Method and system for detecting relationship between signaling identifier and link time slot control in a network

Abstract

The present invention relates to a method and system for detecting the relationship between signaling identifiers and link time slots in a network, which transforms the signaling communication data or voice communication data of one path within a certain time range into feature vectors with the same dimension, and then obtains the relationship between the signaling identifier and the E1 link time slot by calculating the similarity distance between the vectors. The present invention is based on the DTW distance measurement comparison of feature vectors, which is a non-linear adjustment distance measurement comparison based on the overall features of the data, with higher accuracy and a wider range of applicable scenarios, rather than directly checking the time range of single-point data in the prior art solution; the signaling identifier is normalized, and the signaling identifiers are statistically counted for the time slots in the upper and lower half areas respectively, which not only simplifies the statistical process but also avoids the uncertainty of the transmission content of time slot TS16. Finally, the upper and lower half area conditions are jointly filtered to further strengthen the accuracy of the result.

Description

Technical Field

[0001] The present invention relates to the field of Public Switched Telephone Network (PSTN) communication, and particularly to a method and system for detecting the relationship between signaling identifiers and link time slots in a network. Background Art

[0002] The PSTN network is a commonly used telephone network in life. Its physical layer adopts the E1 / T1 transmission standard. The E1 frame structure is as Figure 1 shown. An E1 frame consists of 32 time slots (TS), with a frame length of 256 bit, a frame frequency of 8000 frames / s, and a frame rate of 2.048 Mbit / s. Each time slot consists of 8 bit, and the time slot rate is 64 Kbit / s. Among them, time slot TS0 is mainly used to transmit frame synchronization data, and the remaining 31 time slots are used to transmit user data. Each time slot can be used as an independent channel to transmit voice and signaling data, or multiple time slots can form a complete channel to transmit a path of signaling data.

[0003] During the voice communication process, data in two aspects, namely voice (service plane) and signaling (control plane), are simultaneously transmitted in the E1 link time slots. In the PSTN network, this type of control information used for establishing, maintaining, and releasing the relationship between communication parties is collectively referred to as signaling, and the No. 7 signaling protocol system is adopted, which is a common-channel signaling type. The common-channel signaling system is as Figure 2 shown. The signaling channel and the service channel are completely separated, and the signaling information of all trunk lines and all communication services is transmitted in the form of messages on the common data link.

[0004] In the scenario of voice call monitoring, in order to effectively intercept voice data and its associated control plane information such as the main and called numbers, it is necessary to collect voice data on the corresponding E1 link time slots according to the signaling messages. However, since the signaling messages themselves do not carry the position information of the E1 link time slots where the service plane data is located, and the transmission of signaling and voice data is not in the same channel, how to find the link time slot where the corresponding voice data is located according to the signaling messages becomes a difficult point. The existing technical solution is based on the principle that the start and end time points of signaling and voice in the same call process are basically the same, and the principle that the relationship between the signaling identifier (the signaling point code ODPC and the circuit identification code CIC carried on the signaling message) and the controlled E1 link time slot remains unchanged for a long time. By continuously inputting a large number of multi-channel signaling connection messages and voice connection messages (the connection messages are the information merged from the start and end time points of signaling or voice in the same call process) for a long time, a sliding window is established on the time axis, and the signaling connection messages and voice connection messages within the window are compared and matched one by one at single points in time. When the matching count value reaches a certain threshold, it is considered that there is a corresponding relationship between a certain signaling identifier and a certain E1 link time slot.

[0005] According to the existing technical solutions, the correspondence between signaling identifiers and E1 link time slots is obtained by comparing and counting signaling communication messages and voice communication messages one by one at single points within a sliding window based on the time axis for a long time. In the actual application process, due to the limitations of voice communication detection technology (mainly the technology for detecting the start and end time points of a link time slot based on voice frequency and energy), there are certain misjudgments and time point deviations in the voice communication results; due to the processing capabilities of front-end devices and line problems, signaling communication messages are incomplete or lost; due to the time synchronization problems of server devices, the time point benchmarks marked on signaling and voice communication messages are not unified; and when multiple signaling and voice communication messages need to be processed simultaneously, only comparing and matching the start and end times at single points of signaling communication messages and voice communication messages for counting, there are too few characteristic items for individual comparisons, the matching credibility is low, the interference of each data path is large, the complexity is high, the result accuracy is low, and a high requirement is imposed on the time unity of each server device (the time difference between each server is within 7 seconds), and the applicable scenarios are limited.

[0006] It should be noted that the information disclosed in the above background art is only used to enhance the understanding of the background of the present disclosure, and therefore may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0007] The object of the present invention is to overcome the deficiencies of the prior art, and provides a method and system for detecting the control relationship between signaling identifiers and link time slots in a network, and solves the deficiencies existing in the prior art.

[0008] The object of the present invention is achieved by the following technical solutions: A method for detecting the control relationship between signaling identifiers and link time slots in a network, the method comprising:

[0009] S1. A step of preprocessing communication data: respectively processing the input signaling and voice detection messages to obtain signaling communication data and voice communication data with signaling identifiers and E1 link time slots as key values and communication time data sequences as contents;

[0010] S2. A step of waveform transformation and sampling: converting each signaling communication data and voice communication data sequence into discrete square wave data with a fixed period;

[0011] S3. A step of extracting feature vectors: performing discrete Fourier transform on the square wave data of each signaling communication and voice communication, and taking the coefficients of the first r wave components with the largest amplitudes after the transformation as feature vectors;

[0012] S4. A step of measuring the similarity distance: measuring the similarity distance between the feature vectors between signaling and voice communication based on the DTW distance;

[0013] S5. Similarity comparison and statistics step: For each time slot of the E1 link, select the top p signaling identifiers with the shortest corresponding distance as the candidate signaling identifiers for that time slot, and count the candidate signaling identifiers;

[0014] S6. Control relationship output step: Mark the credibility of the control relationship results of the output signaling identifiers and link time slots, and output the results according to the credibility.

[0015] The communication data preprocessing step specifically includes:

[0016] Perform communication restoration on the input m-channel signaling messages and n-channel voice detection messages respectively, merge the messages in the same call, associate the start and end time points, and remove abnormal communications;

[0017] For the i-th channel of signaling, obtain the signaling communication data with the signaling identifier as the key value and the communication time data sequence as the content, denoted as XL i , i = 1 to m;

[0018] For the j-th channel of voice, obtain the voice communication data with the E1 link time slot as the key value and the communication time data sequence as the content, denoted as HL j , j = 1 to n.

[0019] The feature vector extraction step includes:

[0020] Transform the square wave data in the time domain into a set of frequency domain data through discrete Fourier transform, decompose it into multiple triangular wave components, take the coefficients of the first r wave components with the largest amplitude after transformation as the feature vector, and perform standardization;

[0021] Denote the feature vector of the i-th channel of signaling communication as FOURIER_XL i , denote the square wave data of the j-th channel of voice communication as FOURIER_HL j , r is the number of wave components selected, and the feature vector dimensions of each channel of signaling and voice communication are the same.

[0022] The similarity distance measurement step specifically includes: When the feature vector extraction of each channel of signaling communication and voice communication is completed, based on the recurrence relation formula of DTW distance r(i, j) = d(q i , c j ) + min(r(i - 1, j - 1), r(i - 1, j, r(i, j - 1)) to measure the similarity distance between the feature vectors of signaling communication and voice communication, where r(i, j) is the cumulative distance from [1,1] to [i,j].

[0023] The similarity comparison and statistics step specifically includes:

[0024] After the distance measurement between the call signaling and voice call feature vectors for each path is completed, for each time slot of each E1 link, select the top p signaling identifiers with the shortest corresponding distance as the candidate signaling identifiers for that time slot;

[0025] Perform normalization processing on all candidate signals for each time slot. Divide the time slots of an E1 link into upper and lower half areas. Transform the candidate signaling identifiers of all upper half area time slots into the candidate signaling identifiers corresponding to TS1, and transform the candidate signaling identifiers of all lower half area time slots into the candidate signaling identifiers corresponding to TS17, and count the candidate signaling identifiers.

[0026] The similarity comparison and statistics step further includes:

[0027] Compare the upper and lower half areas of all candidate signaling identifiers that reach a certain counting threshold in descending or ascending order. If the signaling point code ODPC is the same and the difference in the circuit identification code CIC value meets the condition, then consider this candidate signal as the actual controlled signal of TS1 or TS17 and suspend the comparison; otherwise, continue the comparison, so as to obtain the control relationship between all time slots and signaling identifiers in this E1 link.

[0028] The control relationship output step specifically includes:

[0029] Mark the credibility of the control relationship results of the output signaling identifiers and link time slots. When the same result appears, the credibility count is incremented by 1. When it reaches the maximum credibility c, the increment stops;

[0030] When different results appear, the credibility count is decremented by 1. When it is decremented to 0, it is updated to the latest result that appears currently.

[0031] A method for detecting the control relationship between signaling identifiers and link time slots in a network, which includes a communication data preprocessing module, a waveform transformation and sampling module, a feature vector extraction module, a similarity distance measurement module, a similarity comparison and statistics module, and a control relationship output module;

[0032] The communication data preprocessing module is used to process the input signaling and voice detection messages respectively, and obtain signaling communication data and voice communication data with signaling identifiers and E1 link time slots as key values and communication time data sequences as contents;

[0033] The waveform transformation and sampling is used to convert each path of signaling communication data and voice communication data sequence into discrete square wave data with a fixed period;

[0034] The feature vector extraction is used to perform discrete Fourier transform on the square wave data of each path of signaling communication and voice communication, and take the coefficients of the first r wave components with the largest amplitudes after the transformation as feature vectors;

[0035] The similarity distance measurement is used to measure the similarity distance between the feature vectors of signaling and voice communication based on the DTW distance;

[0036] The similarity comparison statistics is used to select the top p signaling identifiers with the shortest corresponding distances for each time slot of each E1 link as the candidate signaling identifiers for that time slot, and count the candidate signaling identifiers;

[0037] The control relationship output is used to mark the credibility of the control relationship result of the output signaling identifier and link time slot, and output the result according to the credibility.

[0038] The present invention has the following advantages: A method and system for detecting the control relationship between signaling identifiers and link time slots in a network, through data transformation and feature extraction, greatly reduces the data dimension, while retaining most of the information of the original data sequence, with higher algorithm efficiency and relatively simple processing; The DTW distance measurement comparison based on feature vectors is a non-linear adjusted distance measurement comparison based on the overall features of the data, with higher accuracy and a wider applicable scenario, rather than directly checking the time range of single-point data in the prior art solution; The normalization processing of signaling identifiers, and the statistical calculation of signaling identifiers for each time slot in the upper and lower half areas respectively, not only simplifies the statistical process, but also avoids the uncertainty of the transmission content of time slot TS16. Finally, through the joint filtering of the upper and lower half area conditions, the accuracy of the result is further enhanced. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 It is a schematic diagram of the E1 frame structure;

[0040] Figure 2 It is a schematic diagram of the common-channel signaling system;

[0041] Figure 3 It is a schematic diagram of the relationship between signaling identifiers and E1 link time slots;

[0042] Figure 4 It is a schematic flowchart of the method of the present invention;

[0043] Figure 5 It is a scatter diagram of the original voice communication data;

[0044] Figure 6 It is a square wave diagram of the voice communication data;

[0045] Figure 7 It is a spectrogram of the square wave data of the voice communication;

[0046] Figure 8 It is a spectrogram of the square wave data of the signaling communication;

[0047] Figure 9 It is a schematic diagram of the alignment of the feature sequences;

[0048] Figure 10 It is a comparison graph of DTW distance. Specific implementation manners

[0049] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part rather than all of the embodiments of the present application. Usually, the components of the embodiments of the present application described and illustrated in the accompanying drawings here can be arranged and designed in various different configurations. Therefore, the detailed description of the embodiments of the present application provided below with reference to the accompanying drawings is not intended to limit the protection scope of the claimed present application, but only represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the protection scope of the present application. The present invention will be further described below with reference to the accompanying drawings.

[0050] The present invention specifically relates to a method for detecting the control relationship between a signaling identifier and an E1 link time slot in a PSTN network, which transforms the signaling communication data or voice communication data of one path within a certain time range into feature vectors with the same dimension, and then obtains the relationship between the signaling identifier and the E1 link time slot by calculating the similarity distance between the vectors.

[0051] In PSTN voice communication, the control relationship between the No. 7 signaling identifier and the E1 link time slot is configured by the network device side, which is relatively stable but can be dynamically changed according to needs. As Figure 3 shown, it shows the control relationship between the signaling identifier and the E1 link time slot, where TS16 is the signaling channel. Within one E1 link, time slot TS0 is not assigned a signaling identifier, TS1 - 15 and TS17 - 31 are assigned signaling identifiers, and TS16 is assigned a signaling identifier according to whether it is used as a voice channel, and the signaling point codes in the signaling identifiers assigned to each time slot are the same, and the circuit identification code CIC is continuous.

[0052] According to the above description, the control relationship between the signaling identifier and the E1 link time slot is configured by the operator's network device, but its control relationship cannot be directly obtained in the voice monitoring scenario and must be detected by software. This proposal provides a detection method for the control relationship between the signaling identifier and the E1 link time slot. The overall process of the technical solution is shown in the following figure, and mainly includes communication data preprocessing, waveform transformation sampling, maximum feature extraction, similarity distance measurement, similarity comparison statistics, etc.

[0053] As Figure 4 shown, it specifically includes the following contents:

[0054] Step 1: Conduct communication restoration on the input m-channel signaling messages and n-channel voice detection messages respectively, merge the messages in the same call, associate the start and end time points, and remove abnormal communications. For the i-th (i = 1 to m) channel of signaling, obtain the signaling communication data with the signaling identifier (ODPC + CIC) as the key value and the communication time (start time, end time) data sequence as the content, denoted as XLi(ODPC_CIC, TL[k,2]), where k is the number of communications, and the k value is different for each signaling channel; for the j-th (j = 1 to n) channel of voice, obtain the voice communication data with the E1 link time slot (LINE + TS) as the key value and the communication time (start time, end time) data sequence as the content, denoted as HLj(LINE_TS, TL[k,2]), where k is the number of communications, and the k value is different for each voice channel.

[0055] Step 2: Convert the communication data sequences of each signaling channel and voice channel into discrete square wave data with a fixed period for subsequent discrete Fourier transform to extract features. As Figure 5 shows the scatter plot of the original voice communication data of a certain E1 link time slot TS1 within a 12-hour range. As Figure 6 shows the corresponding converted square wave diagram. The X-axis of the scatter plot is the start time of the communication, and the Y-axis is the end time of the communication; the X-axis of the square wave diagram is time, and the Y-axis is the amplitude. When there is a call at the corresponding time point, the amplitude value is set to A (A!= 0), otherwise the amplitude value is set to 0. Sample the square wave with a sampling period of T seconds and a sampling frequency of FS = 1 / T. The sampling parameters of each waveform data are the same. The square wave data of the i-th signaling channel is denoted as SQUARE_WARE_XLi(ODPC_CIC, data[s,1]), and the square wave data of the j-th voice channel is denoted as SQUARE_WARE_HLj(LINE_TS, data[s,1]), where s is the total number of samples, and the total number of samples s of the square wave sampling of each signaling and voice communication is the same.

[0056] Step 3: Conduct discrete Fourier transform (DFT) on the square wave data of each signaling channel and voice channel. The discrete Fourier transform formula is

[0057] Convert the square wave data in the time domain into a set of frequency domain data and decompose it into multiple triangular wave components. As Figure 7 shows the spectrogram of the square wave data of the voice communication after Fourier transform in Step 2. As Figure 8shows the spectrogram after Fourier transform of the square wave data corresponding to its corresponding signaling communication. After the transformation, the coefficients of the first r (r>0) wave components with the largest amplitudes are taken as the feature vectors and normalized to balance the weights of each coefficient item (frequency, amplitude, phase). The feature vector of the i-th signaling communication is denoted as FOURIER_XL i (ODPC_CIC, feature[r,3]), and the square wave data of the j-th voice communication is denoted as FOURIER_HL j (LINE_TS, feature[r,3]), where r is the number of wave components selected, and the feature vector dimensions of each signaling and voice communication are the same.

[0058] Step 4. After the feature vectors of each signaling communication and voice communication are extracted, based on the DTW distance (Dynamic Time Warping), the similarity distance between the feature vectors of signaling and voice communication is measured. DTW is based on the dynamic programming strategy to non-linearly align two sequences in the time domain to facilitate the correct calculation of the similarity between them. The recurrence relation formula is r(i, j) = d(q i , c j ) + min(r(i - 1, j - 1), r(i - 1, j), r(i, j - 1)), where r(i, j) is the cumulative distance from [1,1] to [i,j]. As Figure 8 is the schematic diagram of sequence alignment between the feature vectors of voice and signaling communication in Step 3. The feature vector contains 100 feature components, and the black dots are the optimal matching paths under the shortest distance between the two feature vectors.

[0059] Step 5. After the distance measurement between the feature vectors of each signaling communication and voice communication is completed, for each E1 link time slot, the first p (p>0) signaling identifiers with the shortest corresponding distance are selected as the candidate signaling identifiers for this time slot. Figure 9Display the top 20 signaling identifiers with the shortest distances corresponding to the time slot TS1 in the E1 link during voice communication in step 2, where the black line indicates the position of the actual controlled signaling identifier of this link time slot. Then, it is necessary to normalize all candidate signaling for each time slot. An E1 link time slot is divided into two half - zones. TS1 - 16 is the upper half - zone, and TS17 - 31 is the lower half - zone. Transform the candidate signaling identifiers of all upper - half - zone time slots into the candidate signaling identifiers corresponding to TS1, and transform the candidate signaling identifiers of all lower - half - zone time slots into the candidate signaling identifiers corresponding to TS17, and count the candidate signaling identifiers. Compare the candidate signaling identifiers that reach a certain counting threshold in descending order between the upper and lower half - zones. Among them, the signaling point code ODPC needs to be the same, and the circuit identification code CIC value differs by 15 (TS16 is a signaling channel) or 16 (TS16 is a voice channel). If the above conditions are met, it is considered that the candidate signaling is the actual controlled signaling of TS1 or TS17 and the comparison is paused; otherwise, the comparison continues. Since the ODPC is the same and the CIC is continuous among the signaling identifiers corresponding to all time slots of the same E1 link, the control relationship between all time slots and signaling identifiers in this E1 link, as well as whether the TS16 time slot is a voice channel, can be obtained.

[0060] Step six: As Figure 10 shown, mark the credibility of the control relationship result of the output signaling identifier and the link time slot. When the same result appears, the credibility count is incremented by 1, and when it reaches the maximum credibility c (c > 0), the increment stops; when different results appear, the credibility count is decremented by 1, and when it is decremented to 0, it is updated to the latest result that appears currently.

[0061] The present invention transforms non - periodic data with an unfixed length into periodic data with a fixed length. By checking whether there is a call at the corresponding time point, the communication data within a certain time range is transformed into square - wave data, and then discrete square - wave data with a fixed - period property is obtained according to a certain sampling frequency, providing a data basis for subsequent feature extraction;

[0062] Overall data dimensionality reduction and non - linear adjustment for similarity ranging. Perform a discrete Fourier transform on the transformed square - wave data to extract feature vectors, greatly reducing the data dimension while ensuring the efficiency and effectiveness of the algorithm. Then, based on the feature vectors, perform DTW distance measurement to ensure the accuracy of the similarity measurement between feature vectors;

[0063] Signaling identifier normalization processing: perform normalization processing on all obtained time slot candidate signaling identifiers. Divide the time slots in the same E1 link into upper and lower half areas, so that the signaling identifiers corresponding to all time slots are finally transformed into the signaling identifiers corresponding to TS1 or TS17 time slots, and then perform statistical counting. Finally, based on the count value and the joint comparison and filtering of the upper and lower half areas, obtain the control relationship between the signaling identifier and the time slot of this link, which maximally ensures the accuracy and reliability of the result.

[0064] The above are only the preferred embodiments of the present invention. It should be understood that the present invention is not limited to the form disclosed herein, should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be changed within the scope of the concept described herein through the above teachings or the techniques or knowledge in related fields. And the changes and alterations made by those skilled in the art that do not depart from the spirit and scope of the present invention shall fall within the protection scope of the appended claims of the present invention.

Claims

1. A method for detecting the relationship between signaling identifiers and link time slot control in a network, characterized in that: The method includes: S1. Communication data preprocessing step: Process the input signaling and voice detection messages respectively to obtain signaling communication data and voice communication data with signaling identifiers and E1 link time slots as key values and communication time data sequences as contents; S2. Waveform transformation and sampling step: Convert each signaling communication data and voice communication data sequence into discrete square wave data with a fixed period; S3. Feature vector extraction step: Perform discrete Fourier transform on the square wave data of each signaling communication and voice communication, and take the coefficients of the first r wave components with the largest amplitudes after transformation as feature vectors; S4. Similarity distance measurement step: Measure the similarity distance between the feature vectors of signaling and voice communications based on the DTW distance; S5. Similarity comparison and statistics step: Select the first p signaling identifiers with the shortest corresponding distances for each E1 link time slot as the candidate signaling identifiers for that time slot, and count the candidate signaling identifiers; S6. Control relationship output step: Mark the credibility of the control relationship results of the output signaling identifiers and link time slots, and output the results according to the credibility; 2. The method for detecting the control relationship between signaling identifiers and link time slots in a detection network according to claim 1, wherein: The communication data preprocessing step specifically includes: Perform communication restoration on the input m-way signaling messages and n-way voice detection messages respectively, merge the messages in the same call, associate the start and end time points, and remove abnormal communications; For the i-th signaling, obtain the signaling communication data with the signaling identifier as the key value and the communication time data sequence as the content, denoted as XL i , where i = 1 to m; For the j-th voice channel, obtain the voice communication data with the E1 link time slot as the key value and the communication time data sequence as the content, denoted as HL j , where j = 1 to n.

3. The method for detecting the control relationship between signaling identifiers and link time slots in a detection network according to claim 1, wherein: The feature vector extraction step includes: Transform the square wave data in the time domain into a set of frequency domain data through discrete Fourier transform, decompose it into multiple triangular wave components, take the coefficients of the first r wave components with the largest amplitudes after transformation as feature vectors, and perform standardization; Denote the eigenvector of the \(i\)-th signaling communication as FOURIER_XL i and denote the square wave data of the \(j\)-th voice communication as FOURIER_HL j where \(r\) is the number of selected wave components, and the eigenvector dimensions of each signaling and voice communication are the same.

4. The method for detecting the control relationship between a signaling identifier and a link time slot in a detection network according to claim 1, wherein: The specific steps for measuring the similarity distance are as follows: After the feature vectors of each signaling communication and voice communication are extracted, based on the recurrence relation formula of the DTW distance measure the similarity distance between the feature vectors of the signaling communication and the voice communication, where r(i,j) is the cumulative distance from [1,1] to [i,j], i represents the i-th signaling, and j represents the j-th voice.

5. The method for detecting the control relationship between signaling identifiers and link time slots in a detection network according to claim 1, wherein: The similarity comparison and statistics step specifically includes: After the distance measurement between the feature vectors of each signaling communication and voice communication is completed, select the first p signaling identifiers with the shortest corresponding distances for each E1 link time slot as the candidate signaling identifiers for that time slot; Perform normalization processing on all candidate signals of each time slot. Divide an E1 link time slot into upper and lower half regions. Transform the candidate signaling identifiers of all upper half region time slots into the candidate signaling identifiers corresponding to TS1, and transform the candidate signaling identifiers of all lower half region time slots into the candidate signaling identifiers corresponding to TS17, and count the candidate signaling identifiers; 6. The method for detecting the control relationship between signaling identifiers and link time slots in a detection network according to claim 5, characterized in that: The similarity comparison and statistics step further includes: Compare the upper and lower half regions of all candidate signaling identifiers that reach a certain counting threshold in descending order. If the signaling point code ODPC is the same and the difference in the circuit identification code CIC value meets the conditions, then consider this candidate signaling as the actual controlled signaling of TS1 or TS17 and suspend the comparison, otherwise, continue the comparison, so as to obtain the control relationship between all time slots and signaling identifiers in this E1 link; 7. The method for detecting the control relationship between signaling identifiers and link time slots in a detection network according to claim 1, wherein: The control relationship output step specifically includes: Mark the credibility of the control relationship results of the output signaling identifiers and link time slots. When the same result appears, the credibility count is incremented by 1. When the maximum credibility c is reached, the increment stops; When different results appear, the credibility count is decremented by 1. When it is decremented to 0, it is updated to the latest result that appears currently.

8. A system for detecting the relationship between signaling identifiers and link time slot control in a network, characterized in that: It includes a communication data preprocessing module, a waveform transformation and sampling module, a feature vector extraction module, a similarity distance measurement module, a similarity comparison and statistics module, and a control relationship output module; The communication data preprocessing module is used to process the input signaling and voice detection messages respectively, and obtain signaling communication data and voice communication data with the signaling identifier and the E1 link time slot as the key values respectively, and the communication time data sequence as the content; The waveform transformation and sampling is used to convert each path of signaling communication data and voice communication data sequence into discrete square wave data with a fixed period; The feature vector extraction is used to perform discrete Fourier transform on the square wave data of each path of signaling communication and voice communication, and take the coefficients of the first r wave components with the largest amplitude after the transformation as the feature vector; The similarity distance measurement is used to measure the similarity distance of the feature vectors between signaling and voice communication based on the DTW distance; The similarity comparison and statistics is used to select the first p signaling identifiers with the shortest corresponding distance for each E1 link time slot as the candidate signaling identifiers for this time slot, and count the candidate signaling identifiers; The control relationship output is used to mark the credibility of the control relationship result of the output signaling identifier and the link time slot, and output the result according to the credibility.

Images