• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    A computer implemented method of monitoring performance of a communication network. The method comprises receiving (301) a series of performance indicator values comprising a time series of performance indicator values over a measurement period; identifying (302) a change point in the series of performance indicator values, wherein the change point is a point of time; and using (304) the identified change point in network management actions.
    公开号:FI20195664A1
    申请号:FI20195664
    申请日:2019-08-06
    公开日:2021-02-07
    发明作者:Jukka-Pekka Salmenkaita；Petteri Lundén
    申请人:Elisa Oyj；
    IPC主号:
    专利说明:

    [0001] [0001] The present application generally relates to monitoring performance of communication networks.BACKGROUND
    [0002] [0002] This section illustrates useful background information without admission of any technique described herein representative of the state of the art.
    [0003] [0003] For a communication network operator there is a need to monitor performance of the communication network to ensure quality of service for the users of the communication network. For example, key performance indicator, KPI, values and associated threshold values can be monitored to identify possible performance problems in the network and to identify where and when changes may be needed in the network. Examples of the key performance indicators include signal strength, user distribution (e.g. in terms of timing advance or distance from the transmitter), transmission power, number of dropped calls, throughput, and transmission error rate.
    [0004] [0004] In general, the number of cells in communication networks is large and the amount of available performance data is even larger. For this reason, manually finding and analyzing performance problems is slow and tedious. Therefore, automated monitoring solutions are needed.
    [0007] [0007] In an embodiment, identifying the change point comprises identifying degradation in performance indicator values.
    [0008] [0008] In an embodiment, using the identified change point in network management actions comprises identifying a parameter change and/or a network topology change associated with the change point.
    [0009] [0009] In an embodiment, the method further comprises using the combination of identified parameter change and/or network topology change and the associated performance indicator value change at the change point for tuning future network management actions.
    [0010] [0010] In an embodiment, using the identified change point in network management actions comprises identifying a parameter change associated with the change point and performing a roll back of the parameter change.
    [0011] [0011] In an embodiment, using the identified change point in network management actions comprises selecting one or more cells for performance optimization actions based on the identified change point.
    [0012] [0012] In an embodiment, using the identified change point in network management actions comprises identifying a network topology change associated with the change point and selecting one or more cells affected by the network topology change for performance optimization actions.
    [0013] [0013] In an embodiment, identifying of change point is being performed using statistical methods.
    [0014] [0014] In an embodiment, the method further comprises preprocessing the series of performance indicator values to reduce noise in the series of performance O indicator values.
    [0018] [0018] In an embodiment, the measurement period is selected from group 2 of: a week, two weeks, a month, 3 months.
    [0019] [0019] According to a second example aspect of the present invention, there is provided an apparatus comprising a processor and a memory including computer program code; the memory and the computer program code configured to, with the processor, cause the apparatus to perform the method of the first aspect or any related embodiment.
    [0020] [0020] According to a third example aspect of the present invention, there is provided a computer program comprising computer executable program code which when executed by a processor causes an apparatus to perform the method of the first aspect or any related embodiment.
    [0021] [0021] The computer program of the third aspect may be a computer program product stored on a non-transitory memory medium.
    [0022] [0022] Different non-binding example aspects and embodiments of the present invention have been illustrated in the foregoing. The embodiments in the foregoing are used merely to explain selected aspects or steps that may be utilized in implementations of the present invention. Some embodiments may be presented only with reference to certain example aspects of the invention. It should be appreciated that corresponding embodiments may apply to other example aspects as well.BRIEF DESCRIPTION OF THE DRAWINGS
    [0023] [0023] For a more complete understanding of example embodiments of the present invention, reference is now made to the following descriptions taken in connection with the accompanying drawings in which: O [0024] Fig. 1 shows an example scenario according to an embodiment; N [0025] Fig. 2 shows an apparatus according to an embodiment; 3 [0026] Figs. 3 and 4A-4B show flow diagrams illustrating example methods S according to certain embodiments; E [0027] Fig. 5 shows an example of histogram time series of performance 3 indicator values; LO [0028] Fig. 6 is a graph showing an example change point; and > [0029] Fig. 7 is a graph showing another example change point. 3DETAILED DESCRIPTON OF THE DRAWINGS
    [0030] [0030] Example embodiments of the present invention and its potential advantages are understood by referring to Figs. 1 through 7 of the drawings. In this document, like reference signs denote like parts or steps.
    [0031] [0031] As operational load and network complexity increase due to increasing number of cells and base stations automated monitoring and control of communication networks is clearly beneficial. In an embodiment of the invention performance indicator values available in a communication network are automatically analyzed to identify a point of time, a change point, when there is a significant degradation in performance. The identified change point is then used in network management actions to solve the performance problem. The process may first check if there is a parameter/configuration change and/or a network topology change associated with the change point. That is, whether a parameter/configuration change and/or a network topology change took place at the change point or right before the change point (e.g. some minutes, hours or days before the change point depending on the performance indicator in question).
    [0032] [0032] Fig. 1 shows an example scenario according to an embodiment. The scenario shows a communication network 101 comprising a plurality of cells and base stations and other network devices, and an automation system 111 configured to implement automatic performance monitoring according to example embodiments. Further, the scenario shows a performance optimization system 112 and network management personnel 106. It is to be noted that the automation system 111 and the performance optimization system 112 may be separate logical operations in the same physical device or separate physical devices. O [0033] In an embodiment of the invention the scenario of Fig. 1 operates as N follows: In phase 11, the automation system 111 receives performance indicator 3 values from cells and base stations and other network devices of the communication S network 101. The performance indicator values may comprise for example KPI, key z performance indicator, values for different network devices. KPIs are in general 3 readily available from the communication network 101. The performance indicator LO values may be continuously collected, or the automation system may receive the > performance indicator values in series of performance indicator values. In an embodiment, the automation system receives a time series of the performance indicator values over a measurement period. The measurement period may be for 4 example a week, two weeks, a month, 3 months or some other time period. The performance indicator values may be aggregates over the measurement period (such as total number of radio link failures).
    [0034] [0034] In phase 12, the automation system 111 analyzes the series of performance indicator values and identifies a change point. A change point is a point of time where there is an abrupt variation in a time series. The change point may be identified using statistical methods. The change point may comprise a change in at least one of: average value, mean value, and standard deviation value. “A Survey of Methods for Time Series Change Point Detection”, Knowl Inf Syst. 2017 May; 51(2): 339-367, discloses example methods for identifying a change point.
    [0035] [0035] In an embodiment, the automation system 111 may preprocess the series of performance indicator values prior to the change point analysis. In an embodiment, the preprocessing comprises reducing noise from the received series of performance indicator values. The preprocessing is discussed in more detail later in this document.
    [0036] [0036] In an embodiment, the series of performance indicator values comprises multiple series of different performance indicator values and said identifying of change point is performed based on combination of the multiple series of performance indicator values. The multiple series of performance indicator values may be analyzed in parallel and a voting mechanism may be used for determining which series provides the most meaningful or most important results.
    [0037] [0037] The identified change point is used in network management actions in phases 13-16. It is to be noted that phases 13-16 may be alternatives to each other and there is no requirement to perform all of them, although it is possible O perform more than one of the disclosed options in parallel. In an embodiment, the N automation system 111 determined before proceeding to phases 13-16 that the 3 change point is associated with performance degradation that exceeds predefined S threshold. In this way one may avoid taking actions based on insignificantly small E changes.
    [0039] [0039] In phase 14, one or more cells are selected for performance optimization actions based on the identified change point. The one or more cells may comprise the cell where the change point is identified and certain number nearby cells. The cells are submitted to the performance optimization system 112. Additionally or alternatively, a network topology change (e.g. addition of a new cell or new carrier or a change in antenna tilt) associated with the change point may be identified (e.g. from a database storing information about changes made in the network) and one or more cells affected by the network topology change are selected for performance optimization actions in the performance optimization system 112 in phase 14. The performance optimization system 112 may for example determine configuration changes to improve performance in the one or more cells. The configuration changes are then implemented in the communication network 101 (in the one or more cells) in phase 15. In this way it is possible to automatically identify parts of the network that may benefit from performance optimization actions without trying to optimize the whole network.
    [0040] [0040] In phase 16, the automation system 111 outputs the identified change point to network management personnel 106 for manual analysis and/or manual actions and/or for information. In this way, management personnel may be easily informed of problems that may need their attention. In an example embodiment, phase 16 is applied only in the case no likely cause (such as a parameter or configuration change or deployment of a new cell) for the identified change point was found. This way the number of cells needing manual action may be reduced.
    [0041] [0041] The shown phases may be continuously repeated so that continuous monitoring is provided.
    [0042] [0042] Fig. 2 shows an apparatus 20 according to an embodiment. The O apparatus 20 is for example a general-purpose computer or server or some other N electronic data processing apparatus. The apparatus 20 can be used for 3 implementing embodiments of the invention. That is, with suitable configuration the S apparatus 20 is suited for operating for example as the automation system 111 of E foregoing disclosure. 3 [0043] The general structure of the apparatus 20 comprises a processor 21, LO and a memory 22 coupled to the processor 21. The apparatus 20 further comprises > software 23 stored in the memory 22 and operable to be loaded into and executed in the processor 21. The software 23 may comprise one or more software modules and can be in the form of a computer program product. Further, the apparatus 20 6 comprises a communication interface 25 coupled to the processor 21.
    [0044] [0044] The processor 21 may comprise, e.g., a central processing unit (CPU), a microprocessor, a digital signal processor (DSP), a graphics processing unit, or the like. Fig. 2 shows one processor 21, but the apparatus 20 may comprise a plurality of processors.
    [0045] [0045] The memory 22 may be for example a non-volatile or a volatile memory, such as a read-only memory (ROM), a programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), a random-access memory (RAM), a flash memory, a data disk, an optical storage, a magnetic storage, a smart card, or the like. The apparatus 20 may comprise a plurality of memories.
    [0046] [0046] The communication interface 25 may comprise communication modules that implement data transmission to and from the apparatus 20. The communication modules may comprise, e.g., a wireless or a wired interface module. The wireless interface may comprise such as a WLAN, Bluetooth, infrared (IR), radio frequency identification (RF ID), GSM/GPRS, CDMA, WCDMA, LTE (Long Term Evolution) or 5G radio module. The wired interface may comprise such as Ethernet or universal serial bus (USB), for example. Further the apparatus 20 may comprise a user interface (not shown) for providing interaction with a user of the apparatus. The user interface may comprise a display and a keyboard, for example. The user interaction may be implemented through the communication interface 25, too.
    [0047] [0047] A skilled person appreciates that in addition to the elements shown in Fig. 2, the apparatus 20 may comprise other elements, such as displays, as well as additional circuitry such as memory chips, application-specific integrated circuits (ASIC), other processing circuitry for specific purposes and the like. Further, it is O noted that only one apparatus is shown in Fig. 2, but the embodiments of the N invention may egually be implemented in a cluster of shown apparatuses. 3 [0048] Figs. 3 and 4A-4B show flow diagrams illustrating example methods S according to certain embodiments. The methods may be implemented in the E automation system 111 of Fig. 1 and/or in the apparatus 20 of Fig. 2. The methods 3 are implemented in a computer and do not require human interaction unless LO otherwise expressly stated. It is to be noted that the methods may however provide > output that may be further processed by humans. The methods of Figs. 3-4 may be combined with each other and the order of phases conducted in each method may be changed expect where otherwise explicitly defined. Furthermore, it is to be noted 7 that performing all phases of the flow charts is not mandatory.
    [0049] [0049] Fig. 3 shows a flow diagram illustrating a method according to an embodiment of the invention. The method concerns the process of setting the maintenance priority classes. The method comprises following phases:
    [0050] [0050] Phase 301: A series of performance indicator values are received. The series of performance indicator values comprises a time series of performance indicator values over a measurement period. The measurement period may be for example one of: a week, two weeks, a month, 3 months.
    [0051] [0051] Phase 302: A change point is identified in the series of performance indicator values. The change point is a point of time, where a significant change in the performance indicator values takes place.
    [0052] [0052] In an embodiment, the series of performance indicator values comprises multiple performance indicator values and identification of the change point is performed based on combination of the multiple performance indicator values.
    [0053] [0053] Phase 303: It is determined that there is degradation in performance indicator values at the change point. In an embodiment, it is reguired that the performance degradation exceeds a predefined threshold.
    [0054] [0054] Phase 304: The identified change point is used in network management actions.
    [0055] [0055] Fig. 4A shows a flow diagram illustrating a method according to an embodiment of the invention. The method concerns some example details of the process of identifying the change point. The method comprises following phases:
    [0056] [0056] Phase 302: Change point identification is being performed.
    [0061] [0061] Phase 304: The identified change point is being used in network management actions.
    [0062] [0062] Phase 409: A parameter change and/or a network topology change associated with the change point is identified, if possible. This refers to identifying a parameter change and/or network topology change that took place at the change point or right before the change point.
    [0063] [0063] Phase 410: The identified change point is output to network management personnel. This may be done for example, if no reason for the change point can be automatically identified.
    [0064] [0064] Phase 411: A roll back of a parameter or configuration change is performed. In an embodiment, a parameter change associated with the change point is identified in phase 409 and the identified parameter change is rolled back in phase
    [0065] [0065] Phase 412: One or more cells are selected for performance optimization actions based on the identified change point. In an embodiment, a network topology change associated with the change point is identified in phase 409 and and one or more cells affected by the network topology change are selected for performance optimization actions in phase 412.
    [0066] [0066] In yet further embodiment, the combination of identified parameter change and/or network topology change and the associated performance indicator value change at the change point is used for tuning future network management actions. That is, the identified changes and their correlation with each other are taken into account in either manual or automatic network changes in future. In this way, network management may be improved.
    [0069] [0069] One source of noise is that performance indicator values are typically 9 collected in a limited granularity. The collected performance indicator values may be stored in a form of a histogram, each bin of the histogram corresponding to a certain range of values. For example, there may be a histogram of performance indicator values comprising UE timing advance values where each bin in the histogram corresponds to a certain timing advance range. The histogram then provides the number of samples falling within that range of values. As a result, received series of performance indicator values over a measurement period may comprise a multitude of histograms corresponding to different counter and KPI values.
    [0070] [0070] The histograms themselves may be considered as a multi- dimensional time series, but that is prone to be noisy. Hence, in an embodiment the received series of performance indicator values are preprocessed to reduce dimensionality of the performance indicator values by characterizing the data by computing summary statistics of the histograms per measurement period. Examples of these include mean, average, standard deviation, percentiles (such as 10" or 90" percentile). In addition, correlations between different variables can be used.
    [0071] [0071] Fig. 5 illustrates an example histogram time series and conversion of the histogram into a time series of histogram means and standard deviations. Fig. 5 shows histograms of RSRP (Reference Signal Received Power) for five consecutive days and mean and standard deviation for each day. In the example of Fig. 5, the per day mean is obtained by taking a weighted average of the bin number, where the weight is the number of samples in each bin. The per day standard deviation is obtained by taking weighted average of the deviation from the mean. These summary statistics are not necessarily accurate mean and standard deviation values, because of the limited resolution present in the binned histogram. However, O for practical purposes they are good approximations. In some cases, instead of using N statistics of bin numbers, these can be converted into corresponding data values. 3 For example, an RSRP bin can be converted to corresponding RSRP range in dBm, S or e.g. the mid-point of that range.
    [0073] [0073] In an example embodiment, the preprocessing comprises computing mean performance indicator levels in a group of cells that have similar characteristics or that reside in the same geographical area and subtracting the mean value from the received series of performance indicator values prior to the change point analysis. Also this contributes to noise reduction by suppressing seasonality and trends.
    [0074] [0074] For example, when trees sprout in the spring in certain area, mean RSRP level degrades a little in that area. Such degradation caused by seasonal changes is not a sign of any change that should be corrected. By subtracting known mean RSRP value from individual RSRP values in cells prior to the change point analysis, one achieves that changes affecting the whole area do not have an effect on the change point analysis of an individual cell. The known mean may clearly change over time and may be different in different seasons, for example.
    [0075] [0075] As another example, increase in amount of traffic in certain area typically leads to slow degradation of performance in all cells in the area unless new cells are deployed in the area. Effects of such changes may be likewise suppressed by subtracting known mean RSRP value from individual RSRP values in cells prior to the change point analysis.
    [0076] [0076] In this way, changes in general performance level are suppressed before analyzing cell specific performance values to achieve that only cell specific changes are analyzed.
    [0077] [0077] In an embodiment, the change point identification is started after the preprocessing steps. An example implementation of the change point identification phase splits the time series of the performance indicator values into candidate segments (based on time series of one summary statistics or considering several different statistics in multi-variate sense). A gaussian model is fitted to each segment O separately and the fitting error is computed (i.e. how well does the gaussian N distribution fit the time series segment). A penalty term is applied for each introduced 3 change point. This way introducing new change points improves the accuracy of the S fit on each segment, but causes added costs via the penalty term. Different z combinations of change point locations are tried and the configuration leading to 3 smallest cost (fitting error + penalty from the number of change points) is selected. LO [0078] In other words, the process aims at locating the change points £ that > minimize V(t, y) + pen(t), where V() is a cost function associated with the goodness- of-fit (or fitting error) for data y, and pen() is a regularizer function adding cost for each change point. For example, there may be a fixed cost for each change point. 11
    [0079] [0079] Figs. 6 and 7 are graphs showing examples of change points. Fig. 6 illustrates a case, where mean of the performance indicator changes around point 601, while the standard deviation remains close to the same level. Fig. 7 illustrates a case, where standard deviation of the performance indicator changes around point 701, while the mean remains the same or close to the same level.
    [0080] [0080] After identifying a change point, the process may compare the performance statistics before and after the identified point of time to determine if there was a degradation in the performance. If e.g. the user throughput has decreased by more than a certain threshold value (such as 10%) after the change point, that can be considered as a meaningful degradation that would require corrective actions. The appropriate threshold for detecting a meaningful degradation can be based on, for example, identifying the most degraded cells (e.g. the worst 1%), or by using historical information to determine what level of degradation of the performance indicator is abnormal over longer period of time.
    [0081] [0081] Without in any way limiting the scope, interpretation, or application of the claims appearing below, a technical effect of one or more of the example embodiments disclosed herein is ability to automatically monitor performance of communication networks and to identify performance problems. A further technical effect is ability to automatically recover from performance problems. Thereby improved performance monitoring may be achieved.
    [0082] [0082] Another technical effect of one or more of the example embodiments disclosed herein is ability to save manual monitoring work as monitoring is implemented in automated method.
    [0083] [0083] If desired, the different functions discussed herein may be performed O in a different order and/or concurrently with each other. Furthermore, if desired, one N or more of the before-described functions may be optional or may be combined.
    3 [0084] Although various aspects of the invention are set out in the S independent claims, other aspects of the invention comprise other combinations of E features from the described embodiments and/or the dependent claims with the 3 features of the independent claims, and not solely the combinations explicitly set out LO in the claims.
    S [0085] It is also noted herein that while the foregoing describes example embodiments of the invention, these descriptions should not be viewed in a limiting sense. Rather, there are several variations and modifications, which may be made 12 without departing from the scope of the present invention as defined in the appended claims. oO
    N © OO
    I jami a +OO
    LO o Oo
    N 13
    权利要求:
    Claims (15)
    [1] 1. A computer implemented method of monitoring performance of a communication network, the method comprising receiving (301) a series of performance indicator values comprising a time series of performance indicator values over a measurement period; identifying (302) a change point in the series of performance indicator values, wherein the change point is a point of time; and using (304) the identified change point in network management actions.
    [2] 2. The method of claim 1, wherein identifying (302) the change point comprises identifying degradation in performance indicator values.
    [3] 3. The method of any preceding claim, wherein using (303) the identified change point in network management actions comprises identifying (409) a parameter change and/or a network topology change associated with the change point.
    [4] 4. The method of claim 3, further comprising using the combination of identified parameter change and/or network topology change and the associated performance indicator value change at the change point for tuning future network management actions.
    [5] 5. The method of any preceding claim, wherein using (303) the identified change point in network management actions comprises identifying a parameter change O associated with the change point and performing (411) a roll back of the parameter N change.
    [6] S S 6. The method of any preceding claim, wherein using (303) the identified change E point in network management actions comprises selecting (412) one or more cells 3 for performance optimization actions based on the identified change point.
    [7] O 3 S 7. The method of any preceding claim, wherein using (303) the identified change point in network management actions comprises identifying a network topology 14 change associated with the change point and selecting (412) one or more cells affected by the network topology change for performance optimization actions.
    [8] 8. The method of any preceding claim, wherein said identifying of change point is being performed using statistical methods (404).
    [9] 9. The method of any preceding claim, further comprising preprocessing (403) the series of performance indicator values to reduce noise in the series of performance indicator values.
    [10] 10. The method of any preceding claim, wherein the change point comprises a change in at least one of: average value, mean value, and standard deviation value.
    [11] 11. The method of any preceding claim, further comprising determining (303) that the change point is associated with performance degradation that exceeds predefined threshold.
    [12] 12. The method of any preceding claim, wherein the series of performance indicator values comprises multiple performance indicator values and said identifying of change point is performed based on combination of the multiple performance indicator values.
    [13] 13. The method of any preceding claim, wherein the measurement period is O selected from group of: a week, two weeks, a month, 3 months. & 3
    [14] 14. An apparatus (20, 111) comprising S a processor (21), and E a memory (22) including computer program code; the memory and the 3 computer program code configured to, with the processor, cause the apparatus to LO perform the method of any one of claims 1-13.
    O
    N
    [15] 15. A computer program comprising computer executable program code (23) which when executed by a processor causes an apparatus to perform the method of any one of claims 1-13. o
    O
    N ©
    O
    O
    I jami a +
    O
    O
    LO o
    O
    N 16
    类似技术:
    公开号 | 公开日 | 专利标题
    US10039021B2|2018-07-31|Processing of passive intermodulation detection results
    US11263566B2|2022-03-01|Seasonality validation and determination of patterns
    US9900790B1|2018-02-20|Prediction of performance indicators in cellular networks
    US9867067B2|2018-01-09|Measurement module
    CN104065431B|2016-08-31|The tasks carrying of radio monitoring unmanned system and monitoring method
    US20130196685A1|2013-08-01|Web server and method for hosting a web page for presenting location based user quality data related to a communication network
    US20160088502A1|2016-03-24|Method and network device for cell anomaly detection
    CN105917625B|2020-03-27|Classification of detected network anomalies using additional data
    FI129316B|2021-11-30|Monitoring performance of a communication network
    EP3116254A1|2017-01-11|Network system, analysis device, method for processing measurement information, and recording medium
    EP3549366B1|2022-01-05|Forcasting time series data
    CN111026738A|2020-04-17|Regional population monitoring method and system, electronic equipment and storage medium
    FI20195094A1|2020-08-12|Monitoring operation of communication networks
    FI129315B|2021-11-30|Analyzing operation of cells of a communications network
    EP3836599A1|2021-06-16|Method for detecting permanent failures in mobile telecommunication networks
    FI129288B|2021-11-15|Network monitoring
    US10880763B1|2020-12-29|Remote antenna coverage change event detection in mobile wireless networks
    FI129036B|2021-05-31|Automatic neighbor list optimization in communication networks
    US11234164B2|2022-01-25|Forecasting traffic demand experienced by wireless communication system with application to channel allocation
    CN111246564B|2021-11-23|External interference positioning method and device based on MR data
    WO2017220107A1|2017-12-28|Method and network node for detecting degradation of metric of telecommunications network
    FI20195942A1|2021-05-05|Management of predictive models of a communication network
    FI129289B|2021-11-15|Monitoring impacts of parameter changes
    US20210345138A1|2021-11-04|Enabling Prediction of Future Operational Condition for Sites
    US10841821B2|2020-11-17|Node profiling based on combination of performance management | counters using machine learning techniques
    同族专利:
    公开号 | 公开日
    FI129316B|2021-11-30|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    法律状态:
    2021-11-30| FG| Patent granted|Ref document number: 129316 Country of ref document: FI Kind code of ref document: B |
    优先权:
    申请号 | 申请日 | 专利标题
    FI20195664A|FI129316B|2019-08-06|2019-08-06|Monitoring performance of a communication network|FI20195664A| FI129316B|2019-08-06|2019-08-06|Monitoring performance of a communication network|
    [返回顶部]