Signal fast fading identification method, device, equipment, medium and program product

Abstract

The application relates to the technical field of communication, and provides a signal fast-fading identification method, device, equipment, medium and program product. The method comprises the following steps: acquiring MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data; performing sliding window processing on the wireless signal data based on a first window size, performing coarse identification on a first signal segment in a current first window, and determining whether the first signal segment has fast fading; if the first signal segment has fast fading, performing sliding window processing on the first signal segment based on a second window size, and performing mutation detection on a second signal segment in each second window by using a Mann-Kendall algorithm, so as to determine a fast-fading time period in the first signal segment; and the second window size is smaller than the first window size. Through the double sliding window processing and the mutation detection of the Mann-Kendall algorithm, the misjudgment of fast fading caused by instantaneous signal change is avoided, and the accuracy of fast-fading identification is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of communication technology, in particular to a signal fast fading identification method, device, equipment, medium and program product. BACKGROUND

[0002] Fast Fading refers to the phenomenon that the received signal strength of a radio signal changes rapidly in a short time, mainly caused by multipath propagation and Doppler effect. This phenomenon usually occurs in mobile communication systems. When the scattering body, reflector and refractor near the mobile station cause multipath transmission of the signal, these multipath signals superimpose on each other at the receiving point, resulting in rapid fluctuations in the amplitude of the received signal. Fast Fading is an important characteristic of wireless signal propagation. In daily network optimization work, it is difficult to quickly and intuitively locate the fast fading problem point through basic data. Usually, user feedback or active network testing is required to find out, which often causes the network optimization work to be lagging. For the phenomenon of fast fading of wireless signals, a large amount of manpower and time is needed for on-site testing, and daily network testing cannot traverse all network coverage areas, and the comprehensiveness of fast fading identification is insufficient. At present, a more comprehensive way to identify fast fading areas is to compare the reference signal powers of adjacent data points in the MRO (Measurement Report Original, measurement report original data) data reported by the connected UE (User Equipment, user equipment). When the reference signal power fluctuates, the UE is determined to be a fast fading UE. The UE reports instantaneous signal strength, and the instantaneous fluctuation of the radio signal is not necessarily a fast fading phenomenon, so there is a high possibility of misjudgment. SUMMARY

[0003] The present application provides a signal fast fading identification method, device, equipment, medium and program product to solve the defect that the existing fast fading identification method is easy to misjudge fast fading.

[0004] The present application provides a signal fast fading identification method, comprising:

[0005] Obtaining MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data;

[0006] Based on a first window size, performing sliding window processing on the wireless signal data, and performing coarse identification on a first signal segment in the current first window to determine whether the first signal segment has fast fading;

[0007] If the first signal segment has fast fading, the first signal segment is processed by a sliding window based on a second window size, and a Mann-Kendall algorithm is used to detect mutation of a second signal segment in each second window to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

[0008] In one embodiment, the Mann-Kendall algorithm is used to detect mutation of the second signal segment in the current second window to determine the fast fading time period in the first signal segment, comprising:

[0009] The Mann-Kendall algorithm is used to calculate a first order sequence based on an order time sequence, and a first statistical quantity sequence of the first order sequence is calculated according to a preset first statistical quantity algorithm; the order time sequence is composed of signal strengths of sampling points in the second signal segment;

[0010] A second order sequence is calculated based on a reverse order time sequence of the order time sequence, and a second statistical quantity sequence of the second order sequence is calculated according to a preset second statistical quantity algorithm;

[0011] The signal change trend of the second signal segment is determined according to the first statistical quantity sequence, and whether there is fast fading of the signal in the second signal segment is determined based on the signal change trend;

[0012] If there is fast fading of the signal in the second signal segment, the fading start time is determined according to the intersection of the first statistical quantity sequence and the second statistical quantity sequence;

[0013] The fast fading time period in the first signal segment is determined according to the fading start time.

[0014] In one embodiment, the Mann-Kendall algorithm is used to calculate a first order sequence based on an order time sequence, and a first statistical quantity sequence of the first order sequence is calculated according to a preset first statistical quantity algorithm, comprising:

[0015] The first sequence value in the order time sequence is compared with each second sequence value, and a first feature sequence corresponding to the first sequence value is constructed according to the comparison result; wherein, if the comparison result represents that the first sequence value is greater than the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a first preset value; if the comparison result represents that the first sequence value is less than or equal to the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a second preset value; the first sequence value is any one in the order time sequence, and the second sequence value is any one before the first sequence value;

[0016] construct a first rank sequence according to the first target number corresponding to each of the first sequence values based on a first target number of second sequence values less than the first sequence values in the first feature sequence;

[0017] calculate a feature value corresponding to the first rank sequence; the feature value includes a mean value and a variance value;

[0018] calculate each sequence value in the first rank sequence and the feature value according to a preset first statistical quantity algorithm to obtain a first statistical quantity sequence.

[0019] In one embodiment, the second rank sequence is calculated based on the reverse sequence of the sequence time sequence, and a second statistical quantity sequence of the second rank sequence is calculated according to a preset second statistical quantity algorithm, including:

[0020] compare a third sequence value with each fourth sequence value in the reverse sequence of the sequence time sequence, and construct a second feature sequence according to a comparison result; if the comparison result indicates that the third sequence value is greater than the fourth sequence value, a sequence value corresponding to the fourth sequence value in the second feature sequence is a first preset value; if the comparison result indicates that the third sequence value is less than or equal to the fourth sequence value, a sequence value corresponding to the fourth sequence value in the first feature sequence is a second preset value; the third sequence value is any one in the reverse sequence of the sequence time sequence, and the fourth sequence value is any one before the first sequence value;

[0021] construct a second rank sequence according to a second target number corresponding to each of the third sequence values based on a second target number of fourth sequence values less than the third sequence values in the second feature sequence;

[0022] calculate a feature value corresponding to the second rank sequence; the feature value includes a mean value and a variance value;

[0023] calculate each sequence value in the second rank sequence and the feature value according to a preset second statistical quantity algorithm to obtain a first statistical quantity sequence.

[0024] In one embodiment, the first signal segment in the current first window is coarsely identified to determine whether the first signal segment has fast fading, including:

[0025] coarsely identify the first signal segment in the current first window to determine a target sampling point with a signal strength less than a preset signal threshold in the first signal segment;

[0026] based on a target moment corresponding to the target sampling point, obtain a forward sampling point with a first preset time length before the target sampling point and a backward sampling point with a second preset time length after the target sampling point;

[0027] calculate an average value of signal strengths of the forward sampling points based on the signal strengths corresponding to the forward sampling points, to obtain a forward average level of the target sampling point, and calculate an average value of signal strengths of the backward sampling points and the target sampling point based on the signal strengths corresponding to the backward sampling points and the target sampling point, to obtain a backward average level of the target sampling point;

[0028] determine whether the first signal segment has fast fading based on the backward average level and the forward average level; if the backward average level is less than the preset signal threshold, and a difference between the backward average level and the forward average level is greater than a preset level value, the first signal segment has fast fading.

[0029] In one embodiment, the extracting of the wireless signal data of each terminal user from the MRO data comprises:

[0030] extracting an IMSI identifier of each terminal user from the MRO data;

[0031] performing reverse filling on feature data in the MRO data based on the IMSI identifier, to obtain filled feature data; the feature data comprises at least one of a base station number, a cell number, an AMF area number, an AMF set number, an AMF pointing number, a collection time identifier, a measurement object, a measurement signal strength, a physical cell identifier, and a frequency point number;

[0032] extracting the wireless signal data of each terminal user from the filled feature data according to the IMSI identifier.

[0033] The application also provides a signal fast fading identification device, comprising the following modules:

[0034] a data collection module, configured to acquire MRO data collected by a base station, and extract wireless signal data of each terminal user from the MRO data;

[0035] a first identification module, configured to perform sliding window processing on the wireless signal data based on a first window size, and perform coarse identification on a first signal segment in a current first window, to determine whether the first signal segment has fast fading;

[0036] a second identification module, configured to, if the first signal segment has fast fading, perform sliding window processing on the first signal segment based on a second window size, and perform mutation detection on a second signal segment in each second window by using a Mann-Kendall algorithm, to determine a fast fading time segment in the first signal segment; the second window size is smaller than the first window size.

[0037] The application further provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the signal fast fading identification method according to any one of the above when executing the computer program.

[0038] The application further provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the signal fast fading identification method according to any one of the above.

[0039] The application further provides a computer program product, including a computer program, and the computer program is executed by a processor to implement the signal fast fading identification method according to any one of the above.

[0040] The signal fast fading identification method, device, equipment, medium and program product provided by the application perform coarse identification of fast fading of a wireless signal from a user granularity through a sliding window processing manner, and further perform mutation detection on the signal segment with fast fading obtained through coarse identification through a sliding window processing manner and using the Mann-Kendall algorithm to determine a fast fading time period. Through sliding window processing, the Mann-Kendall algorithm is used to further perform mutation detection on the fast fading identified through coarse identification on the basis of coarse identification, so that misjudgment of fast fading caused by instantaneous change of a signal is avoided, and the identification accuracy of fast fading is improved. BRIEF DESCRIPTION OF DRAWINGS

[0041] In order to more clearly illustrate the technical solutions of the application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and those skilled in the art can also obtain other drawings according to these drawings without creative labor.

[0042] Figure 1 is a flowchart of the signal fast fading identification method provided by the embodiments of the application.

[0043] Figure 2 is a structural diagram of the signal fast fading identification device provided by the application.

[0044] Figure 3 is a structural diagram of the electronic device provided by the embodiments of the application. DETAILED DESCRIPTION

[0045] In order to make the purposes, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the present application. Obviously, the described embodiments are only some of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work fall within the scope of protection of the present application.

[0046] Figure 1 is a flowchart of a signal fast fading identification method provided by an embodiment of the present application, as shown in Figure 1 the method comprises the following steps:

[0047] Step 100, acquiring MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data;

[0048] Step 200, performing sliding window processing on the wireless signal data based on a first window size, and performing coarse identification on a first signal segment in a current first window to determine whether the first signal segment has fast fading;

[0049] Step 300, if the first signal segment has fast fading, performing sliding window processing on the first signal segment based on a second window size, and performing mutation detection on a second signal segment in a current second window by using a Mann-Kendall algorithm to determine a fast fading time period in the second signal segment; the second window size is smaller than the first window size.

[0050] Firstly, MRO data collected by a base station is acquired, the MRO data can be reported by a user terminal UE, and the MRO data contains signal level or signal quality information of a cell where the user terminal is located. Wireless signal data of each terminal user is extracted from the MRO data, the wireless signal data includes signal level and / or signal quality, and the MRO data can be extracted according to a device identifier of the user terminal, the MRO data of the same terminal user is extracted and converged, the MRO data of each terminal user is obtained, and the wireless signal data is extracted therefrom.

[0051] Further, the wireless signal data of any terminal user is processed by sliding window based on a first window size, and a first signal segment in a current first window is coarsely identified to determine whether the first signal segment has fast fading. The identification manner of the coarse identification on the first signal segment is not specifically limited, and the change of signal strength in the first signal segment can be detected. If the signal strength fluctuates greatly in a short time, it is considered that the first signal segment has fast fading.

[0052] Optionally, the first window size can be determined based on a sampling rate of the MRO data. Optionally, the first window size is inversely proportional to the sampling rate of the MRO data, the greater the sampling rate of the MRO data, the smaller the first window size, and the smaller the sampling rate of the MRO data, the greater the first window size. Optionally, the first window size can also be a preset first fixed size.

[0053] Optionally, in the process of performing the sliding window processing based on the first window size, the sliding step of the first window can be determined according to the first window size, or can be a preset first sliding step. For example, if the non-overlapping sliding window processing is performed on the wireless signal data, the first sliding step is the same as the window width in the first window size, and if the overlapping sliding window processing is performed on the wireless signal data, the first sliding step is smaller than the window width in the first window size.

[0054] In an embodiment, the first window size and the first sliding step of the first window can be determined according to a detection period of the fast fading, that is, the signal segment in the first window is composed of data points of the wireless signal data collected from a time length corresponding to one detection period.

[0055] On the basis of the coarse identification of the first signal segment, the first signal segment is processed by a sliding window based on a second window size, a Mann-Kendall (MK) algorithm is used to detect a mutation of a second signal segment in a current second window, and a fast fading time period in the second signal segment is determined. The second window size is smaller than the first window size, and optionally, a second sliding step of the second window is smaller than the first sliding step.

[0056] The MK algorithm is a non-parametric statistical test method, the sample data points do not need to comply with a specific distribution, and are not disturbed by a small number of abnormal values, and is more suitable for type variables and order variables.

[0057] Optionally, the second window size can be a preset second fixed size, can be determined according to the first window size, or can be determined according to a sampling rate of the MRO data. In an embodiment, the first window size is an integer multiple of the second window size, and the signal length of the first signal segment is an integer multiple of the length of the second signal segment.

[0058] In an embodiment, the MK algorithm is used to detect a mutation of the second signal segment in each second window, to determine whether there is fast fading in the second signal segment, thereby further accurately identifying the fast fading coarsely identified in the first signal segment, and avoiding misjudgment. Based on the mutation detection of the second signal segment in the second window, when the second signal segment with the fast fading is detected, the fast fading time period of the first signal segment is determined according to the second signal segment with the fast fading.

[0059] In this way, the fast fading of the wireless signal of any terminal user can be identified from the user granularity, and on this basis, the number of users in any cell experiencing fast fading in any time period can be determined according to the cell where the user terminal is located. According to the number of users experiencing fast fading in the cell, it can be determined whether the fast fading in the cell is caused by a few users or caused by the cell, so that the network optimization strategy can be better formulated for network optimization.

[0060] In the embodiment, the fast fading of the wireless signal is coarsely identified from the user granularity by the sliding window processing, and further, the signal segment experiencing fast fading coarsely identified is subjected to mutation detection by the Mann-Kendall algorithm to determine the fast fading time period by the sliding window processing. By the sliding window processing, the fast fading coarsely identified is further subjected to mutation detection by the Mann-Kendall algorithm on the basis of the coarse identification, which avoids the misjudgment of the fast fading caused by the instantaneous change of the signal and improves the identification accuracy of the fast fading.

[0061] In one embodiment, the extraction of the wireless signal data of each terminal user is implemented based on the reverse filling of the unique identification IMSI (International Mobile Subscriber Identity) of the user. In step 100, the wireless signal data of each terminal user is extracted from the MRO data, which can further include:

[0062] Step 101, extracting the IMSI identification of each terminal user from the MRO data.

[0063] Step 102, reverse filling the feature data in the MRO data based on the IMSI identification to obtain the filled feature data; the feature data includes at least one of the base station number, the cell number, the AMF area number, the AMF set number, the AMF pointing number, the collection time identification, the measurement object, the measurement signal strength, the physical cell identification and the frequency point number.

[0064] Step 103, extracting the wireless signal data of each terminal user from the filled feature data according to the IMSI identification.

[0065] Firstly, the IMSI identifier of each terminal user is extracted from the MRO data, the feature data in the MRO data is reversely filled based on the IMSI identifier, so as to associate the AMF (‌Access and Mobility Management Function, access and mobility management function) identifier, time identifier and AMF UE NGAP ID and other data in the measurement report (Measurement Report, MR) of the terminal user, so as to achieve the purpose of accurate matching of wireless data and core network data. Among them, NGAP ID is Non-Access Stratum Protocol Identification, AMF UE NGAP ID is the NGAP identifier between AMF and UE, and in the 5G network, AMF UE NGAP ID is used to uniquely identify a specific communication session between UE and AMF.

[0066] The feature data in the MRO data is reversely filled based on the IMSI identifier, and the feature data filled with the IMSI identifier is obtained. The feature data includes at least one of the base station number, the cell number, the AMF area number, the AMF set number, the AMF pointing number, the collection time identifier, the measurement object, the measurement signal strength, the physical cell identifier and the frequency point number.

[0067] Based on the filled IMSI identifier, the wireless signal data of each terminal user is extracted from the filled feature data. Exemplarily, for the wireless signal data of different terminal users, the user identifier is used as the data set label of the wireless signal data, so as to distinguish the data sets of different terminal users. Optionally, the wireless signal data of the terminal user at least includes the base station identifier of the base station where the user terminal is located, the latitude and longitude of the user data occurrence location, and the time sequence of the reference signal received power (RSRP) of the user terminal based on the time sequence.

[0068] Further, the sliding window processing of the wireless signal data of any terminal user, that is, the sliding window processing of the time sequence of the RSRP of the terminal user, in step 200, the first signal segment in the current first window is coarsely identified to determine whether there is fast fading in the first signal segment. It can also include:

[0069] In step 201, the first signal segment in the current first window is coarsely identified to determine the target sampling point with a signal strength less than a preset signal threshold in the first signal segment;

[0070] In step 202, based on the target time corresponding to the target sampling point, the forward sampling point with a first preset time length before the target sampling point and the backward sampling point with a second preset time length after the target sampling point are obtained;

[0071] In step 203, based on the signal strength corresponding to the forward sampling point, an average value of the signal strength of the forward sampling point is calculated to obtain a forward average level of the target sampling point, and based on the signal strength corresponding to the backward sampling point and the target sampling point, an average value of the signal strength of the backward sampling point and the target sampling point is calculated to obtain a backward average level of the target sampling point.

[0072] In step 204, according to the backward average level and the forward average level, it is determined whether the first signal segment has fast fading; if the backward average level is less than the preset signal threshold, and the difference between the backward average level and the forward average level is greater than a preset level value, then the first signal segment has fast fading.

[0073] The first signal segment in the current first window is coarsely identified to determine a target sampling point in the first signal segment whose signal strength is less than a preset signal threshold, based on a target time corresponding to the target sampling point, a forward sampling point before the target sampling point for a first preset time length is obtained, and a backward sampling point after the target sampling point for a second preset time length is obtained. Optionally, the first preset time length can be the same or different, and no specific limitation is made thereto.

[0074] Further, when a sampling point whose signal strength is less than a preset signal threshold is obtained, a sampling point before the sampling point for a first preset time length and a sampling point after the sampling point for a second preset time length are obtained, and according to the obtained forward sampling point and the backward sampling point, it is determined whether the first signal segment has fast fading at the sampling point.

[0075] In one embodiment, the first preset time length and the second preset time length are the same, and the first preset time length and the second preset time length respectively correspond to one period, when a target sampling point whose signal strength is less than a preset signal threshold is obtained, a forward sampling point before the target sampling point for one period and a backward sampling point after the target sampling point for one period are obtained, and according to the obtained forward sampling point and the backward sampling point, it is determined whether the target sampling point is a fast fading point. If the target sampling point is a fast fading point, then the first signal segment has fast fading, and if the target sampling point is not a fast fading point, then the first signal segment does not have fast fading.

[0076] Optionally, the target sampling point can include one or more, and when there is at least one target sampling point that is a fast fading point, then the first signal segment has fast fading.

[0077] Further, the forward sampling points include one or more sampling points, and the backward sampling points also include one or more sampling points, the average value of the signal strength of the forward sampling points is calculated based on the signal strength of each forward sampling point, to obtain the forward average level of the target sampling point; the average value of the signal strength of the backward sampling points and the target sampling point is calculated based on the signal strength of each backward sampling point and the signal strength of the target sampling point, to obtain the backward average level of the target sampling point.

[0078] For any target sampling point, the forward sampling points and the backward sampling points thereof are obtained, and the forward average level and the backward average level thereof are calculated, and whether the target sampling point is a fast fading point is determined according to the calculated forward average level and the backward average level. Specifically, if the backward average level is less than a preset signal threshold, and the difference between the backward average level and the forward average level is greater than a preset level value, the target sampling point is a fast fading point, and the first signal segment has fast fading.

[0079] Exemplarily, two sampling points correspond to one period, -110 dB is the preset signal threshold, in the first signal segment, if the time corresponding to the sampling point with the signal strength less than -110 dB is t=m+2, and the corresponding signal strength is RSRP m+2 , the forward sampling point is obtained by calculating (RSRP m +RSRP m+1 ) / 2, and the forward average level is calculated; the backward sampling point is obtained by calculating (RSRP m+2 +RSRP m+3 ) / 2, and the backward average level is calculated. If the backward average level is less than -110 dB, and the difference between the backward average level and the forward average level is greater than the preset level value 10 dB, it is considered that the time t=m+2 is a weak coverage point caused by fast fading, that is, a fast fading point.

[0080] Further, the mutation detection of the MK algorithm is realized based on the construction of an order sequence on a time sequence, the wireless signal data of the user terminal contains the signal strength of the user terminal, which can be represented by the RSRP (Reference Signal Receiving Power) level value of the user terminal in the wireless signal data. The wireless signal data of the user terminal is a time sequence composed of sampling points at different times, in step 300, the Mann-Kendall algorithm is used to detect the mutation of the second signal segment in each second window to determine the fast fading time period in the first signal segment, which can also include:

[0081] In step 301, the first rank sequence is calculated based on the sequential time sequence by using the Mann-Kendall algorithm, and the first statistic sequence of the first rank sequence is calculated according to a preset first statistic algorithm; the sequential time sequence is composed of the signal intensity of each sampling point in the second signal segment.

[0082] In step 302, the second rank sequence is calculated based on the reverse sequential time sequence of the sequential time sequence, and the second statistic sequence of the second rank sequence is calculated according to a preset second statistic algorithm.

[0083] In step 303, the signal change trend of the second signal segment is determined according to the first statistic sequence, and whether there is a signal fast fading in the second signal segment is determined based on the signal change trend.

[0084] In step 304, if there is a signal fast fading in the second signal segment, the fading starting time is determined according to the intersection point of the first statistic sequence and the second statistic sequence.

[0085] In step 305, the fast fading time period in the first signal segment is determined according to the fading starting time.

[0086] The first rank sequence is calculated based on the sequential time sequence by using the MK algorithm, and the first statistic sequence of the first rank sequence is calculated according to a preset first statistic algorithm, wherein the sequential time sequence is a time sequence composed of the signal intensity of each sampling point in the second signal segment.

[0087] Further, the second rank sequence is calculated based on the reverse sequential time sequence of the sequential time sequence, and the second statistic sequence of the second rank sequence is calculated according to a preset second statistic algorithm. The signal change trend of the second signal segment is determined according to the first statistic sequence, and whether there is a signal fast fading in the second signal segment is determined based on the signal change trend.

[0088] In the case that there is a signal fast fading in the second signal segment, the fading starting time is determined according to the intersection point of the first statistic sequence and the second statistic sequence, and the fast fading time period in the first signal segment is determined according to the fading starting time.

[0089] The mutation test of the MK algorithm is a non-parametric hypothesis test method for testing the trend change of data in a time series. The algorithm detects the monotonic trend (upward, downward or no trend) in the time series by comparing the size of each data point with that of its previous data point. Specifically, the MK algorithm compares each data point in the time series with all previous data points, calculates the number of data points that are smaller than the current data point, and the number of data points that are larger than the current data point, then compares the size relationship of the two data point numbers, and determines whether there is a monotonic trend.

[0090] Based on this, in step 301, a first rank sequence is calculated based on the sequential time series, and a first statistical quantity sequence of the first rank sequence is calculated according to a preset first statistical quantity algorithm, which can also include:

[0091] Step 311, comparing the first sequence value and each second sequence value in the sequential time series, and constructing a first feature sequence corresponding to the first sequence value according to the comparison result; wherein if the comparison result represents that the first sequence value is greater than the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a first preset value; if the comparison result represents that the first sequence value is less than or equal to the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a second preset value; the first sequence value is any one in the sequential time series, and the second sequence value is any one before the first sequence value;

[0092] Step 321, based on the first feature sequence, the first target number of second sequence values less than the first sequence value is counted, and a first rank sequence is constructed according to the first target number corresponding to each first sequence value;

[0093] Step 331, calculating the feature value corresponding to the first rank sequence; the feature value includes mean and variance;

[0094] Step 341, according to the preset first statistical quantity algorithm, calculating each sequence value in the first rank sequence and the feature value respectively to obtain a first statistical quantity sequence.

[0095] If the time series of the second signal segment extracted from the MRO data of a single terminal user is: , n is the number of sampling points, and the rank sequence is expressed as The cumulative number of sampling points of is defined as:

[0096] ; (1)

[0097] When constructing the first ordered sequence, the first sequence value in the sequential time series is first compared with each second sequence value, and a first feature sequence is constructed based on the comparison result. If the comparison result indicates that the first sequence value is greater than the second sequence value, then the sequence value corresponding to the first sequence value in the first feature sequence is the first preset value. If the comparison result indicates that the first sequence value is less than or equal to the second sequence value, then the sequence value corresponding to the first sequence value in the first feature sequence is the second preset value. The first sequence value is any sequence value in the sequential time series, and the second sequence value is any sequence value in the sequential time series preceding the first sequence value.

[0098] The first characteristic sequence constructed is the ordered sequence. Based on the first feature sequence, and following the method shown in Formula 1, the number of second sequence values ​​less than the first sequence value is counted as the first target number. The first target quantity is statistically determined as the first sequence value and the corresponding sequence value in the first order column. Thus, each sequence value in the sequential time series can be used as the first sequence value. According to the method shown in Formula 1, the sequence values ​​corresponding to each sequence value in the sequential time series in the first order column are statistically obtained, and the first order column is constructed.

[0099] Further, the characteristic values ​​corresponding to the first order column are calculated, including the mean and variance. According to the preset first statistical algorithm, the sequence values ​​and characteristic values ​​in the second order column are calculated respectively to obtain the first statistical sequence.

[0100] Optionally, the algorithm for the first statistic is defined under the assumption that the time series are random and independent, as shown in Equation 2:

[0101] (2)

[0102] Wherein, the initial value of the first statistic sequence , and These are the sequence values ​​in the first order column. The mean and variance of, in When the components are independent and have the same continuous distribution, then:

[0103] ;

[0104] It follows a standard normal distribution and is based on a time series. order The calculated statistic series, at a given significance level In the case of, if This indicates that there is a clear trend change in the statistical sequence.

[0105] The inverse time sequence of the sequential time sequence is used to calculate a second rank sequence and a second statistic sequence in the same way as the sequential time sequence, and step 302 can further include:

[0106] Step 312, comparing a third sequence value in the inverse time sequence of the sequential time sequence with each fourth sequence value, and constructing a second feature sequence according to a comparison result; if the comparison result indicates that the third sequence value is greater than the fourth sequence value, a sequence value corresponding to the fourth sequence value in the second feature sequence is a first preset value; if the comparison result indicates that the third sequence value is less than or equal to the fourth sequence value, a sequence value corresponding to the fourth sequence value in the first feature sequence is a second preset value; the third sequence value is any one in the inverse time sequence, and the fourth sequence value is any one before the first sequence value;

[0107] Step 322, based on a second target number of the fourth sequence values less than the third sequence value in the second feature sequence, constructing a second rank sequence according to the second target number corresponding to each third sequence value;

[0108] Step 332, calculating a feature value corresponding to the second rank sequence; the feature value includes a mean value and a variance;

[0109] Step 342, according to a preset second statistic algorithm, calculating each sequence value in the second rank sequence and the feature value respectively to obtain a first statistic sequence.

[0110] Comparing a third sequence value in the inverse time sequence of the sequential time sequence with each fourth sequence value, and constructing a second feature sequence according to a comparison result; if the comparison result indicates that the third sequence value is greater than the fourth sequence value, a sequence value corresponding to the fourth sequence value in the second feature sequence is a first preset value; if the comparison result indicates that the third sequence value is less than or equal to the fourth sequence value, a sequence value corresponding to the fourth sequence value in the second feature sequence is a second preset value. The third sequence value is any one in the inverse time sequence, and the fourth sequence value is any one before the third sequence value in the inverse time sequence.

[0111] Further, based on the constructed second feature sequence, a second target number of the fourth sequence values less than the third sequence value is counted, and the second target number is taken as a sequence value corresponding to the third sequence value in the second rank sequence, and a second rank sequence is constructed according to the second target number corresponding to each third sequence value.

[0112] Then, the eigenvalue corresponding to the second order sequence is calculated, the eigenvalue includes the mean value and the variance, according to the preset second statistical quantity algorithm, each sequence value in the second order sequence and the eigenvalue are calculated respectively, and the calculation result is taken as the corresponding sequence value of the sequence value in the second order sequence in the second statistical quantity sequence, so as to construct the second statistical quantity sequence.

[0113] That is, according to the reverse order of the order time sequence , the second statistical quantity sequence is calculated in the same way as the order time sequence, and the initial value of , , the second statistical quantity sequence is obtained.

[0114] By analyzing the first statistical quantity sequence and the second statistical quantity sequence , the time node of the mutation of the order time sequence can be further analyzed, so as to determine the mutation region, and the time period corresponding to the mutation region is the fast fading time period.

[0115] Specifically, if , it indicates that the time sequence shows an upward trend, otherwise, if , it indicates that the time sequence shows a downward trend, when exceeding any critical straight line, it indicates that the upward or downward trend is significant. If and exist intersection points, and the intersection points are between the two critical straight lines, then the time corresponding to the intersection point is the mutation start time. Exemplarily, if there is a mutation region, it is marked that indicates that there is a fast fading point, and the mutation region is , if there is no fast fading, it is marked that , indicating that there is no fast fading point.

[0116] In this embodiment, on the basis of the mutation detection of the MK algorithm, the sliding window processing mode is adopted, and the window size can be adaptively adjusted according to the sampling rate, which can better adapt to the characteristics of different signals. For some fast fading problems only reflected in a local range, the sliding window processing mode can be used for more detailed and accurate detection on the basis of rough identification, and the accuracy of fast fading identification is improved.

[0117] It should be noted that the MK algorithm is pre-trained, and the algorithm parameters that need to be pre-trained include the significance level , the value of may be different in different environments, in actual application, the corresponding reference value and confidence interval of can be obtained by analyzing and counting a large amount of historical data, as the basis for selecting dynamic threshold in fast fading identification. ​

[0118] wherein the dynamic threshold is a dynamically selected threshold for the MK algorithm to determine the fast fading, by adjusting the size of the threshold, the false positive rate and the false negative rate can be reduced while ensuring that the fast decay signal is accurately detected, and the robustness and applicability of the algorithm are improved.

[0119] Further, based on the identified fast fading, after confirming whether the identification result is correct, the MK algorithm can be corrected, thereby improving the identification accuracy of the MK algorithm for fast fading.

[0120] In the embodiment, by performing reverse filling of the IMSI identifier on the MRO data, the user's wireless data and core network signaling data can be accurately matched, thereby extracting the wireless signal data of each terminal user, realizing fast fading identification at the user granularity, and improving the identification accuracy of fast fading.

[0121] Further, on the basis of the double sliding window processing, the Mann-Kendall algorithm is used to detect the mutation of the user's wireless signal segment, which improves the identification accuracy of fast fading, and the dynamic threshold and adaptive window size can be used to make the MK algorithm applicable to the dynamic identification of the 5G fast fading scene, thereby improving the applicability and generalization ability of the signal fast fading identification.

[0122] The signal fast fading identification device provided by the embodiments of the present application will be described below. The signal fast fading identification device described below can be mutually referred to the signal fast fading identification method described above.

[0123] Reference Figure 2 The signal fast fading identification device provided by the embodiments of the present application comprises:

[0124] The data acquisition module 10 is configured to acquire the MRO data collected by the base station, and extract the wireless signal data of each terminal user from the MRO data.

[0125] The first identification module 20 is configured to perform sliding window processing on the wireless signal data based on a first window size, and perform coarse identification on a first signal segment in the current first window to determine whether the first signal segment has fast fading.

[0126] The second identification module 30 is configured to, if the first signal segment has fast fading, perform sliding window processing on the first signal segment based on a second window size, and perform mutation detection on a second signal segment in each second window by using the Mann-Kendall algorithm to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

[0127] In one embodiment, the second identification module 30 is further configured to:

[0128] The first rank sequence is calculated based on the sequential time sequence by using the Mann-Kendall algorithm, and the first statistic sequence of the first rank sequence is calculated according to a preset first statistic algorithm;

[0129] The second rank sequence is calculated based on the reverse sequential time sequence of the sequential time sequence, and the second statistic sequence of the second rank sequence is calculated according to a preset second statistic algorithm;

[0130] The signal change trend of the second signal segment is determined according to the first statistic sequence, and whether there is a signal fast fading in the second signal segment is determined based on the signal change trend;

[0131] If there is a signal fast fading in the second signal segment, the fading start time is determined according to the intersection point of the first statistic sequence and the second statistic sequence;

[0132] The fast fading time period in the first signal segment is determined according to the fading start time.

[0133] In one embodiment, the second identification module 30 is further configured to:

[0134] The first sequence value in the sequential time sequence is compared with each second sequence value, and a first feature sequence corresponding to the first sequence value is constructed according to the comparison result; wherein, if the comparison result represents that the first sequence value is greater than the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a first preset value; if the comparison result represents that the first sequence value is less than or equal to the second sequence value, the sequence value corresponding to the second sequence value in the first feature sequence is a second preset value; the first sequence value is any one in the sequential time sequence, and the second sequence value is any one before the first sequence value;

[0135] The first target number of the second sequence values less than the first sequence value is counted based on the first feature sequence, and a first rank sequence is constructed according to the first target number corresponding to each first sequence value;

[0136] The feature value corresponding to the first rank sequence is calculated; the feature value includes mean and variance;

[0137] Each sequence value in the first rank sequence and the feature value are calculated according to a preset first statistic algorithm, and a first statistic sequence is obtained.

[0138] In one embodiment, the second identification module 30 is further configured to:

[0139] a third sequence value and each fourth sequence value in a reverse sequence time sequence of the sequence time sequence are compared, and a second feature sequence is constructed according to a comparison result; wherein, if the comparison result represents that the third sequence value is greater than the fourth sequence value, a corresponding sequence value of the fourth sequence value in the second feature sequence is a first preset value; if the comparison result represents that the third sequence value is less than or equal to the fourth sequence value, a corresponding sequence value of the fourth sequence value in the first feature sequence is a second preset value; the third sequence value is any one in the reverse sequence time sequence, and the fourth sequence value is any one before the first sequence value;

[0140] a second target quantity of the fourth sequence values less than the third sequence value is counted based on the second feature sequence, and a second rank sequence is constructed according to the second target quantity corresponding to each third sequence value;

[0141] a feature value corresponding to the second rank sequence is calculated; the feature value includes a mean value and a variance;

[0142] each sequence value in the second rank sequence and the feature value are calculated according to a preset second statistical quantity algorithm, and a first statistical quantity sequence is obtained.

[0143] In one embodiment, the first identification module 20 is further configured to:

[0144] a first signal segment in a current first window is coarsely identified, and a target sampling point with a signal strength less than a preset signal threshold in the first signal segment is determined;

[0145] a forward sampling point with a first preset time length before the target sampling point and a backward sampling point with a second preset time length after the target sampling point are obtained based on a target moment corresponding to the target sampling point;

[0146] a forward average level of the target sampling point is obtained by calculating a mean value of signal strengths of the forward sampling points based on the signal strengths corresponding to the forward sampling points, and a backward average level of the target sampling point is obtained by calculating a mean value of signal strengths of the backward sampling point and the target sampling point based on the signal strengths corresponding to the backward sampling point and the target sampling point;

[0147] whether fast fading exists in the first signal segment is determined according to the backward average level and the forward average level; if the backward average level is less than the preset signal threshold, and a difference between the backward average level and the forward average level is greater than a preset level value, the first signal segment has fast fading.

[0148] In one embodiment, the data acquisition module 10 is further configured to:

[0149] extracting an IMSI identifier of each terminal user from the MRO data;

[0150] performing reverse filling on feature data in the MRO data based on the IMSI identifier to obtain filled feature data; the feature data includes at least one of a base station number, a cell number, an AMF area number, an AMF set number, an AMF pointing number, a collection time identifier, a measurement object, a measurement signal strength, a physical cell identifier, and a frequency point number;

[0151] extracting wireless signal data of each terminal user from the filled feature data according to the IMSI identifier.

[0152] Figure 3 An example of an entity structure diagram of an electronic device is shown in FIG. 1, which can include a processor 310, a communications interface 320, a memory 330, and a communications bus 340, wherein the processor 310, the communications interface 320, and the memory 330 can communicate with each other through the communications bus 340. The processor 310 can invoke a logical instruction in the memory 330 to execute a signal fast fading identification method, which includes: Figure 3

[0153] obtaining MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data;

[0154] performing sliding window processing on the wireless signal data based on a first window size, and performing coarse identification on a first signal segment in a current first window to determine whether the first signal segment has fast fading;

[0155] if the first signal segment has fast fading, performing sliding window processing on the first signal segment based on a second window size, and performing mutation detection on a second signal segment in each second window using a Mann-Kendall algorithm to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

[0156] ​Further, the logic instructions in the memory 330 described above can be implemented in the form of software functional units and sold or used as independent products, and can be stored in a computer readable storage medium. Based on such understanding, the technical solutions of the present application essentially or the parts that make contributions to the prior art or parts of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present application. The aforementioned storage medium includes various media that can store program codes, such as a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, etc.

[0157] In another aspect, the embodiments of the present application also provide a computer program product, which comprises a computer program, the computer program can be stored on a non-transitory computer readable storage medium, and the computer program is executed by a processor, so that the computer can execute the signal fast fading identification method provided by the above-mentioned methods, and the method comprises the following steps of:

[0158] obtaining MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data;

[0159] performing sliding window processing on the wireless signal data based on a first window size, performing coarse identification on a first signal segment in a current first window, and determining whether the first signal segment has fast fading;

[0160] if the first signal segment has fast fading, performing sliding window processing on the first signal segment based on a second window size, and performing mutation detection on a second signal segment in each second window by using a Mann-Kendall algorithm to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

[0161] In another aspect, the embodiments of the present application also provide a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the signal fast fading identification method provided by the above-mentioned methods, and the method comprises the following steps of:

[0162] obtaining MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data;

[0163] performing sliding window processing on the wireless signal data based on a first window size, performing coarse identification on a first signal segment in a current first window, and determining whether the first signal segment has fast fading;

[0164] If the first signal segment has fast fading, the first signal segment is processed by a sliding window based on a second window size, and a Mann-Kendall algorithm is used to detect mutation in each second window to determine the fast fading time period in the first signal segment; the second window size is smaller than the first window size.

[0165] The device embodiments described above are merely illustrative, wherein the units described as separate components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed on multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement it without creative labor.

[0166] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be realized by means of software and the necessary general hardware platform, and of course, it can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0167] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to some technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of identifying a signal fast fading, characterized by, The method comprises: acquiring MRO data collected by a base station, and extracting wireless signal data of each terminal user from the MRO data; performing sliding window processing on the wireless signal data based on a first window size, and performing coarse identification on a first signal segment in a current first window to determine whether the first signal segment has fast fading; if the first signal segment has fast fading, performing sliding window processing on the first signal segment based on a second window size, and performing mutation detection on a second signal segment in each second window by using a Mann-Kendall algorithm to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

2. The signal fast-fading identification method according to claim 1, characterized in that, The method of performing mutation detection on the second signal segment in the current second window by using the Mann-Kendall algorithm to determine the fast fading time period in the first signal segment comprises: calculating a first order sequence based on an order time sequence by using the Mann-Kendall algorithm, and calculating a first statistic sequence of the first order sequence according to a preset first statistic algorithm; the order time sequence is composed of signal strengths of sampling points in the second signal segment; calculating a second order sequence based on a reverse order time sequence of the order time sequence, and calculating a second statistic sequence of the second order sequence according to a preset second statistic algorithm; determining a signal change trend of the second signal segment according to the first statistic sequence, and determining whether there is signal fast fading in the second signal segment based on the signal change trend; if there is signal fast fading in the second signal segment, determining a fading start time according to a cross point of the first statistic sequence and the second statistic sequence; determining the fast fading time period in the first signal segment according to the fading start time.

3. The signal fast-fading identification method according to claim 2, characterized in that, The method of calculating the first order sequence based on the order time sequence, and calculating the first statistic sequence of the first order sequence according to the preset first statistic algorithm comprises: comparing a first sequence value in the order time sequence with each second sequence value, and constructing a first feature sequence corresponding to the first sequence value according to a comparison result; if the comparison result indicates that the first sequence value is greater than the second sequence value, a sequence value corresponding to the second sequence value in the first feature sequence is a first preset value; if the comparison result indicates that the first sequence value is less than or equal to the second sequence value, a sequence value corresponding to the second sequence value in the first feature sequence is a second preset value; the first sequence value is any one in the order time sequence, and the second sequence value is any one before the first sequence value; constructing a first order sequence based on a first target number of the second sequence values less than the first sequence value according to a first target number corresponding to each first sequence value; calculating a feature value corresponding to the first order sequence; the feature value comprises a mean value and a variance value; calculating each sequence value in the first order sequence and the feature value according to the preset first statistic algorithm to obtain a first statistic sequence.

4. The signal fast-fading identification method according to claim 2, characterized in that, The inverse sequence time sequence based on the sequence time sequence is used to calculate a second rank sequence, and a second statistical quantity sequence of the second rank sequence is calculated according to a preset second statistical quantity algorithm, including: The third sequence value in the inverse sequence time sequence based on the sequence time sequence is compared with each fourth sequence value, and a second feature sequence is constructed according to a comparison result; if the comparison result indicates that the third sequence value is greater than the fourth sequence value, a corresponding sequence value of the fourth sequence value in the second feature sequence is a first preset value; if the comparison result indicates that the third sequence value is less than or equal to the fourth sequence value, the corresponding sequence value of the fourth sequence value in the second feature sequence is a second preset value; the third sequence value is any one in the inverse sequence time sequence, and the fourth sequence value is any one before the third sequence value; A second target quantity of the fourth sequence value less than the third sequence value is counted based on the second feature sequence, and a second rank sequence is constructed according to the second target quantity corresponding to each third sequence value; A feature value corresponding to the second rank sequence is calculated; the feature value includes a mean value and a variance; Each sequence value in the second rank sequence and the feature value are calculated according to a preset second statistical quantity algorithm, and a first statistical quantity sequence is obtained.

5. The method of claim 1, wherein, The first signal segment in the current first window is coarsely identified to determine whether fast fading exists in the first signal segment, including: The first signal segment in the current first window is coarsely identified to determine whether fast fading exists in the first signal segment, including: Based on the target time corresponding to the target sampling point, a forward sampling point with a first preset time length before the target sampling point and a backward sampling point with a second preset time length after the target sampling point are obtained; Based on the signal strength corresponding to the forward sampling point, an average value of the signal strength of the forward sampling point is calculated to obtain a forward average level of the target sampling point, and based on the signal strength corresponding to the backward sampling point and the target sampling point, an average value of the signal strength of the backward sampling point and the target sampling point is calculated to obtain a backward average level of the target sampling point; Whether fast fading exists in the first signal segment is determined according to the backward average level and the forward average level; if the backward average level is less than the preset signal threshold value, and a difference between the backward average level and the forward average level is greater than a preset level value, the first signal segment has fast fading.

6. The method of claim 1, wherein The wireless signal data of each terminal user is extracted from the MRO data, including: The IMSI identifier of each terminal user is extracted from the MRO data; The feature data in the MRO data is reversely filled based on the IMSI identifier to obtain filled feature data; the feature data includes at least one of a base station number, a cell number, an AMF area number, an AMF set number, an AMF pointing number, a collection time identifier, a measurement object, a measurement signal strength, a physical cell identifier and a frequency point number; According to the IMSI identifier, wireless signal data of each terminal user is extracted from the filled feature data.

7. A signal fast fading identification device, characterized in that, The method comprises the steps of: a data collection module is configured to acquire MRO data collected by a base station and extract wireless signal data of each terminal user from the MRO data; a first identification module is configured to perform sliding window processing on the wireless signal data based on a first window size, perform coarse identification on a first signal segment in a current first window, and determine whether the first signal segment has fast fading; a second identification module is configured to, if the first signal segment has fast fading, perform sliding window processing on the first signal segment based on a second window size, and perform mutation detection on a second signal segment in each second window using a Mann-Kendall algorithm to determine a fast fading time period in the first signal segment; the second window size is smaller than the first window size.

8. An electronic device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor executes the computer program to implement the signal fast fading identification method according to any one of claims 1 to 6. 9.A non-transitory computer-readable storage medium having stored thereon a computer program, characterized in that, The computer program is executed by the processor to implement the signal fast fading identification method according to any one of claims 1 to 6.

10. A computer program product comprising a computer program, characterized in that, The computer program is executed by the processor to implement the signal fast fading identification method according to any one of claims 1 to 6.

Images