System for predicting network traffic

Abstract

A system (100) coupled with a wireless communication network (105) is provided. The system comprises:-a predictor module (101) configured to generate a network traffic prediction (NTP) indicating a number of user terminals in a target area (102) under radio coverage of the wireless communication network during a first time period following a second time period. The predictor module comprises: a first sub-module (102 (1)) configured to generate a first operation prediction (P1) of the number of user terminals in the target area during the first time period based on a historical user terminal density indicator (HI), the historical user terminal density indicator indicates the density of user terminals in the target area in a past time period occurring before the second time period; a second sub-module (102 (2)) configured to generate a second operational prediction (P2) of the number of user terminals in the target area during the first time period based on a combination of the historical user terminal density indicator (HI) and the real-time user terminal location indicator (RI), the real-time user terminal location indicator indicates movement of the user terminal at the target area during the second time period; -a third sub-module (120 (3)) configured to provide the network traffic prediction (NTP) during the second time period based on a selected one between the first operational prediction (P1) and the second operational prediction (P2). The system further comprises a network optimization unit (130) configured to adjust a functional parameter of the wireless communication network based on the network traffic prediction (NTP).

Description

Technical Field

[0001] The present invention generally relates to the field of traffic analysis. More specifically, the present invention relates to a method and system for predicting network traffic of a wireless communication network caused by changes in the geographical density of people moving in a geographical area of interest (e.g., a user of a user terminal connected to a wireless communication network). Background Art

[0002] Predicting network traffic of a wireless communication network provides benefits for improving the operation of the wireless communication network. Accurate network traffic prediction allows the communication network to effectively respond to changes in network traffic by correspondingly adjusting the corresponding wireless communication network functional parameters - such as, for example, by increasing the amount of network resources in a specific part of the wireless communication network where high network traffic is expected or by changing the radio parameter configuration based on the predicted network traffic (e.g., by changing the tilt and / or transmission power of an antenna).

[0003] For traffic analysis for predicting the number of user terminals (e.g., smart phones) in a geographical area (hereinafter briefly referred to as the "target area") of a wireless communication network, it is known to utilize data exchanged between a user terminal (e.g., a smart phone) and a base station of the wireless communication network associated with the target area (e.g., a Long Term Evolution (LTE) network or a 5G network).

[0004] Tracking of the cell identifier (cell ID) of a cell identifying the wireless communication network in which network events (e.g., voice calls, data transmissions, periodic updates) providing the interaction between the user terminal and the wireless communication network occur can be advantageously utilized for traffic analysis purposes. For this purpose, it is useful to utilize a network database containing information about the geographical locations of the individual cells of the wireless communication network (e.g., the location of the corresponding base station and the associated cell coverage).

[0005] With reference to a 5G wireless communication network, a known user location detection process called Minimization of Drive Tests (MDT) can be used to track the location and movement of a user terminal. The MDT process stipulates that the wireless communication network regularly reads the Global Navigation Satellite System (GNSS) (e.g., Global Positioning System (GPS), GLONASS, Galileo) location of user terminals connected to the wireless communication network itself, thereby obtaining information about the location and movement of the user terminals. When a user terminal is located in a place where its GNSS location is unavailable (e.g., when the user terminal is inside a building), or when the GNSS function is disabled, the MDT process is able to determine the location of the user terminal by measuring its radio signal.

[0006] Solutions are known where a network traffic prediction is made by processing, via a machine learning engine, an indicator of user terminal density generated based on observations of a wireless communication network during a past observation time period - hereinafter referred to as "historical user terminal density indicator". Examples of historical user terminal density may include the number of user terminals connected to the wireless network, the number of active user terminals, the number of user terminals that have interacted with the wireless communication network, and so on.

[0007] US10862788B2 discloses a method for evaluating and predicting telecommunication network traffic, the method including receiving, via a processor, site data for a plurality of geographic regions. The processor also receives weather data, event data, and demographic data for the geographic regions. The processor also generates predicted occupancy data for each of the geographic regions and for a plurality of time intervals. The processor also determines a predicted telecommunication network metric for each of the geographic regions and for each of the time intervals based on the predicted occupancy data.

[0008] US10862781B2 discloses methods, systems, and apparatuses for improving connectivity and response speed across a network, including computer programs encoded on a computer storage medium. In one aspect, a method includes: receiving network traffic data at an aggregation point on the network, the network traffic data having been sent to or received from one of a plurality of endpoint devices on the network; calculating a performance metric for each endpoint device based on the received network traffic data; for each endpoint device, comparing the performance metric to a respective threshold to determine a performance issue of the network, where the threshold is determined based on historical network data; associating the determined performance issue of the network with an aspect of the endpoint device; and implementing an action to correct the determined performance issue of the network based on the aspect of at least one endpoint device. Summary of the Invention

[0009] The applicant has observed that the solutions known in the art are not efficient and are affected by drawbacks.

[0010] In particular, machine learning solutions that provide predictions using historical user terminal density indicators can give satisfactory results only if the density of user terminals is within an expected range. These solutions are much less efficient in the case of peaks in user terminal density due to abnormal / unexpected reasons - such as when a large number of user terminals move towards a target area due to a public event (e.g., concert, sports game) occurring in the target area. This is particularly exacerbated because the conditions for the occurrence of these peaks are not suitable for training the machine learning engine. Indeed, due to the rarity of the occurrence of such peaks, they are not adequately represented in the usual training datasets.

[0011] In view of the above, the applicant has designed an improved system that is not affected by the above disadvantages and is used to predict network traffic caused by changes in the geographical density of people moving within a geographical area of interest.

[0012] One or more aspects of the present invention are set forth in the independent claims, with the advantageous features indicated in the dependent claims of the same invention, the wording of which is attached verbatim herein by reference (any advantageous features provided with reference to a specific aspect of the present invention are adapted, mutatis mutandis, to any other aspect of the present invention).

[0013] One aspect of the present invention relates to a system coupled to a wireless communication network.

[0014] The system includes a predictor module configured to generate a network traffic prediction that indicates the number of user terminals in a target area under the radio coverage of the wireless communication network during a first time period after a second time period.

[0015] The predictor module includes a first sub-module configured to generate a first operational prediction of the number of user terminals in the target area during the first time period based on a historical user terminal density indicator that indicates the density of user terminals in the target area during a past time period that occurred before the second time period.

[0016] The predictor module further includes a second sub-module configured to generate a second operational prediction of the number of user terminals in the target area during the first time period based on a combination of the historical user terminal density indicator and a real-time user terminal location indicator that indicates the movement of user terminals at the target area during the second time period.

[0017] The predictor module further includes a third sub-module configured to provide the network traffic prediction based on a selected one between the first operational prediction and the second operational prediction during the second time period.

[0018] The system further includes a network optimization unit configured to adjust functional parameters of the wireless communication network based on the network traffic prediction.

[0019] According to an embodiment of the present invention, the historical user terminal density indicator includes at least one of the following:

[0020] - The number of user terminals connected to the wireless communication network;

[0021] - The number of user terminals whose last interaction with the wireless communication network occurred in a cell corresponding to the target area in the wireless communication network;

[0022] - The number of active user terminals;

[0023] - Traffic caused by user terminals connected to the wireless communication network and collected in the past time period that occurred before the second time period.

[0024] According to an embodiment of the present invention, the real-time user terminal location indicator includes at least one of the following:

[0025] - GNSS records, each GNSS record containing at least the GNSS location of a user terminal located at the target area during the second time period;

[0026] - A user terminal location indicator generated by the interaction between a user terminal at the target area and the wireless communication network during the second time period.

[0027] According to an embodiment of the present invention, the first sub-module is configured to generate the first operation prediction by processing the historical user terminal density indicator via a machine learning algorithm.

[0028] According to an embodiment of the present invention, the second sub-module is configured to generate the second operation prediction by comparing the historical user terminal density indicator with the real-time user terminal location indicator.

[0029] According to an embodiment of the present invention, the third sub-module is configured to provide the traffic prediction based on the second operation prediction during the second time period if the following two conditions are both verified:

[0030] - The number of user terminals corresponding to the first operation prediction is lower than the number of user terminals at the target area evaluated based on the real-time user terminal location indicator, and

[0031] - The difference between the number of user terminals at the target area evaluated based on the real-time user terminal location indicator and the number of user terminals corresponding to the first operation prediction is higher than a threshold.

[0032] According to an embodiment of the present invention, the third sub-module is configured to: during the second time period, if the first operation prediction indicates a peak in the number of user terminals in the target area, provide the traffic prediction based on the second operation prediction.

[0033] According to an embodiment of the present invention, the third sub-module is configured to receive calendar data, the calendar data provides an indication of a planned increase in the number of user terminals in the target area, and the third sub-module is further configured to: in the second time period, if the calendar data provides an indication of a planned increase in the number of user terminals in the target area for the first time period, then provide the traffic prediction based on the second operation prediction.

[0034] According to an embodiment of the present invention, the first sub-module is configured to:

[0035] - subdivide a range of possible values of the first operation prediction into a set of sub-intervals, and

[0036] - calculate a prediction probability distribution that provides an indication of the reliability of the first operation prediction corresponding to the value of each sub-interval for each sub-interval, wherein the third calculation sub-module is configured to, in the second time period, provide the traffic prediction based on the second operation prediction if the prediction probability distribution PD is below a reliability threshold for all sub-intervals.

[0037] According to an embodiment of the present invention, the third sub-module is configured to: in the second time period, if the second operation prediction indicates a peak in the number of user terminals in the target area, then provide the traffic prediction based on the second operation prediction.

[0038] According to an embodiment of the present invention, the system further includes a self-organizing network module configured to adjust functional parameters of a wireless communication network based on the network traffic prediction.

[0039] According to an embodiment of the present invention, the system further includes a drone cell module configured to drive a drone of a cell site equipped with a wireless communication network to the target area based on the network traffic prediction. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 is a schematic representation of a system according to an embodiment of the present invention, the system including a predictor module for analyzing traffic in a target area;

[0041] Figures 2A - 2D is a graph showing experimental results of the behavior of the density of user terminals over time;

[0042] Figure 3A –3E depicts a flowchart showing the main operations performed by the Figure 1 system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0043] Referring to the accompanying drawings, Figure 1System 100 including a predictor module 101 according to an embodiment of the present invention is schematically shown. The predictor module 101 is configured to analyze directed traffic to generate a network traffic prediction NTP indicating the number of user terminals UT (e.g., smart phones) in a target area 102. It should be noted that the terms 'unit','system','module' are intended to include, but not be limited to, hardware, firmware, combinations of hardware and software, and software herein.

[0044] The target area 102 is under the radio coverage of a wireless communication network 105 such as a mobile phone network (2G, 3G, 4G, 5G or higher generations), and includes a plurality (three or more) of base stations 105a geographically distributed through a corresponding area including the target area 102. Each base station 105a is adapted to manage the communication of user terminals UT in one or more served areas or cells 105b. In the example under discussion, each base station 105a serves three cells 105b, but similar considerations apply to the case where each base station 105a serves a different number of cells 105b.

[0045] Each base station 105a of the wireless communication network 105 is adapted to interact with any user terminal UT located within one of the cells 105b served by such base station 105a. Such interaction between the user terminal UT and the wireless communication network 105 will generally be represented as a "network event" and may include (non-exhaustively) interactions at power on / off, incoming / outgoing voice calls, sending / receiving SMS, Internet access, general data transfer, etc.

[0046] According to an embodiment of the present invention, the predictor module 101 includes:

[0047] - A first calculation sub-module 120(1), configured to generate a first operational prediction P1 of the number of user terminals UT that will be present in the target area 102 during a corresponding future time period FTP that has not yet occurred;

[0048] - A second calculation sub-module 120(2), configured to generate a second operational prediction P2 of the number of user terminals UT that will be present in the target area 102 during the future time period FTP.

[0049] As will be described in more detail in the following description, the two operational predictions P1 and P2 generated by the calculation sub-modules 120(1), 120(2) according to an embodiment of the present invention are generally different, are obtained based on different principles and using different data sets and different types of data, and are suitable for consideration under different conditions.

[0050] According to an embodiment of the present invention, the predictor module 101 further includes a third computing sub-module 120(3), configured to provide a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 based on one selected between the operation predictions P1 and P2.

[0051] According to an embodiment of the present invention, the system 100 further includes a network optimization unit 130, configured to adjust the functional parameters of the wireless communication network 105 based on the network traffic prediction NTP provided by the third computing sub-module 120(3) of the predictor module 101, so as to allow the wireless communication network 105 to efficiently respond to changes in network traffic in a manner that provides sufficient quality of service without wasting too much network resources. For example, according to an embodiment of the present invention, the network optimization unit 130 may include a self-organizing network module, which is configured to automatically adjust the functional parameters of the wireless communication network based on the network traffic prediction NTP (e.g., the amount of network resources allocated for transmission at the cell 105b included in the target area 102 of the wireless communication network 105, and / or the antenna configuration of the base station 105a serving these cells 105b). According to another exemplary embodiment of the present invention, the network optimization unit 130 may include a drone cell module, which is configured to drive a drone equipped with a cell site of the wireless communication network 105 to the target area 102 based on the network traffic prediction.

[0052] According to an embodiment of the present invention, the first computing sub-module 120(1) is configured to generate a first operation prediction P1 based on a historical user terminal density indicator HI indicating the density of user terminals UT in the target area 102 during a past time period PATP, and thus provide an indication of the historical network traffic behavior in the target area 102.

[0053] By the past time period PATP, it is meant herein a period of a certain length that occurred substantially before the current time when the first computing sub-module 120(1) is generating the first operation prediction P1 (e.g., corresponding to several hours, several days, several weeks, several months, several years). The past time period PATP is temporally prior to the future time period FTP.

[0054] According to an embodiment of the present invention, the historical user terminal density indicator HI includes at least one of the following indicators collected during the past time period:

[0055] - The number of user terminals UT connected to the wireless communication network 105;

[0056] - The number of user terminals UT whose last interaction (network event) with the wireless communication network 105 occurred in the cell 105B included in the target area 102;

[0057] - The number of active user terminals UT;

[0058] - The traffic caused by user terminals UT connected to the wireless communication network 105.

[0059] According to an embodiment of the present invention, a historical user terminal density indicator HI is stored in a corresponding repository module 140, which is included in or coupled to the predictor module 101.

[0060] According to an embodiment of the present invention, the first calculation sub-module 120(1) is configured to generate a first operation prediction P1 by processing the historical user terminal density indicator HI via a machine learning algorithm. For this purpose, according to an embodiment of the present invention, the first calculation sub-module 120(1) is equipped with a machine learning engine, which is trained to generate the first operation prediction P1 by considering the history of network traffic within the target area 102 obtained through the historical user terminal density indicator HI.

[0061] According to an embodiment of the present invention, the machine learning engine of the first calculation sub-module 120(1) is configured to implement a deep learning algorithm, such as, for example, a mixture of experts based on a deep learning model. The concepts of the present invention can also be applied in cases where the machine learning engine of the first calculation sub-module 120(1) is configured to operate according to different neural network methods, such as, for example, a known convolutional neural network method, a long short-term memory network method, a multi-horizon quantile recurrent prediction method, a deep autoregressive recurrent neural network method, a graph convolutional network method. Moreover, the concepts of the present invention can also be extended to those cases where predictions are made according to the autoregressive integrated moving average method and the autoregressive conditional heteroskedasticity method. According to an embodiment of the present invention, in order to improve the accuracy of the first operation prediction P1 generated by the first calculation sub-module 120(1), the content of the repository module 140 is updated, for example, regularly with newly collected historical user terminal density indicators HI.

[0062] According to an embodiment of the present invention, the length of the past time period PATP depends on the algorithm / method implemented by the first calculation sub-module 120(1). For example, for a mixture of experts based on a deep learning model, the past time period PATP can be equal to approximately 2 weeks, while for a graph convolutional network method, the past time period PATP can be equal to approximately 16 weeks.

[0063] According to an embodiment of the present invention, the first calculation sub-module 120(1) can be trained weekly to update the prediction model.

[0064] The longer the past time period PATP considered, the greater the amount of available data, and thus the higher the prediction accuracy (for the same statistical behavior of the communication network cells). However, if the statistical behavior of one or more cells of the communication network changes, considering too long a past time period PATP can be counterproductive, because in such cases the predictor takes more time to adapt to the new behavior.

[0065] The first operation prediction P1 generated by the first calculation sub-module 120(1) using machine learning algorithms with historical observations of network traffic (i.e., through the historical user terminal density indicator HI) is particularly suitable for cases where the network traffic in the considered cell 105b has a regular historical behavior - i.e., in those cases where the behavior of the density of user terminals UT over time has been observed to follow a (quasi)-regular pattern occurring within the expected range - to predict the number of user terminals UT in the target area 102. In these cases, the first calculation sub-module 120(1) is able to correctly predict the occurrence of the peak in the number of user terminals UT in the target area 102, i.e., it is able to generate a first operation prediction P1 that provides a sufficiently correct determination of the time of occurrence of the peak and a sufficiently correct quantification of the number of user terminals UT corresponding to said peak.

[0066] Figure 2A and 2B are graphs of experimental results related to two cells 105b within the target area 102, for which the behavior of the density of user terminals UT over time has been observed to follow a (quasi)-regular pattern occurring within the expected range. The solid line corresponds to the actual evolution of the number of user terminals UT over time, while the dashed line corresponds to the number of user terminals UT corresponding to the first operation prediction P1 generated by the first calculation sub-module 120(1) using a mixture of experts based on a deep learning model to process the historical user terminal density indicator HI. By observing the pictures, it can be seen that the first operation prediction P1 generated by the first calculation sub-module 120(1) is clearly accurate.

[0067] The first operation prediction P1 generated by the first calculation sub-module 120(1) using machine learning algorithms with historical observations of network traffic (i.e., through the historical user terminal density indicator HI), on the contrary, is not suitable for predicting the number of user terminals UT in the target area 102 in cases where the network traffic in the considered cell 105b is subject to peaks occurring due to abnormal / unexpected reasons - such as when a large number of user terminals UT move towards the target area 102 due to a public event occurring in the target area 102 (e.g., a concert, a sports game). In these cases - two examples of which are in Figure 2C and 2DIn the figure, it is shown that the first calculation sub-module 120(1) cannot correctly predict the occurrence of the peak in the number of user terminals UT in the target area 102, that is, the generated first operation prediction P1 provides an incorrect determination of the occurrence time of the peak and / or an incorrect quantification of the number of user terminals UT corresponding to the peak.

[0068] Returning to Figure 1 , according to an embodiment of the present invention, the second calculation sub-module 120(2) is configured to generate a second operation prediction P2 based on a combination between:

[0069] - A historical user terminal density indicator HI, which indicates the density of user terminals UT in the target area 102 in the past (i.e., in the past time period PATP), thereby providing an indication of the historical network traffic behavior in the target area 102, and

[0070] - A real-time user terminal position indicator RI collected from the user terminal UT and indicating the movement of the user terminal UT at the target area 102 during the current time period PRTP, that is, from which a real-time or near-real-time indication of the movement of the user terminal UT towards / away from / within the target area 102 can be obtained.

[0071] By the current time period PRTP, it is meant herein to include the current time when the second calculation sub-module 120(2) is generating the second operation prediction P2 or a time period of a certain length before this current time (e.g., corresponding to several minutes or hours). The past time period PATP corresponding to the historical user terminal density indicator HI is temporally prior to the current time period PRTP. The current time period PRTP is temporally prior to the future time period FTP.

[0072] According to an embodiment of the present invention, the real-time user terminal position indicator RI is collected from the user terminal UT and includes at least one of the following two:

[0073] - GNSS (e.g., GPS, GLONASS, Galileo) records, each GNSS record containing at least the GNSS position of the user terminal UT located at the target area 102 (i.e., within and / or close to it and moving towards / away from it) during the current time period PRTP;

[0074] - A user terminal position indicator generated by the interaction (network events) of the user terminal UT located at the target area 102 (i.e., within and / or close to it and moving towards / away from it) with the wireless communication network 105 during the current time period PRTP.

[0075] According to an embodiment of the present invention, the real-time user terminal location indicator RI is collected by the wireless communication network 105 and sent to the second computing sub-module 120(2) by means of, for example, a known MDT process.

[0076] The second operation prediction P2 generated by the second computing sub-module 120(2) is particularly suitable for predicting the number of user terminals UT in the target area 102 in the case where the network traffic in the considered cell 105b undergoes a peak due to the abnormal and unexpected movement of the user terminal UT within or towards the target area 102, because the prediction is made by the second computing sub-module 120(2) by also considering the real-time user terminal location indicator RI, and a real-time or near-real-time indication of the movement of the user terminal UT at the target area 102 (i.e., within it and / or close to it and moving towards / away from it) can be obtained from the real-time user terminal location indicator RI.

[0077] The second computing sub-module 120(2) is in particular configured to predict in advance a consistent flow towards / away from the target area 102 generated by abnormal and unexpected situations, which can be detected from the movement of the user terminal UT.

[0078] According to an embodiment of the present invention, the second computing sub-module 120(2) is configured to generate the second operation prediction P2 by comparing the historical user terminal density indicator HI with the real-time user terminal location indicator RI. In this way, the prediction is generated by understanding the current (or just upcoming) network traffic situation and thus by understanding how much the network traffic situation deviates from the normal traffic corresponding to the historical user terminal density indicator HI.

[0079] It is noted that when there is no consistent flow of the user terminal UT towards / away from the target area 102, useful information cannot be extracted from the real-time user terminal location indicator RI that can increase the prediction accuracy. On the contrary, in these cases, the real-time user terminal location indicator RI may reduce the prediction accuracy and act as noise data.

[0080] According to an exemplary but non-limiting embodiment of the present invention, the second computing sub-module 120(2) may be based on the method and system disclosed in the published international application WO 2020 / 002094 by the same applicant of the present application, i.e., it is configured to predict the number of user terminals UT in the target area 102 based on the difference between:

[0081] - A first term, which indicates the number of user terminals UT that are moving from the external area to the intermediate area surrounding the target area 102 during the current time period PRTP (in the case where the external area surrounds the intermediate area), and

[0082] - The second item, which indicates the number of user terminals UT that have moved from the external area to the intermediate area during a part of the past time period PATP that matches the current time period PRTP.

[0083] According to an embodiment of the present invention, the first item is calculated by the second calculation sub-module 120(2) through the processing of the real-time user terminal location indicator RI.

[0084] According to an embodiment of the present invention, the second item is calculated by the second calculation sub-module 120(2) through the processing of the historical user terminal density indicator HI.

[0085] As already mentioned above, according to an embodiment of the present invention, the third calculation sub-module 120(3) is configured to provide a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 based on one selected between the operation predictions P1 and P2. As will be described in more detail below, the selection of the operation prediction P1 or the operation prediction P2 according to an embodiment of the present invention can be performed by the third calculation sub-module 120(3) according to one or more of the following:

[0086] - A comparison between the number of user terminals UT corresponding to the first operation prediction P1 and the number of user terminals UT at the target area 102 evaluated based on the real-time user terminal location indicator RI;

[0087] - The first operation prediction P1 itself;

[0088] - Calendar data CDATA received by the third calculation sub-module 120(3) and providing an indication of a planned increase in the number of user terminals UT in the target area 102;

[0089] - A prediction probability distribution PD, which provides an indication of the reliability of the first operation prediction P1 over a sub-interval of predicted values;

[0090] - The second operation prediction P2 itself.

[0091] Figure 3A Depicts a flowchart showing the main operations performed while the system 100 operates during the above-mentioned current time period PRTP according to an embodiment of the present invention.

[0092] According to an embodiment of the present invention, the first calculation sub-module 120(1) uses the historical user terminal density indicator HI as described above to generate a first operation prediction P1 (block 302) of the number of user terminals UT that will be present in the target area 102 during the corresponding future time period FTP.

[0093] According to an embodiment of the present invention, the third calculation sub-module 120(3) then collects real-time user terminal location indicators RI corresponding to the target area 102, and accordingly calculates the number RNT of user terminals UT currently located in the target area 102 based on the collected real-time user terminal location indicators RI (block 304).

[0094] According to an embodiment of the present invention, the third calculation sub-module 120(3) compares the number of user terminals UT of the first operation prediction P1 with the calculated number RNT (block 306).

[0095] According to an embodiment of the present invention, if the calculated number RNT is not much higher than the number of user terminals UT of the first operation prediction P1, that is, if the difference between the calculated number RNT and the number of user terminals UT of the first operation prediction P1 is not higher than the corresponding threshold TH1 - such as, for example, equal to 100 - 300 (exit branch N of block 306), then the third calculation sub-module 120(3) provides a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 that is set to the first operation prediction P1 (block 308).

[0096] According to an embodiment of the present invention, if the calculated number RNT is much higher than the number of user terminals UT of the first operation prediction P1, that is, if the difference between the calculated number RNT and the number of user terminals UT of the first operation prediction P1 is higher than the threshold TH1 (exit branch Y of block 306), then the third calculation sub-module 120(3) activates (block 309) the second calculation sub-module 120(2). The second calculation sub-module 120(2) then collects real-time user terminal location indicators RI corresponding to the target area 102, and accordingly generates a second operation prediction P2 for the number of user terminals UT that will be present in the target area 102 during the future time period FTP using a combination of the historical user terminal density indicator HI and the collected real-time user terminal location indicators RI as described above (block 310). At this time, the third calculation sub-module 120(3) provides a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 that is set to the second operation prediction P2 (block 311).

[0097] According to an embodiment of the present invention, the network traffic prediction NTP provided by the third computing sub-module 120(3) (which may be the first operation prediction P1 or the second operation prediction P2, as described above) is fed into the network optimization unit 130, and the network optimization unit 130 utilizes this information to adjust, if possible, the functional parameters of the wireless communication network 105 (block 312). For example, if the network traffic prediction NTP indicates an upcoming peak in the number of user terminals UT in the target area 120, the network optimization unit 130 may control (or instruct) the wireless communication network 105 to increase the radio resources assigned to the base station 105a serving the cell 105b corresponding to the target area 102, and / or drive the drone equipped with the support cell site of the wireless communication network 105 to the target area 102.

[0098] Another application domain of the solution according to an embodiment of the present invention relates to the radio transmission of the antennas of the wireless communication network 105. Indeed, in order to avoid violating the regulations regarding radio transmission, each antenna of the base station 105a of the wireless communication network 105 should keep its hourly radio transmission below the corresponding threshold. The larger the number of user terminals UT, the higher the radio transmission. Therefore, in the case of an abnormal large concentration of user terminals UT in the target area 102, such a threshold may be exceeded. Since the abnormal large concentration of user terminals UT in the target area 102 can be predicted in advance by the solution according to an embodiment of the present invention, the threshold for the maximum hourly radio transmission of the antennas of the base station 105a serving the cell 105b corresponding to the target area 102 can be temporarily increased.

[0099] Figure 3B A flowchart depicting the main operations performed while the system 100 operates during the above-mentioned current period PRTP according to another embodiment of the present invention is shown. It will be depicted with blocks having the same reference numerals as those Figure 3A used in Figure 3A the flowchart corresponding to the operations equivalent to those already described in Figure 3B the flowchart of

[0100] According to an embodiment of the present invention, the first computing sub-module 120(1) generates the first operation prediction P1 using the historical user terminal density indicator HI (block 302).

[0101] According to an embodiment of the present invention, the third computing sub-module 120(3) evaluates whether the first operation prediction P1 generated by the first computing sub-module 120(1) indicates an upcoming peak in the number of user terminals UT (block 316).

[0102] According to an embodiment of the present invention, if the first operation prediction P1 does not indicate an upcoming peak in the number of user terminals UT (exit branch N of block 316), the third calculation sub-module 120(3) provides a network traffic prediction NTP that is set to the first operation prediction P1 and indicates the number of user terminals UT in the target area 102 (block 308).

[0103] According to an embodiment of the present invention, if the first operation prediction P1 instead indicates an upcoming peak in the number of user terminals UT (exit branch Y of block 316), the third calculation sub-module 120(3) activates (block 309) the second calculation sub-module 120(2), and the second calculation sub-module 120(2) collects real-time user terminal location indicators RI corresponding to the target area 102 and accordingly uses a combination of the historical user terminal density indicator HI (from which a prediction regarding the historical trend can be calculated) and the collected real-time user terminal location indicators RI to generate a second operation prediction P2 (block 310). Then, the third calculation sub-module 120(3) provides a network traffic prediction NTP that is set to the second operation prediction P2 (block 311).

[0104] According to an embodiment of the present invention, the network traffic prediction NTP provided by the third calculation sub-module 120(3) is fed to the network optimization unit 130, and the network optimization unit 130 utilizes this information to adjust, if possible, the functional parameters of the wireless communication network 105 (block 312).

[0105] Figure 3C A flowchart depicting the main operations performed while the system 100 operates during the above-mentioned current time period PRTP according to another embodiment of the present invention is shown. Operations corresponding to those already described in the Figures 3A - 3B flowchart will be depicted using blocks with the same reference numerals as those used in the figures, and their description will be omitted or reasonably compressed for the sake of brevity. Figure 3C flowchart

[0106] According to an embodiment of the present invention, the first calculation sub-module 120(1) generates a first operation prediction P1 using the historical user terminal density indicator HI (block 302).

[0107] According to an embodiment of the present invention, a third computing sub-module 120(3) receives calendar data CDATA, which provides an indication of a planned increase in the number of user terminals UT in a target area 102 (block 320). For example, the calendar data CDATA may include an aggregation of data collected from the Internet - such as from a website / social network - so as to provide an indication of a planned event (e.g., a sports game or a concert) that causes an aggregation of people, which will in turn cause an increase in the density of user terminals UT.

[0108] According to an embodiment of the present invention, the third computing sub-module 120(3) examines the calendar data CDATA to evaluate the presence of an upcoming increase in the number of user terminals UT in the target area 102 (block 322).

[0109] According to an embodiment of the present invention, if the calendar data CDATA does not provide any upcoming increase in the number of user terminals UT in the target area 102 (exit branch N of block 322), then the third computing sub-module 120(3) provides a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 that is set to a first operation prediction P1 (block 308).

[0110] According to an embodiment of the present invention, if the calendar data CDATA does provide an upcoming increase in the number of user terminals UT in the target area 102 (exit branch Y of block 322), then the third computing sub-module 120(3) activates (block 309) a second computing sub-module 120(2), which collects real-time user terminal location indicators RI corresponding to the target area 102 and accordingly generates a second operation prediction P2 using a combination of a historical user terminal density indicator HI and the collected real-time user terminal location indicators RI (block 310). Then, the third computing sub-module 120(3) provides a network traffic prediction NTP that is set to the second traffic prediction P2 (block 311).

[0111] According to an embodiment of the present invention, the network traffic prediction NTP provided by the third computing sub-module 120(3) is fed to a network optimization unit 130, which utilizes this information to adjust, if possible, the functional parameters of the wireless communication network 105 (block 312).

[0112] Figure 3D A flowchart depicts the main operations performed while the system 100 operates during the above-mentioned current time period PRTP according to another embodiment of the present invention. Operations corresponding to those equivalent to the operations already described in Figures 3A - 3C the flowchart will be depicted by blocks having the same reference numerals as those used in the Figure 3DThe operations of the flowcharts, and their descriptions will be omitted or reasonably compressed for the sake of brevity.

[0113] According to an embodiment of the present invention, the first calculation sub-module 120(1) subdivides the range of possible values of the first operation prediction P1 into a set of sub-intervals SP1(i) (i = 1, 2,...), and calculates a prediction probability distribution PD (block 330), where the prediction probability distribution PD provides an indication of the reliability of the prediction corresponding to the value of each sub-interval SP1(i).

[0114] According to an embodiment of the present invention, the third calculation sub-module 120(3) is configured to compare the prediction probability distribution PD with a corresponding reliability threshold RTH (block 332).

[0115] According to an embodiment of the present invention, if the prediction probability distribution PD is higher than the reliability threshold RTH only for a single sub-interval SP1(i) (exit branch Y of block 332), then the third calculation sub-module 120(3) provides a network traffic prediction NTP that is set to the number of user terminals UT in the target area 102 including the selected value included in the sub-interval SP1(i) (block 334).

[0116] According to an embodiment of the present invention, if the prediction probability distribution PD is always lower than the reliability threshold RTH (exit branch N of block 332), then the third calculation sub-module 120(3) evaluates that the prediction made by the first calculation sub-module 120(1) is unreliable, and the third calculation sub-module 120(3) activates (block 309) the second calculation sub-module 120(2), which collects real-time user terminal location indicators RI corresponding to the target area 102 and accordingly uses a combination of the historical user terminal density indicator HI and the collected real-time user terminal location indicators RI to generate a second operation prediction P2 (block 310). Then, the third calculation sub-module 120(3) provides a network traffic prediction NTP that is set to the second traffic prediction P2 (block 311).

[0117] According to an embodiment of the present invention, the network traffic prediction NTP provided by the third calculation sub-module 120(3) is fed to the network optimization unit 130, and the network optimization unit 130 uses this information to adjust the functional parameters of the wireless communication network 105 if possible (block 312).

[0118] Figure 3E A flowchart depicts the main operations performed while the system 100 operates during the above-mentioned current time period PRTP according to another embodiment of the present invention. Blocks having the same reference numerals as those used in the said figures will be used to depict the operations equivalent to those already in Figures 3A - 3Dcorresponding to the operations described in the flowchart of Figure 3E the operations of the flowchart, and for the sake of brevity, their descriptions will be omitted or reasonably compressed.

[0119] According to an embodiment of the present invention, the first calculation sub-module 120(1) and the second calculation sub-module 120(2) operate concurrently to respectively generate a first operation prediction P1 and a second operation prediction P2 (block 350).

[0120] According to an embodiment of the present invention, the third calculation sub-module 120(3) compares the first operation prediction P1 with the second operation prediction P2 (block 355).

[0121] According to an embodiment of the present invention, if the number of user terminals UT corresponding to the first operation prediction P1 is not lower than a corresponding threshold TH2, such as, for example, equal to 100 - 300 (exit branch Y of block 355), than the number of user terminals UT corresponding to the second operation prediction P2, then the third calculation sub-module 120(3) provides a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 that is set to the first operation prediction P1 (block 308).

[0122] According to an embodiment of the present invention, if the number of user terminals UT corresponding to the first operation prediction P1 is lower than the threshold TH2 than the number of user terminals UT corresponding to the second operation prediction P2, because the second operation prediction P2 indicates an upcoming peak in the number of user terminals UT, then the third calculation sub-module 120(3) provides a network traffic prediction NTP indicating the number of user terminals UT in the target area 102 that is set to the second operation prediction P2 (block 360).

[0123] According to an embodiment of the present invention, the network traffic prediction NTP provided by the third calculation sub-module 120(3) is fed to the network optimization unit 130, and the network optimization unit 130 uses this information to adjust the functional parameters of the wireless communication network 105 if possible (block 312).

[0124] According to an embodiment of the present invention, the operation of the system 100 can provide Figures 3A - 3E a combination of two or more of the flowcharts shown in Figures 3A - 3E such that the third calculation sub-module 120(3) can select which one of the first operation prediction P1 and the second operation prediction P2 to use to provide the network traffic prediction NTP by verifying that more than one of the conditions corresponding to blocks 306, 316, 322, 332, 355 in the flowchart shown in

[0125] Naturally, in order to meet local and specific requirements, those skilled in the art can make many logical and / or physical modifications and changes to the present invention described above. More specifically, although the present invention has been described with a certain degree of particularity with reference to the preferred embodiments of the present invention, it should be understood that various omissions, substitutions, and changes in form and detail, as well as other embodiments, are possible. In particular, different embodiments of the present invention can even be practiced without the specific details set forth in the foregoing description for the purpose of providing a more thorough understanding thereof; on the contrary, well-known features may have been omitted or simplified so as not to impede the description with unnecessary details. Moreover, it is expressly intended that the specific elements and / or method steps described in connection with any disclosed embodiment of the present invention can be incorporated into any other embodiment.

Claims

1. A system (100) coupled to a wireless communication network (105), the system comprising: - A predictor module (101) configured to generate a network traffic prediction (NTP), the network traffic prediction indicating the number of user terminals in a target area (102) under the radio coverage of the wireless communication network during a first time period after a second time period, the predictor module comprising: - A first sub-module (102(1)) configured to generate a first operational prediction (P1) of the number of user terminals in the target area during the first time period based on a historical user terminal density indicator (HI), the historical user terminal density indicator indicating the density of user terminals in the target area during a past time period that occurred before the second time period; - A second sub-module (102(2)) configured to generate a second operational prediction (P2) of the number of user terminals in the target area during the first time period based on a combination of the historical user terminal density indicator (HI) and a real-time user terminal location indicator (RI), the real-time user terminal location indicator indicating the movement of user terminals at the target area during the second time period; - A third sub-module (120(3)) configured to provide the network traffic prediction (NTP) based on a selected one between the first operational prediction (P1) and the second operational prediction (P2) during the second time period; - A network optimization unit (130) configured to adjust functional parameters of the wireless communication network based on the network traffic prediction (NTP).

2. The system (100) according to claim 1, wherein, the historical user terminal density indicator (HI) comprises at least one of the following: - The number of user terminals connected to the wireless communication network (105); - The number of user terminals whose last interaction with the wireless communication network occurred in a cell corresponding to the target area in the wireless communication network; - The number of active user terminals; - The traffic caused by user terminals connected to the wireless communication network and collected during the past time period that occurred before the second time period.

3. The system (100) according to claim 1 or 2, wherein, the real-time user terminal location indicator (RI) comprises at least one of the following: - GNSS records, each GNSS record containing at least the GNSS location of a user terminal located at the target area (102) during the second time period; - A user terminal location indicator generated through the interaction of user terminals at the target area (102) with the wireless communication network (105) during the second time period.

4. The system (100) according to any one of the preceding claims, wherein, the first sub-module (120(1)) is configured to generate the first operational prediction (P1) by processing the historical user terminal density indicator (HI) via a machine learning algorithm.

5. The system (100) according to any one of the preceding claims, wherein, The second sub-module (120(2)) is configured to generate the second operation prediction (P2) by comparing the historical user terminal density indicator (HI) with the real-time user terminal location indicator (RI).

6. The system (100) according to any one of the preceding claims, wherein, a third sub-module (120(3)) is configured to, in the second time period, provide the traffic prediction (NTP) based on the second operation prediction (P2) if the following two conditions are both verified: - the number of user terminals corresponding to the first operation prediction (P1) is lower than the number of user terminals at the target area (102) evaluated based on the real-time user terminal location indicator (RI), and - the difference between the number of user terminals at the target area (102) evaluated based on the real-time user terminal location indicator (RI) and the number of user terminals corresponding to the first operation prediction (P1) is higher than a threshold value.

7. The system (100) according to any one of the preceding claims, wherein, a third sub-module (120(3)) is configured to: in the second time period, if the first operation prediction (P1) indicates a peak in the number of user terminals in the target area, provide the traffic prediction (NTP) based on the second operation prediction (P2).

8. The system (100) according to any one of the preceding claims, wherein, a third sub-module (120(3)) is configured to receive calendar data (CDATA), the calendar data providing an indication of a planned increase in the number of user terminals in the target area (102), and the third sub-module (120(3)) is further configured to: in the second time period, if the calendar data provides an indication of a planned increase in the number of user terminals in the target area (102) for the first time period, provide the traffic prediction (NTP) based on the second operation prediction (P2).

9. The system according to any one of the preceding claims, wherein, a first sub-module (120(1)) is configured to: - subdivide a range of possible values of the first operation prediction (P1) into a set of sub-intervals (SP1(i)), and - calculate a prediction probability distribution that provides an indication of the reliability of the first operation prediction (P1) corresponding to the value of each sub-interval, wherein a third calculation sub-module (120(3)) is configured to, in the second time period, provide the traffic prediction (NTP) based on the second operation prediction (P2) if the prediction probability distribution PD is lower than a reliability threshold for all sub-intervals.

10. The system (100) according to any one of the preceding claims, wherein, a third sub-module (120(3)) is configured to: in the second time period, if the second operation prediction indicates a peak in the number of user terminals in the target area (102), provide the traffic prediction (NTP) based on the second operation prediction (P2).

11. The system (100) according to any one of the preceding claims, wherein, The system further includes a self-organizing network module (130), and the self-organizing network module is configured to adjust functional parameters of the wireless communication network (105) based on the network traffic prediction (NTP).

12. The system (100) according to any one of the preceding claims, wherein, the system further includes a drone cell module, and the drone cell module is configured to drive a drone of a cell site equipped with the wireless communication network (105) to the target area (102) based on the network traffic prediction (NTP).

Images