Optical cable fault detection method, system and device based on distributed base station and medium

Abstract

The invention provides an optical cable fault detection method and system based on a distributed base station, electronic equipment and a storage medium, and aims to solve the problem of low detection accuracy, the method comprises the following steps: collecting receiving optical power values of a plurality of RRUs, and normalizing the receiving optical power values; for each RRU, acquiring a normalized optical power value sequence in a corresponding first window before a previous moment, and calculating a first interference coefficient according to a fluctuation condition; determining the length of a second window before the latest moment according to the first interference coefficient and the first window, and obtaining the interference coefficient of each moment to form a sequence; obtaining a change trend of the sequence, and calculating a disturbance intensity accumulated value; calculating DTW distances between the target RRU sequence and other RRUs to obtain a maximum deviation degree; and according to the ratio of the disturbance intensity accumulated value of the target RRU to the average value of the other RRUs and the maximum deviation degree, judging whether the corresponding optical cable has a fault or not. The detection accuracy can be improved.

Description

Technical Field

[0001] This disclosure relates to the field of communication technology, specifically to a method for detecting optical cable faults based on distributed base stations, a system for detecting optical cable faults based on distributed base stations, an electronic device, and a computer-readable storage medium. Background Technology

[0002] As mobile communication networks evolve towards higher bandwidth, lower latency, and higher reliability, distributed base station architecture has become the mainstream deployment method. In this architecture, the Remote Radio Unit (RRU) is connected to the Building Base Band Unit (BBU) via optical fiber. The health status of the optical fiber directly affects signal transmission quality and network coverage performance. Traditional optical fiber fault detection methods are mostly based on statistical analysis of received optical power within a fixed time window.

[0003] Distributed base station installation environments are complex. Short-term interferences such as high-voltage lines near outdoor equipment rooms, construction vibrations, or lightning strikes can cause sudden changes in optical power on a timescale of milliseconds to seconds. Fixed window lengths are often unable to respond quickly to these short-term fluctuations, or the window may be too short and be overwhelmed by noise, or too long and cause signal passivation. Furthermore, optical cable faults often exhibit specific disturbance patterns simultaneously at multiple adjacent ports, making it difficult to accurately locate and reduce false alarms by relying solely on the statistical characteristics of a single port.

[0004] Therefore, a fault detection technology for optical cables that can adaptively cope with short-term strong interference is needed to improve the accuracy and reliability of optical cable monitoring in distributed base stations. Summary of the Invention

[0005] To address the low accuracy of optical cable fault detection caused by short-term strong interference and specific disturbances at adjacent ports in existing technologies, this disclosure provides an optical cable fault detection method, system, electronic device, and computer-readable storage medium based on a distributed base station. This method can dynamically expand and shrink the time window, balancing detection response speed and stability, achieving higher detection accuracy, and precisely locating the corresponding port of the faulty optical cable.

[0006] In a first aspect, this disclosure provides a method for optical cable fault detection based on distributed base stations, the method comprising:

[0007] Collect the received optical power values ​​of multiple RRUs and normalize the received optical power values ​​of each RRU.

[0008] For each RRU, obtain its normalized optical power value sequence within the first adaptive time period before the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence.

[0009] Based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period, the length of the second adaptive time period before the latest moment of the RRU is dynamically determined, and the interference coefficient of the RRU at each moment within the second adaptive time period is obtained to form the interference coefficient sequence of the RRU at the latest moment; and,

[0010] Obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend;

[0011] Calculate the DTW (Dynamic Time Warping) distance between the target RRU's interference coefficient sequence at the latest time and the interference coefficient sequences of each of the other RRUs at the latest time to obtain the maximum deviation of the target RRU at the latest time.

[0012] Calculate the average value of the cumulative disturbance intensity of the remaining RRUs other than the target RRU at the latest time. Determine whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time.

[0013] Furthermore, the calculation of the first interference coefficient of the RRU at the previous moment based on the fluctuation of the normalized optical power value sequence includes:

[0014] Sort the normalized optical power value sequence and take the median as the benchmark;

[0015] Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

[0016] Furthermore, the step of dynamically determining the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period includes:

[0017] Set an initial adaptive time period value, and obtain the initial number of samples based on the sampling time interval;

[0018] If the time from the initial start to the current time is greater than or equal to 2, then a preset reference constant is divided by the first interference coefficient of the RRU at the previous time, and then multiplied by the number of samples in the first adaptive time period before the previous time to obtain the number of samples of the RRU in the second adaptive time period before the latest time. This number of samples should not be less than the preset minimum number of samples.

[0019] The length of the second adaptive time period before the latest time of the RRU is obtained by multiplying the number of samples taken by the RRU in the second adaptive time period before the latest time by a fixed sampling time interval value.

[0020] Furthermore, the step of obtaining the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculating the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend, includes:

[0021] Calculate the first-order difference of the interference coefficient sequence at the latest moment, and accumulate the positive difference values ​​to obtain the accumulated value of the perturbation intensity of the RRU at the latest moment.

[0022] Furthermore, the step of determining whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest moment to the average value, and the maximum deviation of the target RRU at the latest moment, includes:

[0023] The risk coefficient of the target RRU at the latest time is calculated based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time.

[0024] Set a risk coefficient threshold;

[0025] When the risk coefficient of the target RRU is greater than or equal to the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station has experienced a cable fault.

[0026] When the risk coefficient of the target RRU is less than the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station is normal.

[0027] Furthermore, the risk coefficient is calculated using the following formula (1):

[0028] (1)

[0029] in, Let J be the risk coefficient of the i-th RRU at the latest time, J be the total number of RRUs other than the i-th RRU, and j represent the traversal of J;

[0030] This represents the sequence of interference coefficients for the i-th RRU at the latest time. The sequence of interference coefficients of the j-th RRU at the latest time DTW distance between them This represents the maximum deviation of the i-th RRU at the latest time.

[0031] This represents the cumulative disturbance intensity of the i-th RRU at the latest time. This represents the average value of the accumulated disturbance values ​​of all RRUs except the i-th RRU at the latest time.

[0032] Secondly, this disclosure provides an optical cable fault detection system based on a distributed base station, the system comprising:

[0033] The acquisition and processing module is configured to acquire the received optical power values ​​of multiple RRUs and normalize the received optical power values ​​of each RRU.

[0034] The first calculation module is configured to, for each RRU, obtain the normalized optical power value sequence within the first adaptive time period corresponding to the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence.

[0035] The acquisition module is configured to dynamically determine the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous time and the corresponding first adaptive time period, acquire the interference coefficient of the RRU at each time within the second adaptive time period, and form the interference coefficient sequence of the RRU at the latest time.

[0036] The second calculation module is configured to obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend.

[0037] The third calculation module is set to calculate the DTW distance between the interference coefficient sequence of the target RRU at the latest time and the interference coefficient sequence of each of the other RRUs at the latest time, so as to obtain the maximum deviation of the target RRU at the latest time.

[0038] The judgment module is configured to calculate the average value of the cumulative disturbance intensity of the remaining RRUs (excluding the target RRU) at the latest time, and determine whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value, and the maximum deviation of the target RRU at the latest time.

[0039] Furthermore, the first calculation module is specifically configured as follows:

[0040] Sort the normalized optical power value sequence and take the median as the benchmark;

[0041] Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

[0042] Thirdly, this disclosure provides an electronic device, including a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the optical cable fault detection method based on a distributed base station as described in any of the first aspects.

[0043] Fourthly, this disclosure provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the optical cable fault detection method based on any of the first aspects described above.

[0044] Beneficial effects:

[0045] This disclosure provides a method, system, electronic equipment, and storage medium for optical cable fault detection based on distributed base stations. It dynamically expands and contracts the time window length based on the interference coefficient of the previous moment. When an instantaneous disturbance occurs, the window is rapidly contracted to improve sensitivity, and then appropriately enlarged after the disturbance stabilizes to enhance robustness, thus balancing detection response speed and stability. The interference coefficient differences at each moment within the latest adaptive window are summed to form a cumulative disturbance intensity value, which can efficiently capture short-term cumulative effects and has higher detection accuracy for slow drift or continuous small fluctuations. The Dynamic Time Warping (DTW) algorithm is used to measure the similarity of interference sequences at each remote radio frequency unit. By taking the maximum similarity distance, sequences that deviate significantly from other nodes are identified, suppressing network-wide resonance noise interference and accurately locating the corresponding port of the faulty optical cable. Attached Figure Description

[0046] Figure 1 This is a flowchart illustrating a method for optical cable fault detection based on a distributed base station, as provided in Embodiment 1 of this disclosure.

[0047] Figure 2 This is an architecture diagram of an optical cable fault detection system based on a distributed base station, provided in Embodiment 3 of this disclosure;

[0048] Figure 3 This is an architectural diagram of an electronic device provided in Embodiment 4 of this disclosure. Detailed Implementation

[0049] To enable those skilled in the art to better understand the technical solutions of this disclosure, the disclosure will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments and drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0050] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence; furthermore, in the absence of conflict, the embodiments and features in the embodiments of this disclosure can be arbitrarily combined with each other.

[0051] The terminology used in the embodiments of this disclosure is for the purpose of describing particular embodiments only and is not intended to be limiting of this disclosure. The singular forms “a,” “the,” and “the” as used in the embodiments of this disclosure and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise.

[0052] In the following description, the use of suffixes such as “module,” “part,” or “unit” to denote elements is solely for the purpose of illustrative purposes and has no specific meaning in itself. Therefore, “module,” “part,” or “unit” may be used interchangeably.

[0053] The following detailed embodiments illustrate the technical solutions of this disclosure and how they solve the technical problems existing in the prior art. It is understood that in the embodiments of this disclosure, the executing entity may perform some or all of the steps in the embodiments of this disclosure. These steps or operations are merely examples, and the embodiments of this disclosure may also perform other operations or variations thereof. Furthermore, the steps may be performed in different orders as presented in the embodiments of this disclosure, and it is not necessary to perform all the operations in the embodiments of this disclosure. Moreover, the following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments.

[0054] Example 1

[0055] Figure 1 This is a flowchart illustrating a method for optical cable fault detection based on a distributed base station, as provided in Embodiment 1 of this disclosure. Figure 1 As shown, the method includes:

[0056] Step S101: Collect the received optical power values ​​of each of the multiple RRUs, and normalize the received optical power values ​​of each RRU respectively;

[0057] Step S102: For each RRU, obtain its normalized optical power value sequence within the first adaptive time period before the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence.

[0058] Step S103: Based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period, dynamically determine the length of the second adaptive time period before the latest moment for the RRU, obtain the interference coefficient of the RRU at each moment within the second adaptive time period, and form the interference coefficient sequence of the RRU at the latest moment; and,

[0059] Step S104: Obtain the changing trend of the interference coefficient sequence of the RRU at the latest moment, and calculate the cumulative value of the disturbance intensity of the RRU at the latest moment based on the changing trend;

[0060] Step S105: Calculate the DTW distance between the interference coefficient sequence of the target RRU at the latest time and the interference coefficient sequence of each of the other RRUs at the latest time, and obtain the maximum deviation of the target RRU at the latest time;

[0061] Step S106: Calculate the average value of the cumulative disturbance intensity of the remaining RRUs other than the target RRU at the latest time. Based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value, and the maximum deviation of the target RRU at the latest time, determine whether the optical cable corresponding to the target RRU is faulty.

[0062] The purpose of this disclosure is to: provide an adaptive time window mechanism to address the problem that fixed time windows cannot adapt to short-term disturbances, dynamically adjusting the window length based on the disturbance intensity of the previous moment; address the lack of lateral comparison in analyzing a single RRU by introducing DTW distance analysis between multiple RRUs to identify abnormal nodes that deviate from the network-wide disturbance pattern; and address the issues of slow detection response and high false alarm rate in existing solutions by constructing a risk coefficient technical feature that fuses the accumulated disturbance intensity value and the maximum DTW distance to achieve rapid point-to-point fault location.

[0063] Specifically, the received optical power values ​​of multiple RRUs are first collected, and the received optical power values ​​of each RRU are then normalized.

[0064] In a distributed base station, the RRU (Receiving Rush Unit) ports are polled using SNMP (Simple Network Management Protocol) or CLI (Command-Line Interface) scripts to collect data. The received optical power value of each RRU is collected every 30 seconds, and the received optical power value of the i-th RRU at time t is recorded. The 30-second parameter is a hyperparameter that can be adjusted according to the specific implementation scenario.

[0065] To eliminate power bias at each port while taking into account performance differences between different devices, the sequence of received optical power values ​​obtained after acquisition is normalized to obtain normalized received optical power values ​​for multiple RRUs.

[0066] Specifically, the normalization process is as follows: The received optical power value collected by the i-th RRU at the latest time is obtained as the latest received optical power value. The average value of the RRU over the past time period T is subtracted from the latest received optical power value to obtain the average difference. The average difference is then normalized to obtain the normalized value of the latest received optical power value. In this context, "T" refers to the period within the past 5 or 10 minutes since the latest time. The 1 in the figure represents the latest time, i.e., the latest time t=1. The calculation of the mean and standard deviation are well-known techniques and will not be elaborated further.

[0067] By collecting the received optical power value of each RRU and performing independent normalization, the inherent power baseline deviation of different RRU devices due to hardware differences and different transmission distances is eliminated, enabling fair comparison of the interference coefficient and disturbance intensity of all subsequent RRUs under the same dimension. This is the basis for conducting effective horizontal (inter-RRU) comparisons.

[0068] Then, for each RRU, the normalized optical power value sequence within the first adaptive time period corresponding to the previous moment is obtained, and the first interference coefficient of the RRU in the previous moment is calculated based on the fluctuation of the normalized optical power value sequence.

[0069] The first interference coefficient is calculated by the "fluctuation status" of the normalized optical power value sequence within the first adaptive time period. The first interference coefficient reflects whether the RRU is subject to strong interference within this first adaptive time period. The stronger the fluctuation, the larger the first interference coefficient.

[0070] Then, based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period, the length of the second adaptive time period before the latest moment of the RRU is dynamically determined; the interference coefficient of the RRU at each moment within the second adaptive time period is obtained to form the interference coefficient sequence of the RRU at the latest moment.

[0071] The length of the second adaptive time period is not fixed, but is determined by both the interference coefficient of the previous moment and the length of the time period of the previous moment.

[0072] Larger interference leads to a shorter window, which in turn increases the sensitivity to detecting recent sudden anomalies.

[0073] Less interference → longer window → enhanced robustness to long-term slow trends, smoothing out random noise.

[0074] Sequence construction: The length of the second adaptive time period before the latest time is determined, and the interference coefficients of each sampling time within this time period are extracted to form a short-term time series of interference coefficients that can reflect the latest dynamics.

[0075] Specifically, this is achieved by saving and updating two state variables for each RRU: 1) the current adaptive window length. 2) Current interference coefficient At each step, the state of the previous time step is used to calculate the state of the current time step.

[0076] The interference coefficient of the "previous moment" This requires a window from the "previous moment" (adaptive time period). To calculate. And the "latest moment" window. And it needs to go through To calculate.

[0077] The system employs recursive and incremental computational approaches. Each RRU maintains a state, which is updated at each time step based on the state from the previous time step and new observation data.

[0078] Core state variables:

[0079] For each RRU, the system only needs to persist (save) two core state variables: the adaptive window length of the previous time step and the disturbance coefficient of the previous time step;

[0080] For each RRU, at the latest sampling time t:

[0081] Obtain historical status: Read the status values ​​of the RRU saved at the previous time t-1: adaptive window length and interference coefficient. Calculate new window length: Based on the interference coefficient at the previous time, dynamically calculate the adaptive time window length at the current time t using a preset formula, then calculate the interference coefficient at the current time. t is continuously increased to form a sequence, and the newly calculated interference coefficient is appended to the end of the historical interference coefficient sequence of the RRU.

[0082] Take the interference coefficient corresponding to each time point in the second adaptive time period backward from the current time t, and perform subsequent calculations.

[0083] Each RRU maintains a state vector, including its latest adaptive window length and interference coefficient. For each sampling time, the system executes a recursive state update process: first, the window length at the current time is calculated based on the interference coefficient at the previous time, then the interference coefficient at the current time is calculated based on the data within the current window, and the state vector is updated.

[0084] Obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend;

[0085] For example, a linear fit can be performed on the entire disturbance coefficient sequence to obtain the slope k; the cumulative disturbance intensity can be defined as the absolute value of the slope k. This method is highly sensitive to slow, continuous linear drift faults, but insensitive to sudden, transient spikes. Alternatively, the variance or standard deviation of the entire disturbance coefficient sequence can be calculated to capture the dispersion of the sequence values ​​around their mean; a larger variance indicates more severe sequence fluctuations. This method is sensitive to any deviation from the mean in the sequence, but insensitive when the fault manifests as a stable shift. Alternatively, the cumulative deviation after change point detection can be used to identify points in the sequence where significant changes occur. Assuming a change occurs at time point tcp, the cumulative deviation of the sequence values ​​after the change point from the mean (or median) before the change point can be calculated. This method is effective for stationary shift faults after the fault occurs, but it has high computational complexity, requires running a change point detection algorithm, and requires a certain sequence length to effectively detect change points.

[0086] Calculate the DTW distance between the target RRU's interference coefficient sequence at the latest time and the interference coefficient sequences of each of the other RRUs at the latest time to obtain the maximum deviation of the target RRU at the latest time.

[0087] DTW: A powerful algorithm for measuring the similarity between two time series of different lengths and speeds. It finds the best alignment between two sequences by “bending” the time axis and calculates their cumulative distance.

[0088] Since the adaptive window length of each RRU changes independently and dynamically, their interference coefficient sequence lengths are likely to differ. DTW perfectly solves the problem of similarity comparison for unequal-length sequences. The DTW distance between the interference coefficient sequence of the target RRU and the interference coefficient sequences of every other RRU in the cluster is calculated, and the maximum value (Max DTW) is taken. This maximum value represents the degree of difference between the target RRU's behavior pattern and the RRU in the cluster that is "least similar" to it, i.e., the maximum deviation, thus obtaining the maximum deviation of the target RRU at the latest time. The anomalous patterns of faulty RRUs typically result in large DTW distances between them and all normal RRUs.

[0089] After obtaining the cumulative disturbance intensity value and maximum deviation of each RRU, the average value of the cumulative disturbance intensity value of the remaining RRUs (excluding the target RRU) at the latest time is calculated. Based on the ratio of the cumulative disturbance intensity value of the target RRU at the latest time to the average value, and the maximum deviation of the target RRU at the latest time, it is determined whether the optical cable corresponding to the target RRU is faulty.

[0090] By calculating its own perturbation strength (longitudinal evidence): D target / D mean-others A ratio greater than 1 indicates that the target RRU's own disturbance growth is significantly higher than the cluster average; and, behavioral pattern deviation (lateral evidence): Max DTW. The larger this value, the greater the difference between the target RRU's disturbance pattern and the behavioral pattern of any normal member in the cluster; For the final decision, the logic is: the final fault determination is the product of these two indicators or some form of weighted / fused sum. Only when the target RRU simultaneously meets both the conditions of "severe self-disturbance" and "abnormal behavioral pattern" will it be determined as faulty, thus greatly reducing false alarms. Judgment is performed on all detected RRUs, thereby quickly identifying the faulty optical cable.

[0091] This embodiment dynamically expands and contracts the time window length based on the interference coefficient of the previous moment. When a sudden disturbance occurs, the window is rapidly contracted to improve sensitivity, and then appropriately enlarged after the disturbance stabilizes to enhance robustness, thus balancing detection response speed and stability. The interference coefficients at each moment within the adaptive window at the latest moment are differentiated and positively summed to form a cumulative disturbance intensity value, which can efficiently capture short-term cumulative effects and has a higher detection accuracy for anomalies with slow drift or continuous small fluctuations. The Dynamic Time Warping (DTW) algorithm is used to measure the similarity of each RRU interference sequence. By taking the maximum similarity distance, sequences that deviate significantly from other nodes are identified, suppressing network-wide resonance noise interference and accurately locating the corresponding port of the faulty optical cable.

[0092] Furthermore, the calculation of the first interference coefficient of the RRU at the previous moment based on the fluctuation of the normalized optical power value sequence includes:

[0093] Sort the normalized optical power value sequence and take the median as the benchmark;

[0094] Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

[0095] In distributed base station optical cable monitoring, environmental noise and transient disturbances (such as lightning pulses and construction vibrations) can cause significant deviations in short-term optical cable power values. If a fixed-length window is used directly for optical cable fault analysis, it is easily affected by interference, resulting in a large number of false detections.

[0096] To achieve time window adaptation, the window length is automatically set based on the specific disturbance conditions. The window length for the current time is calculated using the disturbance coefficient from the previous time step. The median (a robust statistic against disturbances) of the normalized optical power sequence within the first adaptive time period is taken as the benchmark. The absolute deviation of each data point in the sequence from this median is calculated. The average of all these absolute deviations is then calculated (i.e., the Mean Absolute Deviation (MAD)). This value (MAD) effectively quantifies the overall dispersion and volatility of the RRU optical power sequence at the previous time step; a larger value indicates stronger historical disturbances.

[0097] The specific calculation method is as follows:

[0098] Since the i-th RRU is continuously acquired between the previous and latest times, the optical cable power value does not change much in a short period of time. Therefore, we need to obtain the time period before the previous time for the i-th RRU. Internal interference coefficient Available Adjust the time period of the i-th RRU at the latest time. The length of the time window is adaptive, which improves the response strength to short-term changes in optical cable power value in distributed base station optical cable monitoring. Here, t=2 represents the previous moment and t=3 represents the second moment before.

[0099] Here, the previous time point refers to the time when the received optical power value of the i-th RRU was last collected during the polling process. That is, each time point during the polling process for the received optical power value of the i-th RRU has a corresponding time window length. When the latest time point is the time point when the received optical power value of the i-th RRU was first collected during the polling process, since there is no previous time point, the time window preceding the previous time point is initialized. Internal interference coefficient .

[0100] Get the time period before the previous time step for the i-th RRU. Internal interference coefficient The specific process is as follows:

[0101] Get the time interval of the i-th RRU up to the previous time. The received optical power value at each time step is used to obtain the received optical power value sequence of the i-th RRU. Each received optical power value in the received optical power value sequence of the i-th RRU is then normalized to obtain a sequence of normalized received optical power values ​​for the i-th RRU. ...

[0102] right By sorting in ascending order, we can obtain Sorted sequence Get the sorted sequence the median of This is the median of the i-th RRU at the previous time step. Similarly, we obtain the median of the i-th RRU within the time period. The normalized received optical power value corresponding to the t-th time is itself Where t represents the time period. Traversal of the number of internal samples.

[0103] If the i-th RRU is in the time interval before the previous time... If the normalized received optical power at each time step is close to the median, it indicates that the i-th RRU has achieved a certain level of performance in the time interval up to the previous time step. The lower the disturbance received within, the higher the disturbance received within, indicating that the i-th RRU experiences less disturbance in the time interval before the previous time. The internal interference is relatively high.

[0104] Therefore, calculate the time period before the previous time step for the i-th RRU. Internal interference coefficient :

[0105]

[0106] in, Time period Total number of samplings, t represents the number of samples taken within the time frame. The traversal.

[0107] For the i-th RRU in the time period The normalized received optical power value corresponding to the t-th time.

[0108] It is the median of the i-th RRU at the previous time step.

[0109] The larger the value, the more likely the i-th RRU is to be active within the time period. The normalized received optical power value at time t is significantly different from the median value at the previous time.

[0110] so The larger the value, the more time the i-th RRU has been in the period up to the previous time. The higher the level of interference within.

[0111] Furthermore, the step of dynamically determining the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period includes:

[0112] Set an initial adaptive time period value, and obtain the initial number of samples based on the sampling time interval;

[0113] If the time from the initial start to the current time is greater than or equal to 2, then a preset reference constant is divided by the first interference coefficient of the RRU at the previous time, and then multiplied by the number of samples in the first adaptive time period before the previous time to obtain the number of samples of the RRU in the second adaptive time period before the latest time. This number of samples should not be less than the preset minimum number of samples.

[0114] The length of the second adaptive time period before the latest time of the RRU is obtained by multiplying the number of samples taken by the RRU in the second adaptive time period before the latest time by a fixed sampling time interval value.

[0115] To achieve adaptive time window, an initial time period S is set. The time period S represents the sampling time required when the number of received optical power values ​​is s, and s is the total number of samples within the time period S. An empirical value of 50 is taken, which can be adjusted according to the specific implementation scenario.

[0116] When the time interval is less than 2, the number of samples taken by a single RRU at the latest time interval is the initial number of samples.

[0117] When the time is greater than or equal to 2, the length of the second adaptive time period before the latest time needs to be calculated based on the interference coefficient of the previous time.

[0118] After obtaining the interference coefficient of the i-th RRU at the previous time step Subsequently, to achieve adaptive window shrinking / expansion, it is necessary to calculate the number of samples taken at the latest time step. The specific calculation formula is as follows:

[0119]

[0120] The baseline constant is a preset constant used to stabilize the computational scale, and it is recommended to take the value as the square of the initial number of samplings (i.e., s²).

[0121] The interference coefficient calculated for the i-th RRU at the previous time step;

[0122] For the i-th RRU in the time period Total number of internal samples;

[0123] Minimum number of samplings: A lower limit protection value to prevent the window from being too small. An empirical value of 10 is used, which can be adjusted by the implementer according to the specific real-time scenario.

[0124] floor(·) represents the floor operation, ensuring that the number of samples is an integer.

[0125] Calculated Then, multiplying it by the fixed sampling interval Δt, we can obtain the adaptive time period length of the i-th RRU at the latest time. .

[0126] The above calculation method allows for a shorter window when the interference is greater, and a longer window when the interference is smaller. This solves the problem of "fixed window response lag". A short window when interference is large: it can immediately capture sudden strong interference (such as lightning strikes), improving detection sensitivity and response speed. A long window when interference is small: under stable conditions, using a longer window can better smooth out small random fluctuations, reduce the false alarm rate, and obtain a more stable baseline median.

[0127] Furthermore, the step of obtaining the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculating the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend, includes:

[0128] Calculate the first-order difference of the interference coefficient sequence at the latest moment, and accumulate the positive difference values ​​to obtain the accumulated value of the perturbation intensity of the RRU at the latest moment.

[0129] When calculating the cumulative disturbance intensity of the RRU at the latest moment, the disturbance coefficient sequence is calculated. The difference sequence, i.e., ΔC t = - .

[0130] The cumulative value of the disturbance intensity is obtained by summing all positive values ​​in the difference sequence.

[0131] Technical significance: A positive difference value ΔC indicates that the interference level is more severe at time t compared to time t-1. Accumulating these positive values ​​(Σ max(0, ΔC)... tThis method can efficiently capture the cumulative effect of increasing interference intensity within a window, while ignoring negative values ​​(increased interference), thus focusing on identifying the "deteriorating" trend. Fiber optic cable faults are typically a gradual deterioration process (e.g., the fiber being gradually broken, or the connector being continuously oxidized by water ingress) or a sudden high-level sustained process (e.g., a complete interruption). Both of these situations lead to a continuous or step-incrementing increase in the interference coefficient, resulting in a series of positive first-order differences. By accumulating only positive values ​​and ignoring negative values ​​(the improving trend), focusing on the signal "being damaged," the accumulated disturbance intensity value can highly specifically characterize the fault occurrence process, distinguishing it from transient, recoverable jitter.

[0132] It specifically captures instantaneous changes in "deterioration" or "intensification." The difference is only factored in when the disturbance coefficient is higher in the next moment than in the previous moment. It ignores the recovery and improvement process, focusing only on the cumulative increase in disturbance intensity. The calculation is simple and efficient, with clear directionality, and is highly sensitive to "rising" anomalies, effectively detecting sudden spikes and step-like rising faults.

[0133] Furthermore, the step of determining whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest moment to the average value, and the maximum deviation of the target RRU at the latest moment, includes:

[0134] The risk coefficient of the target RRU at the latest time is calculated based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time.

[0135] Set a risk coefficient threshold;

[0136] When the risk coefficient of the target RRU is greater than or equal to the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station has experienced a cable fault.

[0137] When the risk coefficient of the target RRU is less than the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station is normal.

[0138] The time period before the latest time of the i-th RRU is obtained. The sequence of interference coefficients at each time step Similarly, obtain the sequence of interference coefficients for each time step of the remaining RRUs within the time period preceding the latest time step, where the interference coefficients for the j-th RRU within the time period preceding the latest time step are... The sequence of interference coefficients at each time step is as follows .

[0139] Since the power received values ​​of different RRUs have been normalized to eliminate dimensional interference, if the operating environments of different RRUs are similar, the sequence of interference coefficients for each time period before the latest time of different RRUs should be approximate.

[0140] However, in reality, different RRUs operate in different environments, resulting in differences in their working conditions. Furthermore, when a fault occurs in the optical cable of a distributed base station, the sequence of interference coefficients for each moment in the time period preceding the latest moment for the RRU corresponding to the faulty optical cable differs significantly from the sequence for the other RRUs.

[0141] Furthermore, optical cable fault detection can be performed by analyzing the differences in the sequence of interference coefficients at each time point in the time period preceding the latest time for different RRUs. Adaptive windows can improve the response strength for optical cable fault detection. However, since the optical cable fault does not necessarily occur when the first data in the time window experiences abnormal disturbance, the disturbance intensity of the sequence of interference coefficients at each time point in the time period preceding the latest time for different RRUs is accumulated. Combined with the differences in the sequence of interference coefficients at each time point in the time period preceding the latest time for different RRUs, the risk coefficient of different RRUs at the latest time can be calculated.

[0142] Get the time period of the i-th RRU before the latest time. The sequence of interference coefficients at each time step is as follows ,calculate The first-order difference sequence is used to collect the difference values ​​that are greater than 0, and these values ​​are accumulated to serve as the accumulated perturbation intensity value of the i-th RRU at the latest time. Similarly, the cumulative value of the disturbance intensity of the remaining RRUs at the latest time can be obtained.

[0143] The average value of the cumulative disturbance intensity of the remaining RRUs at the latest time is calculated to obtain the cumulative average disturbance value of the remaining RRUs at the latest time.

[0144] The differences in the sequence of interference coefficients at each time step in the time period before the latest time for different RRUs can be represented by data approximation. However, since the adaptive window lengths corresponding to the time periods before the latest time for different RRUs are different, the lengths of the sequence of interference coefficients at each time step in the time period before the latest time for different RRUs are different. Therefore, the DTW algorithm is used to obtain the DTW distance value of the sequence of interference coefficients at each time step in the time period before the latest time for different RRUs.

[0145] Then, the risk coefficient value of the i-th RRU at the latest time is calculated.

[0146] Introducing two indicators—DTW distance and cumulative disturbance intensity—for a two-factor judgment can effectively distinguish between "global disturbance" and "local fault." For example:

[0147] Scenario A: A fiber optic cable of a certain RRU experiences a fault (partial fault).

[0148] Faulty RRUs: They exhibit significant self-perturbation and behave differently from all normal RRUs. Their DTW distance is also large. Therefore, the combined risk factor calculated based on both factors is also very high.

[0149] Normal RRUs: They have relatively small disturbances. Although they are far from the DTW of the faulty node, the ratio of the cumulative disturbance intensity to the average value is very small. The risk coefficient calculated by combining the two is moderate.

[0150] Result: Only the risk coefficient of the faulty RRU significantly exceeded the standard, and the location was accurate!

[0151] Scenario B: Lightning strike or power fluctuation occurs (global common-mode interference)

[0152] All RRUs have large self-perturbations D (large numerators).

[0153] All RRUs: Since all nodes jitter in a similar pattern at the same time, their interference coefficient sequences are highly similar in shape. Therefore, the DTW distance between any two RRUs will be very small.

[0154] All RRUs have a low risk coefficient.

[0155] Result: The risk coefficients of all RRUs did not reach the threshold, and the system was correctly determined to be fault-free!

[0156] Based on the actual situation, a risk coefficient threshold is set, and optical cable fault detection of distributed base stations is achieved by comparing the thresholds. The risk coefficient value of any RRU at the latest moment is obtained. If the risk coefficient value of any RRU at the latest moment is greater than or equal to the risk coefficient threshold, it indicates that there is an optical cable fault in the optical cable collected by the corresponding RRU; otherwise, it indicates that the optical cable collected by the corresponding RRU is operating normally.

[0157] By calculating the risk coefficient, global disturbances (such as regional heavy rainfall or power grid fluctuations affecting the entire data center) and local disturbances (such as optical cable failures of a single RRU) were effectively identified.

[0158] Furthermore,

[0159] The risk coefficient is calculated using the following formula (1):

[0160] (1)

[0161] in, Let J be the risk coefficient of the i-th RRU at the latest time, J be the total number of RRUs other than the i-th RRU, and j represent the traversal of J;

[0162] This represents the sequence of interference coefficients for the i-th RRU at the latest time. The sequence of interference coefficients of the j-th RRU at the latest time DTW distance between them This represents the maximum deviation of the i-th RRU at the latest time.

[0163] This represents the cumulative disturbance intensity of the i-th RRU at the latest time. This represents the average value of the accumulated disturbance values ​​of all RRUs except the i-th RRU at the latest time.

[0164] The risk coefficient value of the i-th RRU at the latest time is calculated using the following formula (1). :

[0165] (1)

[0166] Where J is the total number of RRUs except for the i-th RRU, and j represents the traversal of J.

[0167] Represents the sequence corresponding to the i-th RRU Sequence corresponding to the j-th RRU The larger the DTW distance value, the greater the correlation between the two sequences. This indicates a greater perturbation difference between the sequences corresponding to the i-th RRU and the j-th RRU in a distributed base station, which in turn indicates a higher probability of an abnormal fault in the optical cable corresponding to the i-th RRU and the j-th RRU. Therefore, the maximum DTW distance value between the sequences corresponding to the i-th RRU and the other RRU sequences is taken. , representing the maximum perturbation risk of the i-th RRU. The method for calculating the DTW distance between two unequal-length sequences is a well-known technique and will not be elaborated further.

[0168] This represents the cumulative disturbance intensity of the i-th RRU at the latest time. The larger the value, the greater the increase in disturbance of the i-th RRU within the corresponding time period at the latest time.

[0169] This represents the cumulative average of the perturbations of all RRUs except the i-th RRU at the latest time. The larger the value, the higher the risk posed by the increase in disturbance of the i-th RRU in the corresponding time period at the latest moment.

[0170] Therefore, the risk coefficient value of the i-th RRU at the latest time. The larger the value, the higher the probability that the i-th RRU in the latest distributed base station's optical cable will experience an optical cable failure at the latest moment.

[0171] An adaptive window mechanism ensures that the window can be shortened immediately in the event of sudden interference, and anomalies can be captured quickly. The final decision mechanism integrates information from two dimensions: vertical (self-change) and horizontal (group comparison). By combining self-trend analysis and group anomaly detection, the double verification greatly reduces the false alarm rate (e.g., normal disturbances common to the entire network will no longer be falsely reported as faults).

[0172] Logically, the risk coefficient of the remaining RRUs at the latest moment can be obtained. A risk coefficient threshold is set, and fiber optic cable fault detection of the distributed base station is achieved through threshold comparison. The risk coefficient threshold is taken as an empirical value of 2.6, which can be adjusted by the implementer according to the specific implementation scenario. This completes the fiber optic cable fault detection of the distributed base station.

[0173] The embodiments disclosed herein can achieve the following:

[0174] Balancing high responsiveness and strong robustness: Through an adaptive time window mechanism, it can respond quickly to short-term strong disturbances and suppress noise interference when the environment is stable, thus resolving the inherent contradiction of the fixed window algorithm.

[0175] Accurately capture cumulative and trend anomalies: By designing the cumulative value of perturbation intensity, it can not only detect abrupt changes at a single point, but also effectively detect hard-to-detect fault precursors such as slow drift, intermittent attacks, or continuous small deterioration.

[0176] Strong common-mode interference resistance: By comparing data across groups using DTW distance, it can effectively distinguish between global interference (such as mains power fluctuations and temperature changes, where all RRU modes are similar) and local faults (where only the faulty RRU mode is abnormal). Only local anomalies will be identified as faults, thus avoiding large-scale false alarms caused by weather, global power issues, etc.

[0177] Reduce false alarms and improve accuracy: Employing multi-dimensional fusion decision-making (individual disturbance + group deviation) is equivalent to setting up double insurance. An RRU will not be easily judged as faulty if it only fluctuates greatly (which may be due to normal sudden business) or if its mode is only slightly different (which may be due to individual device differences). An alarm will only be triggered if both conditions are met, making the diagnostic results highly reliable.

[0178] Suitable for complex real-world environments: The solution processes each RRU independently, does not require data synchronization or window alignment across all devices, and solves the problem of inconsistent sequence lengths through DTW, making the algorithm very suitable for real-world distributed base station scenarios with numerous devices and varying environments.

[0179] Example 2

[0180] This disclosure also provides a specific implementation process for a fiber optic cable fault detection method based on a distributed base station, as follows:

[0181] Thunderstorm occurred (12:00):

[0182] The electromagnetic interference generated by the lightning strike instantly affected a large area, causing severe fluctuations in the received optical power of dozens of RRUs (especially in the eastern area).

[0183] Adaptive window: The interference coefficients of these RRUs increase sharply, causing their time windows to shorten rapidly from 60 to 10. The system enters a high-sensitivity monitoring state.

[0184] DTW Analysis: At this point, all affected RRUs exhibit a highly consistent disturbance pattern (a sudden spike followed by a drop). Therefore, the DTW distance between any two affected RRUs will be very small. max Although it has increased, it will not be abnormally high because it is possible to find "partners" with similar behaviors.

[0185] Risk assessment: Their cumulative disturbance intensity will increase, but DTW max The risk level is not high; the fusion risk coefficient R may reach 1.5 (medium risk), but will not exceed the threshold of 2.6. The system marks it as a "regional common disturbance" but does not generate a fault alarm.

[0186] The fiber optic cable was severed (12:30):

[0187] After the thunderstorm, the optical power of other RRUs quickly returned to normal, and the window gradually lengthened. However, the optical power of RRU_47 permanently dropped to an extremely low level due to the physical interruption (or generated a lot of noise), and its interference coefficient remained high.

[0188] Perturbation accumulation: The perturbation coefficient of RRU_47 remains high, and its perturbation intensity accumulation value continues to accumulate due to the continuous positive change, becoming very large.

[0189] Pattern Deviation: At this point, the interference pattern of RRU_47 is "Permanent Step Change," while the pattern of all other normal RRUs is "Brief Spike Recovery." These two patterns are distinctly different.

[0190] DTW Analysis: Calculating the DTW distance between RRU_47 and any other normal RRU would result in a very large distance because an extreme twist would be required to match a straight line (post-fault) with a brief pulse line (normal). Therefore, the DTW distance of RRU_47 is... max The value became abnormally high.

[0191] Risk Assessment: The ratio of the cumulative disturbance intensity of RRU_47 at the latest moment to the average of the cumulative disturbance intensity of other RRUs, and the maximum deviation at the latest moment, are both much greater than 1. Substituting these values ​​into the risk formula, the calculated value is 20 >> 2.6, and the system immediately generates the highest level alarm: "Fiber optic cable fault, located at RRU_47".

[0192] Maintenance personnel could clearly see a single, precisely located alarm on the central control room's large screen, instead of hundreds of generic "abnormal optical power" alarms. This allowed for the immediate dispatch of a repair team to the location of RRU_47, significantly improving efficiency.

[0193] Example 3

[0194] This disclosure also provides a fiber optic cable fault detection system based on a distributed base station, such as... Figure 2 As shown, the system includes:

[0195] The acquisition and processing module 11 is configured to acquire the received optical power values ​​of multiple RRUs and normalize the received optical power values ​​of each RRU.

[0196] The first calculation module 12 is configured to obtain, for each RRU, a normalized optical power value sequence within a first adaptive time period corresponding to the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence.

[0197] The acquisition module 13 is configured to dynamically determine the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous time and the corresponding first adaptive time period, acquire the interference coefficient of the RRU at each time within the second adaptive time period, and form the interference coefficient sequence of the RRU at the latest time.

[0198] The second calculation module 14 is configured to obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend.

[0199] The third calculation module 15 is configured to calculate the DTW distance between the interference coefficient sequence of the target RRU at the latest time and the interference coefficient sequence of each of the other RRUs at the latest time, so as to obtain the maximum deviation of the target RRU at the latest time.

[0200] The judgment module 16 is configured to calculate the average value of the cumulative disturbance intensity of the remaining RRUs other than the target RRU at the latest time, and determine whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value, and the maximum deviation of the target RRU at the latest time.

[0201] Furthermore, the first calculation module 12 is specifically configured as follows:

[0202] Sort the normalized optical power value sequence and take the median as the benchmark;

[0203] Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

[0204] Furthermore, the acquisition module 13 is specifically configured as follows:

[0205] Set an initial adaptive time period value, and obtain the initial number of samples based on the sampling time interval;

[0206] If the time from the initial start to the current time is greater than or equal to 2, then a preset reference constant is divided by the first interference coefficient of the RRU at the previous time, and then multiplied by the number of samples in the first adaptive time period before the previous time to obtain the number of samples of the RRU in the second adaptive time period before the latest time. This number of samples should not be less than the preset minimum number of samples.

[0207] The length of the second adaptive time period before the latest time of the RRU is obtained by multiplying the number of samples taken by the RRU in the second adaptive time period before the latest time by a fixed sampling time interval value.

[0208] Furthermore, the second calculation module 14 is specifically configured as follows:

[0209] Calculate the first-order difference of the interference coefficient sequence at the latest moment, and accumulate the positive difference values ​​to obtain the accumulated value of the perturbation intensity of the RRU at the latest moment.

[0210] Furthermore, the judgment module 16 is specifically configured as follows:

[0211] The risk coefficient of the target RRU at the latest time is calculated based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time.

[0212] Set a risk coefficient threshold;

[0213] When the risk coefficient of the target RRU is greater than or equal to the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station has experienced a cable fault.

[0214] When the risk coefficient of the target RRU is less than the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station is normal.

[0215] Furthermore,

[0216] The risk coefficient is calculated using the following formula (1):

[0217] (1)

[0218] in, Let J be the risk coefficient of the i-th RRU at the latest time, J be the total number of RRUs other than the i-th RRU, and j represent the traversal of J;

[0219] This represents the sequence of interference coefficients for the i-th RRU at the latest time. The sequence of interference coefficients of the j-th RRU at the latest time DTW distance between them This represents the maximum deviation of the i-th RRU at the latest time.

[0220] This represents the cumulative disturbance intensity of the i-th RRU at the latest time. This represents the average value of the accumulated disturbance values ​​of all RRUs except the i-th RRU at the latest time.

[0221] The optical cable fault detection system based on distributed base stations in this disclosure is used to implement the optical cable fault detection method based on distributed base stations in embodiments one and two. Therefore, the description is relatively simple. For details, please refer to the relevant descriptions in the previous method embodiments, which will not be repeated here.

[0222] In addition, such as Figure 3 As shown, Embodiment 4 of this disclosure also provides an electronic device, including a memory 100 and a processor 200. The memory 100 stores a computer program. When the processor 200 runs the computer program stored in the memory 100, the processor 200 executes the various possible methods described above.

[0223] The memory 100 is connected to the processor 200. The memory 100 can be a flash memory, a read-only memory, or another type of memory. The processor 200 can be a central processing unit or a microcontroller.

[0224] Furthermore, embodiments of this disclosure also provide a computer-readable storage medium storing a computer program, which is executed by a processor using the various possible methods described above.

[0225] The computer-readable storage medium includes volatile or non-volatile, removable or non-removable media implemented in any method or technology for storing information, such as computer-readable instructions, data structures, computer program modules or other data. Computer-readable storage media include, but are not limited to, RAM (Random Access Memory), ROM (Read-Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), flash memory or other memory technologies, CD-ROM (Compact Disc Read-Only Memory), Digital Video Disc (DVD) or other optical disc storage, cartridges, magnetic tapes, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and is accessible to a computer.

[0226] It is understood that the above embodiments are merely exemplary embodiments used to illustrate the principles of this disclosure, and this disclosure is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and substance of this disclosure, and these modifications and improvements are also considered to be within the scope of protection of this disclosure.

Claims

1. A method for optical cable fault detection based on distributed base stations, characterized in that, The method includes: Collect the received optical power values ​​of multiple remote radio frequency units (RRUs) and normalize the received optical power value of each RRU. For each RRU, obtain its normalized optical power value sequence within the first adaptive time period before the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence. Based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period, the length of the second adaptive time period before the latest moment of the RRU is dynamically determined, and the interference coefficient of the RRU at each moment within the second adaptive time period is obtained to form the interference coefficient sequence of the RRU at the latest moment; and, Obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend; Calculate the dynamic time-normalized (DTW) distance between the target RRU's interference coefficient sequence at the latest time and the interference coefficient sequences of each of the other RRUs at the latest time, and obtain the maximum deviation of the target RRU at the latest time. Calculate the average value of the cumulative disturbance intensity of the remaining RRUs other than the target RRU at the latest time. Determine whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time.

2. The method according to claim 1, characterized in that, The calculation of the first interference coefficient of the RRU at the previous moment based on the fluctuation of the normalized optical power value sequence includes: Sort the normalized optical power value sequence and take the median as the benchmark; Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

3. The method according to claim 2, characterized in that, The step of dynamically determining the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous moment and the corresponding first adaptive time period includes: Set an initial adaptive time period value, and obtain the initial number of samples based on the sampling time interval; If the time from the initial start to the current time is greater than or equal to 2, then a preset reference constant is divided by the first interference coefficient of the RRU at the previous time, and then multiplied by the number of samples in the first adaptive time period before the previous time to obtain the number of samples of the RRU in the second adaptive time period before the latest time. This number of samples should not be less than the preset minimum number of samples. The length of the second adaptive time period before the latest time of the RRU is obtained by multiplying the number of samples taken by the RRU in the second adaptive time period before the latest time by a fixed sampling time interval value.

4. The method according to claim 1, characterized in that, The step of obtaining the changing trend of the interference coefficient sequence of the RRU at the latest time and calculating the cumulative value of the perturbation intensity of the RRU at the latest time based on the changing trend includes: Calculate the first-order difference of the interference coefficient sequence at the latest moment, and accumulate the positive difference values ​​to obtain the accumulated value of the perturbation intensity of the RRU at the latest moment.

5. The method according to claim 1, characterized in that, The step of determining whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest moment to the average value, and the maximum deviation of the target RRU at the latest moment, includes: The risk coefficient of the target RRU at the latest time is calculated based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value and the maximum deviation of the target RRU at the latest time. Set a risk coefficient threshold; When the risk coefficient of the target RRU is greater than or equal to the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station has experienced a cable fault. When the risk coefficient of the target RRU is less than the risk coefficient threshold at the latest moment, it indicates that the optical cable of the target RRU in the distributed base station is normal.

6. The method according to claim 5, characterized in that, The risk coefficient is calculated using the following formula (1): （1） in, Let J be the risk coefficient of the i-th RRU at the latest time, J be the total number of RRUs other than the i-th RRU, and j represent the traversal of J; This represents the sequence of interference coefficients for the i-th RRU at the latest time. The sequence of interference coefficients of the j-th RRU at the latest time DTW distance between them This represents the maximum deviation of the i-th RRU at the latest time. This represents the cumulative disturbance intensity of the i-th RRU at the latest time. This represents the average value of the accumulated disturbance values ​​of all RRUs except the i-th RRU at the latest time.

7. A fiber optic cable fault detection system based on distributed base stations, characterized in that, The system includes: The acquisition and processing module is configured to acquire the received optical power values ​​of multiple RRUs and normalize the received optical power values ​​of each RRU. The first calculation module is configured to, for each RRU, obtain the normalized optical power value sequence within the first adaptive time period corresponding to the previous moment, and calculate the first interference coefficient of the RRU in the previous moment based on the fluctuation of the normalized optical power value sequence. The acquisition module is configured to dynamically determine the length of the second adaptive time period before the latest time of the RRU based on the first interference coefficient of the RRU at the previous time and the corresponding first adaptive time period, acquire the interference coefficient of the RRU at each time within the second adaptive time period, and form the interference coefficient sequence of the RRU at the latest time. The second calculation module is configured to obtain the changing trend of the interference coefficient sequence of the RRU at the latest time, and calculate the cumulative value of the disturbance intensity of the RRU at the latest time based on the changing trend. The third calculation module is set to calculate the DTW distance between the interference coefficient sequence of the target RRU at the latest time and the interference coefficient sequence of each of the other RRUs at the latest time, so as to obtain the maximum deviation of the target RRU at the latest time. The judgment module is configured to calculate the average value of the cumulative disturbance intensity of the remaining RRUs (excluding the target RRU) at the latest time, and determine whether the optical cable corresponding to the target RRU is faulty based on the ratio of the cumulative disturbance intensity of the target RRU at the latest time to the average value, and the maximum deviation of the target RRU at the latest time.

8. The system according to claim 7, characterized in that, The first calculation module is specifically configured as follows: Sort the normalized optical power value sequence and take the median as the benchmark; Calculate the mean absolute deviation of each normalized received optical power value in the normalized optical power value sequence from the reference, and use the mean absolute deviation as the first interference coefficient of the RRU at the previous moment.

9. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the optical cable fault detection method based on a distributed base station as described in any one of claims 1-6.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the optical cable fault detection method based on a distributed base station as described in any one of claims 1-6.

Images