TWI731636B - System and method for monitoring broadband loop cable quality - Google Patents

Abstract

A system and a method for monitoring broadband loop cable quality are provided. The system and the method can quickly and periodically search for possible malfunctions in cable apparatuses of a large number of broadband circuits by using methods such as threshold settings of network management parameters related to broadband equipment, degrees of deterioration, and simple loop malfunction prediction model. The system and the method further generate quantized quality indicators of the cable apparatuses by using double end line tests and a broadband loop malfunction diagnosis model in order to quickly and precisely predict deterioration malfunctions of cable apparatuses and improve broadband service quality.

Description

Broadband loop cable quality monitoring system and method

The present invention relates to a broadband loop cable quality monitoring system and method, especially after using daily relevant network management parameters as the characteristic value input, the field experts with broadband circuits and rich experience in inspection and repair are responsible for the labeling of obstacle classification. Then apply the fast screening method of obstacle circuit, Double End Line Test (DELT) and artificial intelligence (AI) machine learning technology to provide a system and method for broadband loop cable quality monitoring.

With the vigorous development of Over-The-Top (OTT) services, the demand for network bandwidth is increasing day by day. In the face of the difficulty of fiber access to homes, VDSL2 dedicated lines installed with copper cables still account for the largest proportion of broadband installations. Among them, VDSL is the abbreviation of Very-high-bit-rate Digital Subscriber Line, and VDSL2 represents the second generation technology. However, copper cables have distance limitations and may have obstacles such as oxidation, corrosion, poor connections, T-connections, crosstalk, and noise. These physical layer obstacles will affect the performance of the service layer.

VDSL2 DSLAM equipment conforming to ITU-T G.992.3, G.993.2 and G.997.1 can perform DELT test when the equipment is connected, and the test result can determine the equivalent length of the loop, Information such as loop quality, background noise, and crosstalk interference. The aforementioned ITU is the abbreviation of the International Telecommunication Union, and the DSLAM is the abbreviation of the Digital Subscriber Line Access Multiplexer. As shown in Figure 1, in the VDSL2 loop, the cable exits from the DSLAM equipment connector 10 of the transfer box and reaches the user through the distribution cable 20 and the distribution box cable 30. Among them, the cable itself or the connection point may cause obstacles. This leads to a decrease in communication quality. Therefore, it is hoped that the quality of the above cable equipment can be monitored, and an alarm can be raised when the deterioration reaches a certain standard for the maintenance unit to replace or repair.

The cable equipment is composed of a variety of different numbers of circuits. For example, the distribution cable 20 has 100 pairs; the distribution box cable 30 has 10-100 pairs; the DSLAM device connector 10 of the transfer box has 24 pairs. At present, VDSL2 loop tests are implemented for a single circuit of a single user, and the analysis results are qualitative, which must be analyzed and interpreted by experts and cannot be converted into quantitative quality index values. In addition, because the VDSL2 single-channel DELT test takes about 30 seconds, and the total amount of VDSL2 circuits is huge, it is impossible to test all the circuits every day, and it is not a good method in terms of labor cost and time cost. . It can be seen that there are still many shortcomings in the above-mentioned customary methods, which are not good designs and need to be improved urgently.

The purpose of the present invention is to provide a broadband loop cable quality monitoring system and method, using broadband equipment related network management parameters and broadband loop obstacle diagnosis model to predict VDSL2 circuit obstacles, and then classify poor quality related circuits into distribution cables and branch lines Box cable and DSLAM equipment connectors, and pre-processed before users report a large number of obstacles, and timely cable replacement and maintenance To achieve the goal of quickly and accurately predicting the degradation obstacles of broadband loop cables, reducing the reporting rate of broadband obstacles and improving the quality of broadband circuit services.

In order to achieve the above-mentioned purpose of the invention, the present invention provides a broadband loop cable quality monitoring system and method, which utilizes the daily relevant network management parameters of the DSLAM equipment provided to provide an original method and analysis method to provide the broadband loop cable and related equipment joint quality Analyze the results.

As shown in Figure 2, the broadband loop cable quality monitoring system consists of at least six modules, namely the network management performance parameter collection module 100, the obstacle circuit quick screening module 200, the obstacle circuit test module 300, and the loop obstacle diagnosis module. Group 400, quality index conversion module 500 and cable obstruction warning module 600.

The quality monitoring process of the distribution cables, junction box cables and equipment connectors performed by the above six modules includes the following steps:

(1) The network management performance parameter collection module 100 collects the network management performance parameters of all VDSL2 circuits every day.

(2) There are three methods for screening possible obstacle circuits, and they can be used at the same time:

(2a) The average value of each network management performance parameter plus or minus 1~2 standard deviations is used as the threshold value. If the threshold is exceeded, it is regarded as an obstacle. The network management performance parameters include LOL (Loss of Link), CV (Code Violation), ES (Error Second), SES (Severely Error Second), FEC (Forward Error Correction) and SNRM (Signal to Noise Ratio Margin).

(2b) The achievable speed and electrical length changes in the network management performance parameters exceeding the set value are considered as obstacles.

(2c) Use network management performance parameters to establish a simple broadband loop obstacle classification model to quickly predict whether the user circuit part of the loop is obstructed.

(3) After screening the suspicious obstacle circuits, first carry out the classification of the cable distribution, the distribution box cable and the DSLAM equipment connector and the calculation of the obstacle ratio, and then further conduct the DELT test for each circuit of the cable equipment with a high obstacle ratio to greatly reduce The number of DELT tests.

(4) Establish a broadband loop fault diagnosis model based on the historical data of circuit link (Link), frequency spectrum (Carrier) and performance parameter (Performance Parameter, PM for short), so as to determine whether the aforementioned circuit (3) is a loop fault and generate its Obstacle probability value.

(5) The quality index conversion module 500 recalculates the distribution cable, the distribution box cable and the DSLAM equipment connector obstacle ratio according to the probability value of the loop obstacle diagnosis model after the DELT test, and converts it into a quality index value.

(6) The cable obstruction alarm module 600 sets the quality indicator alarm threshold value, and provides alarm messages and a list of severely deteriorated cables, junction box cables and DSLAM equipment connectors to be repaired.

10: DSLAM equipment connector of the transfer box

20: VDSL2 copper cable broadband loop distribution cable

30: Distribution box cable for VDSL2 copper cable broadband loop

100: Network management performance parameter collection module

110: Screening test module

120: Diagnostic alarm module

200: Obstacle circuit quick screen module

210: Network management performance parameter threshold obstacle judgment module

220: Achievable speed and electrical length degradation extent judgment module

230: Simple circuit obstacle classification model obstacle prediction module

300: Obstacle circuit test module

400: Circuit failure diagnosis module

500: Quality Index Conversion Module

600: Cable Obstruction Alarm Module

S700~S720: process steps

Figure 1 shows the topology of the VDSL2 copper cable broadband loop.

Figure 2 is a schematic block diagram of a broadband loop cable quality monitoring system according to an embodiment of the present invention.

Figure 3 is a flowchart of a method for monitoring the quality of a broadband loop cable according to an embodiment of the present invention.

The following specific examples illustrate the implementation of the present invention. Those familiar with the art can easily understand the other advantages and effects of the present invention from the content disclosed in this specification.

The present invention uses broadband equipment related network management parameters and broadband loop obstacle diagnosis model to predict VDSL2 circuit obstacles, and then classifies poor quality related circuits into distribution cables, junction box cables and related equipment connectors, and converts the probability of obstacles In order to provide quality indicators and raise alarms, it can be processed in advance before a large number of users report, and the cables can be replaced and maintained in time.

As shown in FIG. 2, the broadband loop cable quality monitoring system according to an embodiment of the present invention includes a network management performance parameter collection module 100, a screening test module 110, and a diagnostic alarm module 120. The network management performance parameter collection module 100 is used to obtain the network management performance parameters and basic data of the broadband loop. This basic information includes the structure of the entire broadband loop. The entire broadband loop includes multiple cable devices, and each cable device can be a distribution cable, a junction box cable, or a DSLAM device connector. Each distribution box cable includes the distribution box itself and multiple cables between the distribution box and downstream users. Circuit refers to the copper cable line between the transfer box and each user in the broadband loop, and each user has a corresponding circuit. The above basic information includes which circuits each cable device passes through, as well as the distribution box cable, distribution cable and DSLAM device connector to which each circuit belongs. The screening test module 110 is used to select a part of all circuits of the broadband loop according to the network management performance parameters and basic data, perform DELT test on each circuit in the part, and generate test data. The diagnostic alarm module 120 is used for outputting a list of the cable equipment of the broadband loop to be repaired according to the test data.

As shown in FIG. 2, the screening test module 110 includes an obstacle circuit quick screening module 200 and an obstacle circuit test module 300. The diagnostic alarm module 120 includes a circuit failure diagnostic module 400, a quality index conversion module 500, and a cable failure alarm module 600. Among them, the obstacle circuit quick screening module 200 includes three sub-modules, namely the network management performance parameter threshold obstacle judgment module 210, the reachable speed and electrical length degradation extent judgment module 220, and the simple circuit obstacle classification model obstacle prediction module. Module 230. Each of the above modules can be software (such as computer programs) or hardware (such as Such as electronic devices such as computers or servers that can execute computer programs) or firmware. There can be multiple modules included in the same software or the same hardware or the same firmware or combination.

The main function of the network management performance parameter collection module 100 in Figure 2 is to obtain the network management performance parameters of all circuits of the broadband loop from the network management equipment daily and update the basic information of the distribution cable, the distribution box cable and the DSLAM equipment connector of the transfer box. The types of network management performance parameters are determined by experts with inspection and repair experience, and serve as the input data required by the obstacle circuit quick screening module 200 and the characteristic input value of the obstacle prediction module 230 of the simple circuit obstacle classification model. These network management performance parameters have a total of 21 items, including DSLAM boot duration, VDSL user terminal equipment (remote VDSL terminating unit, referred to as VTUR) connection duration, uplink and downlink link second loss (Loss of Link Second, referred to as LOLS), power Loss of Power (LPR), Loss of Link (LOL), Code Violation (CV), Error Second (ES), Severely Error Second , SES for short), Forward Error Correction (FEC for short), Signal to Noise Ratio Margin (SNRM for short), reachable rate, and electrical length.

The obstacle circuit quick screening module 200 in Fig. 2 has a main function of quickly screening a large number of possible obstacle circuits. The quality of the VDSL2 circuit is more accurate with the DELT test, but a DELT test takes about 30 seconds. Due to the huge number of VDSL2 circuits, it is impossible to complete all circuit tests every day. Therefore, the present invention adopts three methods to quickly screen the circuits with possible obstacles. These three methods correspond to the network management performance parameter threshold obstacle judgment module 210, the reachable rate and electrical length degradation extent judgment module 220, and the simple circuit obstacle classification model obstacle prediction module. Judgment module 230. Use the three judgment modules 210, 220, and 230 to filter out possible obstacles from all circuits of the broadband loop according to the performance parameters of the network management, and then select the three judgment modules to filter Select the union of the results, and then classify them into each cable equipment, and then use the DELT test to retest the cable equipment with a high percentage of obstacles.

The network management performance parameter threshold obstacle judgment module 210 in Figure 2 has a function of judging criteria to screen obstacle circuits based on the pros and cons of important network management performance parameters. For example, LOL, CV, ES, SES, FEC, SNRM are used as screening targets, and the (average value + standard deviation) of all network management performance parameters of all VDSL2 circuits in a day is used as the screening threshold value (SNRM threshold value is the average value-standard Poor X2). When the value of any network management performance parameter of the circuit read daily exceeds the threshold value, the circuit to which the network management performance parameter belongs is deemed to be a possible obstacle. The above-mentioned excess can be greater or less than, depending on the nature of the various network management performance parameters.

The achievable rate and electrical length degradation degree judgment module 220 in Figure 2 mainly depends on whether the ratio of (current achievable rate/original achievable rate) of each circuit of the broadband loop is lower than the threshold setting value (eg 0.6) , The other is whether (current electrical length-original electrical length) exceeds the threshold setting value (eg 50 meters). When the achievable rate and electrical length change of any circuit read daily exceed the threshold value, it is considered as a possible obstacle.

In another embodiment, the achievable rate and electrical length degradation extent judgment module 220 can use other network management performance parameters to screen out possible impediment circuits. The reachable rate and electrical length degradation extent judgment module 220 can calculate the degradation extent of at least one network management performance parameter of each circuit of the broadband loop compared with the corresponding original value. The extent of degradation is the difference or ratio between the network management performance parameter and the corresponding original value. If each degradation amplitude of any circuit exceeds the corresponding threshold value, the circuit is a possible obstacle circuit.

The obstacle prediction module 230 of the simple circuit obstacle classification model shown in Fig. 2 has the function of establishing a first circuit obstacle prediction model, and quickly predicting circuit obstacles based on this model. The sample data of the model is to retrieve the DELT test history record of the broadband copper cable loop on weekdays. The inspection experts refer to channel gain (also known as HLOG) and Quiet Line Noise (Quiet Line Noise). QLN for short), Signal-to-Noise Ratio (SNR for short) spectrum graphics, and network management performance parameter values such as LOL, LPR, CV, ES, etc., manual labeling according to the classification of "loop obstacles" and "non-loop obstacles" . After labeling, only take DSLAM boot duration, VTUR connection duration, uplink and downlink LOL, LOLS, LPR, CV, ES, SES, FEC, SNRM and other important network management performance parameters as characteristic input values, and perform broadband copper cable loop Obstacle classification model training and evaluation. The models are established and evaluated using AI machine learning algorithms such as Decision Tree, Random Forest, Ada Boost, and Support Vector Machine. Accuracy, Recall and Precision. The resulting model can predict whether the circuit is obstructed and obtain its probability of obstruction. The simple circuit obstacle classification model obstacle prediction module 230 can input the above-mentioned network management performance parameters of each circuit of the broadband loop into the above-mentioned model to generate the obstacle probability value of each circuit. If the failure probability value of any circuit is greater than the set threshold value, the circuit is a possible obstacle circuit.

In another embodiment, the obstacle circuit quick screening module 200 can be simplified, that is, the network management performance parameter threshold obstacle judgment module 210, the reachable rate and electrical length degradation extent judgment module 220, and the simple circuit obstacle classification model are omitted. At least one of the judging modules 230.

The obstacle circuit test module 300 in Figure 2 has a function for each possible obstacle circuit screened out by the obstacle circuit quick screening module 200, according to the basic data of the broadband circuit to carry out cable distribution, junction box cables and DSLAM equipment connectors. The calculation of the obstacle ratio of each cable equipment. The above classification is to count the number of possible obstacle circuits under each cable device. For each cable device, the above-mentioned obstacle ratio is the ratio of the number of possible obstacle circuits passing through the cable device to the total number of circuits passing through the cable device. The obstacle circuit test module 300 performs DELT tests on each circuit of each cable device whose obstacle ratio is greater than the set threshold value, and then obtains more detailed link (Link), frequency spectrum (Carrier) and performance parameter (PM) tests. The data is used for the circuit fault diagnosis module 400 to perform fault prediction.

The circuit obstacle diagnosis module 400 in Figure 2 has a function of establishing a broadband circuit obstacle diagnosis model, and predicting circuit obstacles based on this model. The sample data of the model is the same as the obstacle prediction module 230 of the simple circuit obstacle classification model. The characteristic input values are the uplink and downlink SNRM, 300KHz signal attenuation (300K_Decline), QLN background noise value, QLN unstable noise ratio, and uplink and downlink signals. Attenuation, VDSL Transceiver Unit (VTUC) link duration, DSLAM startup (Power ON) duration, LOL, LOLS, LPR, ES, SES, CV, FEC and other performance indicators and HLOG, QLN, SNR Spectrum data. Similarly, AI machine learning algorithms such as decision trees, random forests, propulsion methods, vector support machines, etc. are used to train and evaluate broadband copper cable circuit obstacle diagnosis models to generate models that can predict whether the circuit is malfunctioning and its probability value. The loop failure diagnosis module 400 inputs the test data of each circuit that has passed the DELT test into the above model to generate the failure probability value of each of the above circuits.

The quality index conversion module 500 in Figure 2 has the function of recalculating the obstruction ratios of the distribution cables, junction box cables, and DSLAM equipment connectors based on the results of the loop obstruction diagnosis module 400, and converting them into quality indicators (1 ~10). In more detail, if the failure probability value predicted by the circuit failure diagnosis module 400 for a certain circuit is greater than the set threshold value, the quality index conversion module 500 determines that the circuit is a failure circuit. The quality index conversion module 500 classifies each obstacle circuit among the multiple cable devices of the broadband circuit according to the basic data of the broadband circuit, and calculates the obstacle ratio of each cable device. The above classification is to count the number of obstacle circuits under each cable device. For each cable device, the above-mentioned obstacle ratio is the ratio of the number of obstacle circuits passing through the cable device to the total number of circuits passing through the cable device. The quality index conversion module 500 outputs the quality index of each cable device, and the quality index is linearly inversely proportional to the obstacle ratio of the cable device to which it belongs.

The cable obstruction alarm module 600 shown in Figure 2, its function is to issue alarm messages for the distribution cables, junction box cables and DSLAM equipment connectors whose quality index scores are lower than the set threshold. And output a list of obstacles to be repaired for related maintenance personnel to inquire and repair. The above list to be repaired includes every cable equipment whose quality index is lower than the set threshold.

In another embodiment, the quality index of each cable device is the obstacle ratio of the same cable device, and the list to be repaired includes each cable device whose quality index is higher than the set threshold. The cable obstruction alarm module 600 will issue an alarm message for the cable equipment whose quality index is higher than the set threshold.

Figure 3 is a flowchart of a method for monitoring the quality of a broadband loop cable according to an embodiment of the present invention. First, in step S700, the network management performance parameters and basic data of the broadband loop are obtained. In step S710, a part of the circuits in all the circuits of the broadband loop is selected according to the performance parameters of the network management and the basic data, and the line double-ended test is performed on each circuit in the part of the circuits, and test data is generated. In step S720, a list of the cable equipment of the broadband loop to be repaired is output according to the test data.

The broadband loop cable quality monitoring method in Figure 3 can be implemented by the broadband loop cable quality monitoring system in Figure 2. Among them, steps S700, S710, and S720 are executed by the network management performance parameter collection module 100, the screening test module 110, and the diagnostic alarm module 120, respectively. The details of each step have been described in the previous embodiment, so it will not be repeated here.

The broadband loop cable quality monitoring method using machine learning provided by the present invention has the following advantages when compared with other conventional methods:

(1) In the present invention, experts in the field with professional experience in broadband copper cable circuits and inspections and repairs will jointly participate in the extraction of relevant network management parameters and the labeling of sample data of obstacle prediction models, and then apply the fast screening method of network management parameter threshold values and AI machine learning Modeling and combining with the DELT test function can obtain the predictive result of VDSL2 loop obstacles, and then diagnose whether the broadband cable equipment is abnormal.

(2) The present invention only needs to add value through the network management information to provide telecommunication cable obstacle diagnosis and monitoring, and find out the obstacles of the VDSL2 circuit's distribution cable, the distribution box cable and the junction box DSLAM equipment connector on the broadband copper cable loop, and realize The goal of rapid and large-scale diagnosis.

(3) The present invention reduces the professional threshold and difficulty of telecommunications cable inspection and repair, and the introduction of AI technology can play an auxiliary effect and provide better network service quality.

(4) The present invention uses AI machine learning technology to implement cable obstacle diagnosis on broadband networks, and provides cable maintenance personnel with intelligent analysis of repair targets through various original algorithms and based on big data collection. Therefore, not only can it save a lot of time for inspection and repair and improve efficiency, on the other hand, it can also greatly improve the satisfaction of broadband users.

(5) The present invention can reduce the personnel cost of network maintenance and operation, and can also ensure the reliability and stability of the broadband copper cable loop, thereby improving maintenance efficiency, and its economic benefits are very obvious.

The above-mentioned embodiments only exemplarily illustrate the principles and effects of the present invention, and are not intended to limit the present invention. Anyone who is familiar with this technique can modify and change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the rights of the present invention should be listed in the scope of patent application described later.

S700~S720: process steps

Claims (10)

A broadband loop cable quality monitoring system, wherein the broadband loop has a plurality of circuits, and the system includes: The network management performance parameter collection module is used to obtain the network management performance parameters and basic data of the broadband loop; The screening test module is used to select a part of the circuit from the plurality of circuits of the broadband loop according to the performance parameters of the network management and the basic data, and perform a line double-ended test on each of the circuits of the part to Generate test data; and The diagnostic alarm module is used to output the waiting list of the cable equipment of the broadband loop according to the test data. For example, the broadband loop cable quality monitoring system described in item 1 of the scope of patent application, the screening test module includes: Obstacle circuit quick screening module for selecting at least one possible obstacle circuit from the plurality of circuits of the broadband loop according to the network management performance parameters; and The barrier circuit test module is used to select the part of the circuit from the plurality of circuits of the broadband loop according to each possible barrier circuit and the basic data, and perform the circuit for each of the parts of the circuit Double-ended testing to generate the test data. The broadband loop cable quality monitoring system as described in item 2 of the scope of patent application, in which The obstacle circuit quick screening module includes at least one of a first judgment module, a second judgment module, and a third judgment module, wherein: The first judgment module is used to compare at least one network management performance parameter of each circuit of the broadband loop with a corresponding first threshold value. If any one of the network management performance parameters exceeds the first threshold value, Then the circuit to which the network management performance parameter belongs is the possible obstacle circuit; The second judgment module is used to calculate the degradation amplitude of at least one of the network management performance parameters of each circuit of the broadband loop compared with the corresponding original value. If there is any one of the circuits, the degradation amplitude exceeds the corresponding The second threshold value of, then the circuit is the possible obstacle circuit; and The third judgment module is used to input the network management performance parameters of each circuit of the broadband loop into an artificial intelligence model to generate the obstacle probability value of the circuit. If the obstacle probability value is greater than the third threshold value, the circuit is the May obstruct the circuit. For example, the broadband loop cable quality monitoring system described in item 2 of the scope of patent application, wherein the obstacle circuit test module is further used to classify each possible obstacle circuit in a plurality of cable devices in the broadband loop based on the basic data Class, and calculate the obstacle ratio of each cable device, wherein the circuit of the part includes each circuit of each cable device whose obstacle ratio is greater than the threshold value. The broadband loop cable quality monitoring system described in item 4 of the scope of patent application, wherein, for each cable device, the obstacle ratio is the number of possible obstacle circuits passing through the cable device among the total number of circuits passing through the cable device The percentage. For example, the broadband loop cable quality monitoring system described in item 1 of the scope of patent application, the diagnostic alarm module includes: The loop failure diagnosis module is used to generate the failure probability value of each of the circuits in the part according to the test data; The quality index conversion module is used to calculate the quality index of each cable device of the broadband loop according to the failure probability value; and The cable obstruction alarm module is used to output the waiting list according to the quality index. According to the broadband loop cable quality monitoring system described in item 6 of the scope of patent application, the loop fault diagnosis module is further used to input the test data of each of the circuits of the part into an artificial intelligence model to generate the The value of the probability of the failure of each of the parts of the circuit. For example, the broadband loop cable quality monitoring system described in item 6 of the scope of patent application, wherein the quality index conversion module is further used for each of the fault probability values in the circuit of the part greater than the threshold value according to the basic data Classify the multiple cable devices in the broadband loop to calculate the obstacle ratio of each cable device, where the quality index is the obstacle ratio or is linearly inversely proportional to the obstacle ratio. Such as the broadband loop cable quality monitoring system described in item 8 of the scope of patent application, in which, The quality index is the obstacle ratio, and the list to be repaired includes each cable device whose quality index is lower than the first threshold, or The quality index is linearly inversely proportional to the barrier ratio, and the to-be-repaired list includes each cable device whose quality index is higher than the second threshold value. A method for monitoring the quality of broadband loop cables, including: Obtain the network management performance parameters and basic data of the broadband loop; According to the network management performance parameters and the basic data, select a part of the circuit from the plurality of circuits in the broadband loop; Perform a line double-ended test on each of the circuits in this part to generate test data; and According to the test data, the list of cable equipment to be repaired for the broadband loop is output.

Images