, reducing operation and maintenance costs, and improving the quality of mobile communication services.

[0123] Working principle: By capturing and recording signaling messages during a call in real time, analyzing the real-time signaling messages based on the signaling process model, anomalies and potential faults are discovered in a timely manner; an alarm mechanism is set to trigger an alarm when an anomaly is detected; the cause of the fault is located based on the anomaly detection result, and an adjustment strategy for the same fault type is obtained from the decision tree model for dynamic adjustment, thereby improving the quality of mobile communication services.

[0124] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "having" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article or device.

[0125] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made therein without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A voice signaling tracking service system for tracking calls, characterized in that: The system comprises: a signaling tracking unit and a fault processing unit, wherein the fault processing unit comprises a model building module, a signaling analysis module and a tracking service module; The signaling tracking unit is configured to capture and record signaling messages during a call in real time, and to make an abnormality determination on the signaling tracking quality based on the operating parameters of the signaling tracking unit; The model building module is configured to build a signaling process model of normal calls based on historical signaling messages, define anomaly detection rules and signaling fault types based on expert experience, use a rule engine to detect anomalies in the process, and build a decision tree model to classify fault types, historical signaling messages, and adjustment strategies; The signaling analysis module is configured to analyze real-time signaling messages based on the signaling process model to promptly detect anomalies and potential faults; and to set an alarm mechanism to trigger an alarm when an anomaly is detected; The tracking service module is configured to locate the cause of the fault based on the abnormal detection result, and obtain the adjustment strategy of the same fault type from the decision tree model for dynamic adjustment.

2. The voice signaling tracking service system for tracking calls according to claim 1, characterized in that: The voice signaling tracking service system further includes: The signaling tracking quality monitoring module is used to monitor the operating parameters of the signaling tracking unit in real time and make abnormal judgments on the signaling tracking quality; The signaling tracking quality monitoring module performs the following steps: Real-time monitoring of operating parameters of the signaling tracking unit, wherein the operating parameters include signaling capture rate, message capture delay ratio and storage bandwidth ratio; Retrieve the storage bandwidth ratio and message capture delay ratio corresponding to each unit time in real time; The available storage bandwidth ratio corresponding to the storage bandwidth ratio per unit time; Compare the available storage bandwidth ratio corresponding to each unit time with the message capture delay ratio; When the available storage bandwidth ratio corresponding to each unit time is lower than the message capture delay ratio, the signaling capture rate, the message capture delay ratio and the storage bandwidth ratio are used to determine the operation quality of the signaling tracking unit.

3. The voice signaling tracking service system for tracking calls according to claim 2, characterized in that: When the available storage bandwidth ratio corresponding to each unit time is lower than the message capture delay ratio, the signaling capture rate, the message capture delay ratio and the storage bandwidth ratio are used to determine the operation quality of the signaling tracking unit, including: When the available storage bandwidth ratio corresponding to each unit time is lower than the message capture delay ratio, the available storage bandwidth ratio is processed with the message capture delay ratio to obtain reference ratio data; Performing conversion processing on the reference ratio data to obtain a reference ratio corresponding to the converted reference ratio data; Extracting the signaling capture rate variation amplitude between the signaling capture rate corresponding to each unit time and the signaling capture rate corresponding to the previous unit time; Obtain the signaling capture rate change rate by using the signaling capture rate change amplitude between the signaling capture rate corresponding to each unit time and the signaling capture rate corresponding to the previous unit time and the signaling capture rate corresponding to the current unit time; Comparing the signaling capture rate change rate with the reference ratio; When the signaling capture rate change rate exceeds the reference ratio, it is determined that the operation of the signaling tracking unit has an operation abnormality risk, and an operation quality abnormality risk warning is issued.

4. The voice signaling tracking service system for tracking calls according to claim 1, characterized in that: The model building module comprises: A process definition module is configured to define a standard signaling process model based on historical signaling messages of normal calls and determine the normal behavior of each step, including: time range, message format; A rule definition module is configured to define anomaly detection rules in the signaling process model based on expert experience, forming a rule engine mode, so that the signaling process model implements anomaly detection rules based on the use of the rule engine; The model training module is configured to train the signaling process model based on a training set of historical signaling messages so that the signaling process model can accurately describe the normal signaling process; and to verify the accuracy of the signaling process model based on a test set of historical signaling messages so that the signaling process model can correctly identify the normal signaling process.

5. The voice signaling tracking service system for tracking calls according to claim 4, characterized in that: The model building module further includes: A fault definition module, configured to define signaling fault types based on expert experience, and provide a detailed description and possible causes for each fault type as well as corresponding adjustment strategies; The fault classification module is configured to classify the fault type, the historical signaling message and the adjustment strategy based on the key features of the historical signaling message, so that the decision tree model can correctly classify the fault type and the adjustment strategy.

6. The voice signaling tracking service system for tracking calls according to claim 1, characterized in that: The signaling analysis module comprises: The model integration module is configured to integrate the signaling process model into the existing system based on the message queue, and configure the topic and partition of the message queue; the message queue is used as the transmission and buffer layer of the signaling message, and after capturing the real-time signaling message, the signaling message is sent to the message queue; The signaling analysis module is configured as a signaling process model to obtain signaling messages from the message queue in real time and use the rule engine to perform real-time analysis; The alarm definition module is configured to define alarm rules based on the abnormal detection results of historical signaling messages. When the signaling process model analyzes that there are abnormalities in real-time signaling messages, the alarm rules are triggered and synchronously fed back to the management department; the alarm rules include: signaling loss alarm and delay abnormality alarm.

7. The voice signaling tracking service system for tracking calls according to claim 1, characterized in that: The tracking service module includes: The fault location module is configured to locate the cause of the fault based on the anomaly detection results, combined with the network topology and signaling logs, including: network congestion, equipment failure, and configuration error; A fault matching module, configured to match the fault cause located with the fault type in the decision tree model to obtain an adjustment strategy; The dynamic repair module is configured based on the acquired adjustment policy, and uses the Ansible automation tool to execute the adjustment policy to dynamically adjust the network configuration or signaling process.

8. The voice signaling tracking service system for tracking calls according to claim 7, characterized in that: The tracking service module further includes: Repair verification module, configured to continuously monitor network status and signaling processes, verify whether the adjustment strategy is effective, and ensure that the system is restored to normal; A repair feedback module is configured to collect the execution results of the adjustment strategy and optimize the decision tree model and the adjustment strategy.

9. The voice signaling tracking service system for tracking calls according to claim 1, characterized in that: The signaling tracking unit includes: The signaling collection module is used to collect historical signaling messages, and use the signaling tracking tool to capture and record all signaling messages during the call to form a signaling log; the signaling messages include: Invite initiates a session request, 100 Trying request is being processed, 180 Ringing the called party is ringing, 200 OK the called party accepts the request and establishes the call, and ACK the calling party confirms the call establishment; The process checking module is configured to check whether the signaling process is complete based on the signaling log.

10. The voice signaling tracking service system for tracking calls according to claim 9, characterized in that: The signaling tracking unit further includes: The data processing module is configured to clean the signaling messages, remove noise and invalid data; standardize the signaling messages, including: unifying the timestamp format and field naming; converting the signaling messages into a format suitable for the signaling process model input; The feature extraction module is configured to extract key features of the signaling message, including: message type, timestamp, delay, and SDP information, calculate statistical features, including: average delay, maximum delay, and message loss rate, and use feature importance to filter feature information that has an impact on anomaly detection.

Images