Method for determining a route of a mobile terminal from data about multiple network events using said mobile terminal, device and computer program therefor

Description

Domain of invention

[0001] The field of the invention is that of the localization of mobile objects connected to at least one communication network.

[0002] More specifically, the invention relates to a method for determining a route taken by a mobile terminal along roads of a transport network using collected signaling data and the corresponding devices, computer programs and media. Art antérieur et ses inconveniences

[0003] The signaling data collected by a telecommunications operator within the communications network(s) it operates allows it to understand how its users use the resources it makes available to them. With this knowledge, a telecommunications operator can then plan development and maintenance operations for the equipment that makes up the communications networks it operates, enabling it to meet the needs and expectations of its users.

[0004] In recent years, with the development of the Internet of Things or loT (I Internet of Things) and the emergence of connected vehicles, telecommunications operators have realized that the signaling data in their possession could be of interest to other players and that it then becomes an asset to be valued.

[0005] Signaling data from mobile terminals used during travel are particularly of interest for the study of human mobility, whether for example mobile terminals carried by a user or mobile terminals embedded in a vehicle.

[0006] There are many techniques for estimating a mobility situation, and more specifically a route, of a mobile terminal. Among these different techniques, some allow estimating a route taken by a mobile terminal along roads of a transport network using signaling data. According to these techniques, a position of the mobile terminal is estimated, approximated by the center of a coverage area, or cell, of a base station to which the mobile terminal is connected. To do this, a Voronoi partitioning of the territory covered by the cells constituting a radio communication network is used. With this partitioning, each network event is positioned on the center of the cell in which it occurs. Such events are time-stamped, which makes it possible to calculate a speed, and / or a direction of movement knowing the coordinates of the centers of the cells.This information is then used, in conjunction with transport network graphs, to determine the route taken by the mobile terminal.

[0007] However, such techniques have the following limitations specific to Voronoi partitioning: all cells are assumed to be omnidirectional, cell characteristics (radiation power, height and inclination of base station antennas) are not taken into account, overlap of cell action areas is not taken into account, no location a priori is used.

[0008] Regarding the cross-referencing of this information with transport network graphs in order to determine the route taken by the mobile terminal, several approaches are known.

[0009] The probabilistic approach, based on hidden state Markov models or Hidden Markov Models(HMM) which considers the transition between events is a very popular approach, particularly in location and route determination applications such as Microsoft Bing Maps, OpenStreetMap or Mapbox.

[0010] A hidden-state Markov model defines hidden states, in this case road segments or nodes of a transport network, and observable states, network events involving a mobile terminal. The idea of such a probabilistic approach is to associate a road segment with each network event involving a mobile terminal using two types of probabilities: the connection probability (also called transmission probability in this application) and the transition probability. The connection probability evaluates the probability of the mobile terminal considered to connect to a cell of a radio communication network knowing that it is located on a given road segment. The transition probability evaluates the probability of passing from the last road segment before a given road segment to the given road segment. Once these probabilities are obtained, the most probable sequence of road segments is selected as the output of the model.

[0011] Empirically, hidden-state Markov models assume that roads closer to a given position of a mobile terminal, or to the position of the base station serving the mobile terminal when the estimation of the mobile terminal's position is uncertain, have higher probabilities of connection (or transmission). In many cases, the spatial position error, which is obtained using GPS data (Global Positioning System) is modeled as a Gaussian distribution with zero mean. The connection (or emission) probabilities therefore also follow a Gaussian distribution.

[0012] For the transition probability, it is often assumed that the distance between two observable states is close to the road distance between projections of these two observable states on the road network. The transition probability is then modeled according to an exponential distribution as a function of the difference between these two distances.

[0013] When the position of the mobile terminal is obtained using data collected from the radio communication network to which the mobile terminal is attached, the routes taken by a mobile terminal determined using hidden state Markov models based on this information are very uncertain. This is partly due to the fact that the cells of a radio communication network sometimes cover large areas, e.g. greater than 5 km in radius, the position estimates obtained are very uncertain and so are the resulting speed and direction of travel estimates.

[0014] The article COLERI ERGEN SINEM ET AL: “RSSI-Fingerprinting-Based Mobile Phone Localization With Route Constraints”, IEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, vol.63, no.1, pages 423-428, XP011537078 describes a mobile phone localization device based on an RSSI fingerprint with a route constraint. The application FR 3 046 006 A1 describes a method for estimating trajectories using mobile data. The article LV JINHUA ET AL: “BSLoc: Base Station ID-Based Telco Outdoor Localization”, ADVANCES IN DATABASES AND INFORMATION SYSTEMS; PAGE(S) 206-219, XP047502705 describes a localization technique based on an outdoor base station identifier. These documents are related to the technological background.

[0015] There is therefore a need for a solution that does not have all of the above drawbacks for identifying a route taken by a mobile terminal along roads of a transport network by means of signaling data. Exposition of the invention

[0016] The invention meets this need by proposing a method for determining a route of a mobile terminal from data relating to a plurality of network events involving said mobile terminal as defined in claim 1.

[0017] In this application, a network event is understood to mean any event giving rise to a transmission or reception of signaling data between a mobile terminal and a base station of a communication network, such as the establishment of a communication between the mobile terminal and the base station, for example in the event of an incoming or outgoing call or in the event of transmission or reception of a short message or SMS, the triggering of an attachment procedure to the base station, the transmission of a “paging” message to the mobile terminal asking it to exit a standby state, etc.

[0018] This solution helps to overcome all or part of the constraints of the state of the art by using a base station support likelihood map in order to calculate a probability of connection of the mobile terminal to a given base station.

[0019] More particularly, such a base station support likelihood map represents the probability that a mobile terminal has of connecting to a base station at a location in the base station's coverage area. Such a likelihood map does not directly correspond to a spatial density of probability of presence of the mobile terminal.

[0020] Indeed, these likelihood maps of support by a base station offer the advantage of taking into account the direction of the cells, their radiation characteristics, the overlap of their action zones unlike certain state-of-the-art techniques and more particularly certain techniques using Voronoi partitioning.

[0021] More particularly, the first event and the second event are selected from a plurality of network events involving the mobile terminal, occurring during a first time window. In some embodiments, the first event and the second event are two consecutive events.

[0022] This makes it possible to estimate the mobility conditions of a mobile terminal over a short time window of the order of one or several tens of minutes, for example less than or equal to 15 minutes, which is not the case with state-of-the-art techniques which have difficulty extracting relevant information over short time windows due to their poor consideration of the uncertainties in the location of mobile terminals. In addition, this makes the present solution compatible with the legislative provisions relating to the retention of event history relating to mobile terminals limiting the duration of history retention (for example a limitation to a retention duration of the order of 15 minutes), unlike certain state-of-the-art solutions.

[0023] This solution proposes to use a chain of mobility data relating to a mobile terminal in association with one or more transport network graphs, e.g. road or motorway network, railway network, metropolitan network, etc., in order to determine a route taken by the mobile terminal. To this end, this solution implements a hidden-state Markov model in which the observed states correspond to the data relating to network events involving the mobile terminal and the hidden states correspond to the route segments of the transport network graphs. Expressed in other words, the proposed solution is adapted to the spatial uncertainty of a radio communication network and to the sporadic data relating to network events.

[0024] In some embodiments, the selected route is the one for which the product of the set of connection (transmission) probabilities with the set of previously determined transition probabilities is the highest.

[0025] In some embodiments, the probability of connection (or transmission) P ( A | s 1) for a given network event is determined according to the following formula: P A s 1 = ∑ j = 1 n P A p j n in which A represents the first base station involved in the first event, s 1 represents the first candidate road segment located within the coverage area of said first base station, pjrepresents a pixel of a likelihood map representing a probability of connection of said mobile terminal to the first base station when the terminal is located on a pixel of said likelihood map, and n represents the number of pixels of said non-zero weight likelihood map crossed by the first candidate road segment s 1 .

[0026] Thus, this probability of connection (or transmission) is determined on the basis of mobile terminal location data that is more precise and more relevant than certain state-of-the-art solutions.

[0027] In some embodiments, a selection of said at least one first candidate road segment taking into account a distance between said at least one first candidate road segment and at least one candidate road segment at a time t0 prior to time t1

[0028] More particularly, in certain embodiments, the first candidate road segment is selected from among the road segments located at a distance less than or equal to a first distance, called the connectivity tolerance distance, from all the candidate road segments at a time t0 prior to time t1.

[0029] In some embodiments, the value of the first connectivity tolerance distance may take into account a geographic distribution density of the base stations and a median radius of a coverage area of base stations near said first base station.

[0030] In some embodiments, the method may comprise a step of selecting the first candidate road segment comprising: a determination of a convex hull consisting of all the candidate road segments at a time t0 prior to time t1, a selection of said at least one first candidate road segment located at a distance less than or equal to a first threshold, called the connectivity tolerance threshold.

[0031] This may help, at least in some embodiments, to reduce the number of candidate road segments and thus reduce the computation time and / or the required computing power.

[0032] In some embodiments, the value of the connectivity tolerance threshold is the product of a weighting coefficient representative of a geographic distribution density of the base stations with a median radius of a base station coverage area knowing a radius of a coverage area of at least one base station located at a temporal distance from said first base station less than or equal to a second threshold.

[0033] Thus, the lower the density of base stations and therefore the more distant the base stations are from each other, as is the case in the countryside, the higher the weighting coefficient. Conversely, the higher the density of base stations and therefore the closer the base stations are from each other, as is the case in urban areas, the lower the weighting coefficient.

[0034] In some embodiments, the connection (or transmission) probability is weighted using a penalty coefficient representative of a deviation of a direction of the first candidate road segment with a reference direction, the penalty coefficient tending towards zero as the deviation between the direction of the first candidate road segment and the reference direction increases.

[0035] This may help, at least in some embodiments, to favor candidate road segments with the smallest deviation from a reference direction.

[0036] In some embodiments, the transition probability P ( s 2 | s 1) for a given network event is determined according to the following formula: P s 2 s 1 = ∫ d min d max f u du in which: s1 represents the first candidate road segment located within the coverage area of said first base station, s 2 represents the second candidate road segment located within the coverage area of said second base station, f ( u ) represents a density of distribution of distances between the coverage areas of the first base station and the second base station, d min represents the minimum distance separating the first candidate road segment and the second candidate road segment on a graph representing said transport network and d max represents the sum of d min with the length of the first candidate road segment and the length of the second candidate road segment

[0037] According to this aspect of the present method, the transition probability is determined by taking into account a distribution density of distances between the coverage areas of the first base station and the second base station. Such a choice can help, at least in certain embodiments, to take into account an uncertainty of movement of the mobile terminal inherent in the radio communication network. Indeed, an uncertainty as to the location of the cells of a radio communication network can lead, in certain cases, to an uncertainty as to the distance traveled by a mobile terminal within this radio communication network.

[0038] Thus, this distance distribution density relates to actual movements of the mobile terminal, in other words. This corresponds to a situation in which the mobile terminal has changed its home base station following an (actual) movement, excluding situations in which the mobile terminal has not left the coverage area of a base station and is therefore immobile from the point of view of the radio communications network.

[0039] In some embodiments, the transition probability P ( s 2 | s 1) for a given network event is determined according to the following formula: P s 2 s 1 = ∫ v min v max f v dv in which s 1 represents the first candidate road segment located within the coverage area of said first base station, s 2 represents the second candidate road segment located within the coverage area of said second base station, f (v ) represents an average speed density of movement of the mobile terminal, v min = d min / Δ t Or d min represents the minimum distance separating the first candidate road segment and the second candidate road segment on a graph representing said transport network and v max = d max / Δ t Or d max represents the sum of d min with the length of the first candidate road segment and the length of the second candidate road segment and where Δ t = t2-t1..

[0040] In some embodiments, the transition probability P ( s 2 | s 1) for a given network event is determined according to the following formula: P s 2 s 1 = ∫ θ min θ max f θ dθ in which s 1 represents the first candidate road segment located within the coverage area of said first base station, s2 represents the second candidate road segment located within the coverage area of said second base station, f ( i ) represents an average direction density of movement of the mobile terminal between the coverage areas of the first base station and the second base station, θ min represents the lower bound of an intersection of a first angular sector defining possible directions of movement for the mobile terminal with a second angular sector defining a maximum angle between a first end of the first candidate road segment and a second end of the second candidate road segment and θ max represents the upper bound of the intersection of the first angular sector and the second angular sector.

[0041] In some embodiments, the transition probability is determined based on at least two densities representative of a movement of the mobile terminal among the density of distribution of distances between the coverage areas of the first base station and the second base station f ( u ), the average speed density of movement of the mobile terminal f ( v ) and the average direction density of movement of the mobile terminal between the coverage areas of the first base station and the second base station f ( i ).

[0042] This may help, at least in some embodiments, to improve the determination of the route taken by the mobile terminal since the data provided as input to the hidden state Markov model takes into account many aspects of the movement of a mobile terminal such as its speed of movement as well as the direction of this movement and the distance traveled.

[0043] In some embodiments, the transition probability is weighted based on the number of intersections of transportation network road segments encountered along the candidate route.

[0044] In some embodiments, the transition probability is weighted using a penalty coefficient representative of a number of intersections of transportation network road segments encountered along the candidate route, the penalty coefficient tending to zero as the number of intersections increases.

[0045] This may help, at least in some embodiments, to favor transitions between candidate road segments involving the fewest possible nodes.

[0046] The invention also relates to a device capable of determining a route taken by a mobile terminal from data relating to a plurality of network events involving said mobile terminal as defined in claim 2.

[0047] Such a device, suitable for implementing the method which is the subject of the invention in any of its embodiments or according to any combination of the embodiments described in the present document, can for example be embedded in a server belonging to the telecommunications operator operating the radio communication network to which the base stations belong.

[0048] The invention also relates to a computer program product comprising program code instructions for implementing a method as described above, according to any one of its embodiments, when executed by a processor.

[0049] The invention also relates to a computer-readable recording medium on which is recorded a computer program comprising program code instructions for executing the steps of the method according to the invention as described above according to any one of its embodiments.

[0050] Such a recording medium may be any entity or device capable of storing the program. For example, the medium may include a storage medium, such as a ROM, for example a CD-ROM or a microelectronic circuit ROM, or a magnetic recording medium, for example a USB flash drive or a hard disk.

[0051] On the other hand, such a recording medium may be a transmissible medium such as an electrical or optical signal, which may be conveyed via an electrical or optical cable, by radio or by other means, so that the computer program contained therein is remotely executable. The program according to the invention may in particular be downloaded over a network, for example the Internet.

[0052] Alternatively, the recording medium may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the method which is the subject of the aforementioned invention. List of figures

[0053] Other aims, characteristics and advantages of the invention will appear more clearly on reading the following description, given as a simple illustrative, and non-limiting, example, in relation to the figures, among which: [ Fig. 1] : this figure represents a diagram of the steps of a method for determining a route taken by a mobile terminal from data relating to a plurality of network events involving said mobile terminal; [ Fig. 2A ] : this figure represents an example of a likelihood map of support by a cell associated with the base station obtained with a a priori of uniform presence of the mobile; [ Fig. 2B ] : this figure represents an example of a cell support likelihood map obtained with a a priori presence of the mobile along a road axis and a high-speed train line; [ Fig. 3 ] : this figure represents the notion of range of a base station; [ Fig. 4 ]: this figure represents the sequence of steps leading to obtaining a density of distribution of distances between the coverage areas of a first base station A 1 and a second base station A 2 ; [ Fig. 5 ] : this figure represents the sequence of steps leading to obtaining an average speed density of movement of the mobile terminal; [ Fig. 6 ] : This figure represents the sequence of steps leading to obtaining an average direction density of movement of the mobile terminal between the coverage areas of the first base station A 1 and the second base station A 2 ; [ Fig. 7A ] : this figure represents a situation in which the mobile terminal is stationary from the point of view of the radio communications network; [ Fig. 7B ] :this figure represents a situation in which the mobile terminal is mobile from the point of view of the radio communications network; [ Fig. 8 ] : this figure represents a device capable of implementing certain steps of the solution previously described. Detailed description of the methods of implementation of the invention

[0054] The general principle of the invention is based on the use of signaling data collected by a telecommunications operator operating at least one radio communication network to determine a route taken by a mobile terminal along roads of one or more transport networks. More particularly, in at least some of the detailed embodiments, one of the signaling data used in the present solution may be a likelihood map of support by a base station. As already explained above (part "disclosure of the invention"), such a likelihood map represents the probability that a mobile terminal has of connecting to a base station at a location in the coverage area of the base station.Thus, in at least some embodiments of the present method, the use of a likelihood map can help to improve the results obtained, compared to prior art solutions, in terms of, for example, reliability, accuracy and / or realism.

[0055] Such knowledge can also help, for example, to better classify the different types of goods transport vehicles and their uses, to better manage fleets of bicycles or scooters made available to the public, to better track postal parcels equipped with connected trackers, and / or to better plan the flow of passengers on public transport, etc.

[0056] There [ FIG. 1 ]represents a diagram of the steps of a method for determining a route taken by a mobile terminal from data relating to a plurality of network events involving said mobile terminal.

[0057] Such a method is based on the implementation of a hidden-state Markov chain. The present solution then consists of using a chain of mobility data relating to a mobile terminal in association with one or more transport network graphs, e.g. road or motorway network, railway network, metropolitan network, etc., in order to determine a route taken by the mobile terminal. To do this, the method implements a hidden-state Markov model in which the observed states correspond to the data relating to network events involving the mobile terminal and the hidden states correspond to route segments of transport network graphs.

[0058] In a step G1, signaling data is collected, for example by means of probes arranged in a radio communication network. This signaling data is then stored in one or more databases. In such a database, each entry corresponds, for example, to a network event.

[0059] In the detailed embodiments, the signaling data collected for a network event includes, among other things: an identifier of a mobile terminal (e.g. a pseudonym uniquely identifying the mobile terminal for the operator), timestamp data of the occurrence of the network event, an identifier of the base station or cell that is associated with the network event.

[0060] In the detailed embodiments, enriched signaling data may also be stored in the database. Thus, a network event may also be associated with a cell support likelihood map associated with the base station with which the mobile terminal interacted when the network event occurred.

[0061] A cell support likelihood map associated with a base station may, for example, be obtained by simulation. Knowing that the coverage area of a base station consists of a plurality of cells, a plurality of support likelihood maps may be associated with the same base station. For example, for a base station comprising an antenna A i , a support likelihood map may represent a probability P (A i |(X, Y) ∈ pk ) of connection of the mobile terminal to the antenna A i (with pkone of the pixels of the map), at the point of coordinates (X, Y) knowing that the mobile terminal is located on the pixel pk.

[0062] The geographical map representing the coverage area of the base station being partitioned into pixels, the formula for total probabilities makes it possible to obtain the probability, for a mobile terminal to be located on the pixel pk knowing that it is connected to the antenna A i of the base station: P X Y ∈ p k A i = P X Y ∈ p k P A i X Y ∈ p k P A i = P X Y ∈ p k P A i X Y ∈ p k ∑ k ′ P X Y ∈ p k ′ P A i X Y ∈ p k ′ which is equivalent, noting P((X, Y) ∈ pk ) = Prior(k) the a priori of mobile presence on the pixel pk, at P X Y ∈ p k A i = Prior k P A i X Y ∈ p k ∑ k ′ Prior k ′ P A i X Y ∈ p k ′ .

[0063] From these probabilities and choosing a a priori of the presence of the mobile terminal we obtain a spatial density map of presence.

[0064] Such a a priorican be chosen, for example, from a uniform presence, a presence along road axes, and / or a presence according to a population density, etc. Such a geographical map representative of the coverage area of the base station is, for example, obtained by simulations. The a priori presence of the chosen mobile may depend on the population of mobile terminals that we wish to monitor. According to a first example, we can choose a a priori mobile presence when we want to determine the distance traveled over the entire coverage area of the base station. According to a second example, we can choose a a priori presence along the roads if we want to determine the distance traveled by mobile terminals embedded in vehicles. According to a third example, we can choose a a prioripresence based on population density if we wish to monitor the movement of mobile terminals in residential areas.

[0065] An example of a cell support likelihood map associated with the base station obtained with a a priori uniform presence of the mobile is represented at the [ FIG. 2A ].

[0066] Another example of a cell support likelihood map obtained with a a priori of the presence of the mobile along a road axis and a high-speed train line is represented in the [ FIG. 2B ].

[0067] In some embodiments, an event may also be associated, in the database, with a value of a range of an antenna of the corresponding base station. This range may for example be expressed in meters. Such a range may range from a few tens of meters to several hundreds of meters.

[0068] Thus, for a cell of a base station (for example each cell of the base station), it may be possible to define a radius of action R such that the probability of presence of a mobile terminal in a disk of radius R around the center of this cell knowing that the probability of the occurrence of a network event at a given instant is greater than or equal to a first probability (such as 85% in the detailed example). Such a network event is, for example, a connection of a mobile terminal to an antenna of the base station serving the cell considered at a given instant.

[0069] The range R may, for example, depend in particular on the technology used (depending on whether it complies with 2nd, 3rd, 4th (LTE) or 5th generation telecommunications standards) and / or the nature of the geographical area concerned, e.g. rural, urban or peri-urban.

[0070] In other words, for an antenna A i whose coordinates on the geographic map are ( xi , yi ), having a radius of action R i , and noting D Ri the disk of radius R i centered on the center of the cell considered, we have: P (( X, Y ) ∈ D Ri | A i ) ≥ 85% by noting (X, Y) the pair of random variables giving the position of the mobile terminal at time ti.

[0071] There [ FIG. 3 ] illustrates this notion of range of action.

[0072] The present method is based on the use of signaling data relating to mobile terminals, and therefore to their users. Also, in certain embodiments its implementation must comply with regulatory constraints of anonymization, and / or pseudonymization, of the data. These constraints may also sometimes impose (relatively short) deadlines for anonymization, and / or pseudonymization, of the data. Thus, in certain embodiments, the calculations to be carried out use signaling data whose history (i.e. conservation) must not exceed a certain duration. Such a duration, or time window, may be one or several tens of minutes depending on the embodiments, for example 15 minutes. The calculations themselves must therefore be carried out in a time less than this duration.

[0073] In a step G2, one or more transport network graphs, such as maps of road or motorway networks, rail networks, metropolitan networks, etc. corresponding to a geographical area in which the mobile terminal has moved are obtained.

[0074] In a G3 step, a connection (or emission) probability P ( A 1 | s 1) for a first network event E1 is determined using the following formula: P A 1 s 1 = ∑ j = 1 n P A p j n in which A 1 represents the base station involved in the first E1 network event, s 1 represents a candidate road segment located within the coverage area of the base station A 1 , pj represents a pixel of the likelihood map representing a probability of connection of said mobile terminal to the base station A1 when the terminal is located on a pixel of this likelihood map, and n represents the number of pixels of the non-zero weight likelihood map crossed by the candidate road segment s 1. Such a candidate road segment s 1 is for example a section of road or motorway, a section between two metro stations, etc.

[0075] In other words, P ( A 1 | s 1) represents the average of the probabilities of connection of the mobile terminal to the base station A 1 knowing that the mobile terminal is located, at the time t1 at which the network event E1 occurred, on a pixel of the likelihood map crossed by the candidate road segment s 1

[0076] In some embodiments, in order to reduce the computation time, prior to calculating the connection (or transmission) probability P ( A 1 | s1), candidate road segments can be selected yes .

[0077] To do this, firstly, a convex hull consisting of a set of candidate road segments at a time t0 prior to time t1 is determined.

[0078] In a second step, a value of an STC connectivity tolerance threshold is calculated, for example as the product of a weighting coefficient oh representative of a geographic distribution density of base stations with a median radius of a base station coverage area knowing the radius of a coverage area of at least one base station located at a temporal distance δ from the base station considered less than or equal to a second threshold S. A value of d allowing the determination of a relevant value of the STC connectivity tolerance threshold is for example 15 minutes.

[0079] Thus, the lower the density of presence of base stations and therefore the more distant the base stations are from each other, as is the case in the countryside, the higher the weighting coefficient oh is large. Conversely, the greater the density of presence of base stations and therefore the closer the base stations are to each other, as is the case in urban areas, the higher the weighting coefficient oh is weak.

[0080] Then selected as candidate road segments for the network event E1 are all the candidate road segments located at a distance from the connected envelope consisting of all the candidate road segments at a time t0 less than or equal to a first connectivity tolerance distance STC.

[0081] In order to improve the accuracy of the connection (or emission) probability calculation, the connection probability may be weighted using a penalty coefficient representative of a deviation from a direction of a candidate road segment yes with a reference direction θ model, the penalty coefficient tending to zero as the deviation between the direction of the candidate road segment yes considered and the reference direction the model increase.

[0082] For example, the mathematical law linking the value of the penalty coefficient with the value of the deviation between the direction of the candidate road segment yes considered and the reference direction the model is given by the following formula: y = λ e − λx Or y corresponds to the value of the penalty coefficient, x to the value of the deviation between the direction of the candidate road segment yesconsidered divided by an angle of 45° and the reference direction the model And l a constant. In the remainder of this document it is considered that l is worth 1. Of course, l can take other values depending on the specific case. Similarly, the angle value can be different from 45°. The angle value can be chosen so that the value of the member e - λx < does not drop abruptly when the deviation increases. This improves the performance obtained.

[0083] To further reduce the number of candidate road segments for a given network event E1, once a connection (or transmission) probability has been calculated for each of the candidate road segments, they can be ranked in order of increasing connection probabilities. The n candidate road segments with the highest connection probabilities are then selected. The number n can, for example, be determined arbitrarily by applying the following formula: n = max 15 , ceil n 1 / n 2 Or n 1 is the number of candidate road segments for network event E1, the number n2 is chosen arbitrarily, ceiling is the function that rounds up to the next higher number. The choice of the value of the number n2, here 10 for example, impacts the performance obtained. Thus, depending on the cases encountered, a value other than 10, smaller or larger, can help improve performance.

[0084] At the end of step G3, a connection (or emission) probability was calculated for all candidate road segments for a given network event.

[0085] In a G4 step, a transition probability P ( s 2 | s 1) between a first candidate route segment corresponding to a first network event E1 and a second candidate route segment corresponding to a second network event E2 occurring at a time t2 and involving a second base station A 2 is determined.

[0086] In a first implementation, the transition probability P ( s 2 | s 1) is determined as follows: P s 2 s 1 = ∫ d min d max f u du in which s 1 represents the candidate road segment s 1 located within the coverage area of the base station A 1 , s2 represents the second candidate road segment located within the coverage area of the base station A 2, f ( u ) represents a density distribution of distances between the coverage areas of the first base station A 1 and the second base station A 2, d min represents the minimum road distance separating the candidate road segment s 1 and the second candidate road segment s 2 and d max represents the sum of d min with the length of the first candidate road segment s 1 and the length of the second candidate road segment s 2. In other words, d max represents the maximum road distance separating the first candidate road segment s 1 and the second candidate road segment.

[0087] Such a density of distribution of distances between the coverage areas of the first base station A 1 and the second base station A 2 f ( u ) is obtained in the following manner with reference to the [ Fig. 4 ].

[0088] During a first step E1, a first network event ER1 and a second network event ER2 are selected from a plurality of network events ER i involving the mobile terminal. Such a selection consists of taking all the pairs of network events such that the two network events ER1 and ER2 constituting a pair of network events are present in the time window considered, that they are separated by a duration at least equal to a certain duration (called minimum), for example 5 minutes, and are such that the network event ER2 is later than the network event ER1.

[0089] The first and second network events ER1, ER2 occur respectively at the times t 1 and t 2 and involve the antennas A 1 and A 2 carried by two base stations which can be either different or the same.

[0090] For i ∈ {1, 2} we note ( X i , Y i ) the random variables giving respectively the longitude and latitude of the mobile terminal at the instant you .

[0091] In the remainder of the document, the following hypotheses are made: random variables ( X i , Y i ) associated with the probability densities of presence of the two antennas A 1 and A 2 are independent, the movement of the mobile terminal is assumed to be uniformly rectilinear if the time separating two events is less than a first duration (for example a constant duration playing the role of a minimum threshold).

[0092] In a step E2, a distance density is determined f D 12 for two network events ER1 and ER2 belonging to the first set of network events JER1 Using the formula for total probabilities and the assumptions cited above, we can prove that the random variables ( X i , Y i ) follow a density law fi . Of course, other calculation methods can be used to obtain the distance density. f D 12 .

[0093] We then pose D 12 = X 2 − X 1 2 + Y 2 − Y 1 2 Or D 12 represents the random variable giving the distance traveled by the mobile terminal between the instants t 1 and t 2 .

[0094] The random variable D 12 follows a density law f D 12, such that: ∀ d ∈ ℝ + : f D 12 d = ∫ ℝ 3 f 1 x 1 y 1 ∑ ± f 2 x 2 , y 1 ± d 2 − x 2 − x 1 2 d 1 d > x 2 − x 1 x 1 x 2 d d 2 − x 2 − x 1 2 dx 1 dx 2 dy 1

[0095] By integrating this density over a distance interval [ yes, db ] included in ℝ + and noting E d a d b = x , y , x ′ , y ′ ∈ ℝ 4 / y ′ − y ≤ d b 2 − x ′ − x 2 et d a ≤ x ′ − x ≤ d b , we obtain, after inversion and using changes of variables y 2 = y 1 ± d 2 − x 2 − x 1 2 in the appropriate integrals, a much more convenient form for numerical calculation: ∫ d a d b f D 12 d = ∫ ℝ 4 f 1 x 1 y 1 f 2 x 2 y 2 1 E d a d b x 1 x 2 y 1 y 2 dx 1 dx 2 dy 1 dy 2

[0096] A method to calculate this distance density f D 12 perhaps, in certain embodiments, to create a spatial mesh of the two geographical maps representative of a coverage area of a base station associated with an antenna A i , a first geographic map corresponding to the base station involved in the network event ER1 and a second geographic map corresponding to the base station involved in the network event ER2. Each pixel pk of this mesh can be associated with a probability of support by the antenna Ai conditioned by the presence of the mobile terminal in this pixel P(A i |(X, Y) ∈ pk ). Such information is provided for example by the cell support likelihood map associated with the antenna A i .

[0097] In a second implementation, the transition probability P ( s 2 | s 1) is determined as follows: P s 2 s 1 = ∫ v min v max f v dv in which s 1 represents the first candidate road segment within the coverage area of the base station A 1 , s 2 represents the second candidate road segment located within the coverage area of said second base station A 2, f ( v ) represents an average speed density of movement of the mobile terminal, v min = d min / Δ t Or d min represents the minimum road distance separating the first candidate road segment s 1 and the second candidate road segment s 2 and vmax = d max / Δ t Or d max represents the sum of d min with the length of the first candidate road segment and the length of the second candidate road segment and where Δ t = t2-t1.

[0098] In reference to the [ Fig. 5 ], in a step H1, a first network event ER1 and a second network event ER2 are selected from a plurality of network events ER i involving the mobile terminal. Such a selection consists of taking all the pairs of network events such that the two network events ER1 and ER2 constituting a pair of network events are present in the time window considered, that they are separated by a duration at least equal to a certain duration (called minimum), which can be set in the order of a few seconds to a few minutes, for example 5 or 6 minutes, and are such that the network event ER2 is later than the network event ER1.

[0099] The first and second network events ER1, ER2 occur respectively at the times t 1 and t 2 and involve the antennas A 1 and A 2 carried by two base stations which can be different or the same

[0100] For i ∈ {1, 2} we note ( X i , Y i ) the random variables giving respectively the longitude and latitude of the mobile terminal at the instant you .

[0101] In the remainder of the document, the following hypotheses are made: random variables ( X i , Y i ) associated with the probability densities of presence of the two antennas A 1 and A 2 are independent, the movement of the mobile terminal is assumed to be uniformly rectilinear if the time separating two events is less than a first duration (for example a constant duration playing the role of a minimum threshold).

[0102] In a step H2, we determine a distance density f D 12 . Using the total probability formula and the assumptions mentioned above, we can prove that a pair of variables ( X i , Y i ) follows a density law fi . Of course, other calculation methods can be used to obtain the distance density. f D 12 .

[0103] We then pose D 12 = X 2 − X 1 2 + Y 2 − Y 1 2 Or D 12 represents the random variable giving the distance traveled by the mobile terminal between the instants t 1 and t 2 .

[0104] The random variable D 12 follows a density law f D 12 , such as : ∀ d ∈ ℝ + : f D 12 d = ∫ ℝ 3 f 1 x 1 y 1 ∑ ± f 2 x 2 , y 1 ± d 2 − x 2 − x 1 2 d 1 d > x 2 − x 1 x 1 x 2 d d 2 − x 2 − x 1 2 dx 1 dx 2 dy 1

[0105] Since the mobile terminal is assumed to move in a uniform rectilinear motion, a velocity density is then calculated using the following change of variable v = d t 2 − t 1 , which gives: f V 12 v = t 2 − t 1 f D 12 d

[0106] By integrating this density over a distance interval [ yes, db ] included in ℝ + and noting E d a d b = x , y , x ′ , y ′ ∈ ℝ 4 / y ′ - y ≤ d b 2 − x ′ - x 2 et d a ≤ x ′ - x ≤ d b , we obtain, after inversion and using changes of variables y 2 = y 1 ± d 2 − x 2 − x 1 2 in the appropriate integrals, a much more convenient form for numerical calculation: ∫ d a d b f D 12 d dd = ∫ ℝ 4 f 1 x 1 y 1 f 2 x 2 y 2 1 E d a d b x 1 x 2 y 1 y 2 dx 1 dx 2 dy 1 dy 2

[0107] A method to calculate this distance density f D12 may create a spatial mesh of the two geographic maps representing a coverage area of a base station associated with an antenna A i Each pixel pk of this mesh can be associated with a probability of support by the antenna Ai conditioned by the presence of the mobile terminal in this pixelP (A i |(X, Y) ∈ pk ). Such information is provided for example by the cell support likelihood map associated with the antenna A i .

[0108] Thus, in certain embodiments, at the end of a step H2, a distance density can be obtained f D 12 and therefore, through the change of variable v = d t 2 − t 1 , a speed density for a pair of network events ER1, ER2.

[0109] In order to improve the accuracy of the value of the movement speed of a mobile terminal, steps H1 and H2 may be repeated, in certain embodiments, for a plurality of pairs of events (on the condition, for example, that the instants youassociated with each of the events are included in the historical time window (15 minutes for example) defined above). Of course, the duration of the time window can take any other value depending on needs, legislation, etc.

[0110] Thus, once all the pairs of events ER i , ER j taken within the time window have been formed, steps H1 and H2 are implemented for each of these pairs of events ER i , ER j . At the end of these different iterations of steps H1 and H2, we can obtain as many densities of speed of movement of the mobile terminal as there are pairs of events ER i , ER j .

[0111] In certain embodiments, a step H3 during which a combination of the different densities of movement speed of the mobile terminal obtained can be implemented. Such a combination can for example result in a probability density of average movement speed of the mobile terminal for a duration less than or equal to that of the time window considered.

[0112] Such embodiments may for example be based on at least one additional hypothesis. Thus, the existence of a speed law relating to the movement of the mobile terminal may be assumed. For example, it may be assumed that this speed law relating to the movement of the mobile terminal may be obtained from the different densities of movement speeds of the mobile terminal corresponding to the pairs of events ER i , ER j .

[0113] In the detailed embodiments, in order to obtain the probability density of the average speed of movement of the mobile terminal over the duration of the time window considered, it is possible, for example, to consider two pairs of events, a first pair of events C 1 consisting of events ER1 and ER2 and a second pair of events C 2 consisting of events ER3 and ER4, as well as the corresponding densities of speed of movement of the mobile terminal obtained at the end of the implementation of steps E1 and E2, and noted respectively V 1 and V 2 .

[0114] We denote by V the random variable representing the average speed density of movement of the mobile terminal, which we seek to obtain from the speed densities of movement of the mobile terminal V 1 and V 2 .

[0115] Knowing the following property: ∀ v , v ′ ∈ ℝ + 2 , P V ∈ v , v ′ > 0 ⇔ P V 1 ∈ v , v ′ > 0 et P V 2 ∈ v , v ′ > 0 which says that for the event, in the sense of probabilities, "V ∈ [v, v']" to occur with a non-zero probability, it is necessary and sufficient that the events, in the sense of probabilities, "V i ∈ [v, v']" occur with a non-zero probability.

[0116] Noting that the mobile terminal's movement speed densities V 1 and V 2 are independent of each other by construction, property (3) is then rewritten: ∀ v , v ′ ∈ ℝ + 2 , P V ∈ v , v ′ > 0 ⇔ P V 1 > 0 ∩ V 2 > 0 > 0

[0117] Considering the average speed density of the mobile terminal V verifying equation (3') as the result of two random experiments whose order does not matter, it is then possible to write, noting I v,v' = [v, v'] for all v, v': P V ∈ I v , v ′ = P V 1 , V 2 ∈ I v , v ′ 2 + P V 1 ∉ I v , v ′ ∩ V 2 ∈ I v , v ′ 1 P V 1 ∈ I v , v ′ > 0 v , v ′ + P V 1 ∈ I v , v ′ ∩ V 2 ∉ I v , v ′ 1 P V 2 ∈ I v , v ′ > 0 v , v ′ which can be rewritten, thanks to the independence of the two densities of speed of movement of the mobile terminal V 1 and V 2 , in the form: P V ∈ I v , v ′ = 1 − P V 1 ∉ I v , v ′ P V 2 ∉ I v , v ′ 1 P V 1 ∈ I v , v ′ ∩ V 2 ∈ I v , v ′ > 0 v , v ′

[0118] Such an expression is easily generalizable to n independent mobile terminal movement speed densities where n corresponds to the number of pairs C i of events constituted for a given time window.

[0119] Equation (4) can then be normalized to verify the following property: P V ∈ ℝ + = 1 , which reflects the fact that the average speed of movement of the mobile terminal sought is positive or zero.

[0120] In certain embodiments, from the average movement speed density of the mobile terminal as determined for example above, it is possible to obtain, in a step H4, a value of an average movement speed of the mobile terminal by calculating the expectation of a probability law associated with the average movement speed density of the mobile terminal V.

[0121] Thus, a 95% confidence interval can be determined, for example, to obtain a value of the average speed of movement of the mobile terminal over the time window considered. Of course, other methods of determining a value of the average speed of movement of the mobile terminal over the time window considered can be implemented in certain embodiments, such as calculating the median.

[0122] In some embodiments, to mitigate the network oscillation phenomenon, the instants you And tjcorresponding respectively to a network event ER i and to an event ER j constituting a pair C i of network events, can be chosen to be separated by a minimum duration. Thus, the duration elapsed between the occurrence of the event ER i and the event ER j is greater than or equal to a first duration. Such a duration, acting as a threshold, can for example when the time window considered lasts 15 minutes, be fixed in the order of a few seconds to a few minutes (for example from 3 to 9 minutes), such as a duration of 5 or 6 minutes. Of course other values of this first duration can be envisaged.

[0123] In a third implementation, the transition probability P ( s 2 | s 1) is determined as follows: P s 2 s 1 = ∫ θ min θ max f θ dθ in which s 1 represents the first candidate road segment located within the coverage area of said first base station,s 2 represents the second candidate road segment located within the coverage area of said second base station, f ( i ) represents an average direction density of movement of the mobile terminal between the coverage areas of the first base station A 1 and the second base station A 2, θ min represents the lower bound of an intersection of a first angular sector defining possible directions of movement for the mobile terminal with a second angular sector defining a maximum angle between a first end of the first candidate road segment and a second end of the second candidate road segment and θ maxrepresents the upper bound of the intersection of the first angular sector and the second angular sector. More specifically, the first angular interval defines a 95% confidence interval for the estimated direction of the mobile terminal. The second angular sector can also be defined as the angular sector between the upper bound and the lower bound of the four angles formed by the lines connecting the ends of the first candidate road segment s 1 and the ends of the second candidate road segment s 2 .

[0124] In reference to the [ Fig. 6 ], the variable representing a movement of the mobile terminal is the direction of movement.

[0125] Since obtaining a value for a direction of movement of a mobile terminal is based on the use of signaling data relating to mobile terminals, and therefore to their users, it may be necessary to comply with anonymization, or pseudoanonymization, constraints in the short term. Thus, the calculations to be carried out use signaling data whose history does not exceed a certain duration. Such a duration is, for example, 15 minutes.

[0126] In order to determine the value of a direction of movement of a mobile terminal, in a step S1 a first network event ER1 and a second network event ER2 are selected from among a plurality of network events ER i involving the mobile terminal.

[0127] The first and second network events ER1, ER2 occur respectively at the times t 1 and t 2 and involve the antennas A 1 and A2 carried by two different base stations.

[0128] For i ∈ {1, 2} we note ( X i , Y i ) the random variables giving respectively the longitude and latitude of the mobile terminal at the instant you .

[0129] In a step S2, a direction density is determined f i 12 . Using the total probability formula and the hypotheses cited above, we prove directly that the pair of variables ( X i , Y i ) follows a density law fi .

[0130] We then pose θ 12 = arctan Y 2 − Y 1 X 2 − X 1 Or i 12 represents the random variable giving the direction of movement of the mobile terminal between the instants t 1 and t 2 .

[0131] The random variable i 12 follows a density law f i 12 , such that: ∀ i ∈ [0, 2π[: f θ 12 θ = ∫ ℝ 3 f 1 x 1 y 1 f 2 x 2 , y 1 + x 2 − x 1 tan θ x 2 − x 1 1 + tan 2 θ 1 x 2 ≠ x 1 1 θ ≠ k + 1 2 π , k ϵ ℤ dx 1 dx 2 dy 1

[0132] Since the mobile terminal is assumed to move in a uniform rectilinear motion, a direction density is then calculated as follows: By integrating this density over an angle interval [ θ a , θ b ] included in [0, 2π[\ k + 1 2 π , k ϵ ℤ , and noting E θ a θ b = x , y , x ′ , y ′ ∈ ℝ 4 − y ′ ϵ y + x ′ − x tan θ a , (x' - x ) tan ( θ b )]}, we obtain, after inversion and using changes of variables y 2 = y 1 + (x 2 - x 1 ) tan ( i ) in the appropriate integrals, a much more convenient form for numerical calculation: ∫ θ a θ b f θ 12 θ dθ = ∫ ℝ 4 f 1 x 1 y 1 f 2 x 2 y 2 1 E θ a θ b x 1 x 2 y 1 y 2 dx 1 dx 2 dy 1 dy 2

[0133] Thus, at the end of step S2, we obtain a direction density f i 12 for a pair of network events ER1, ER2.

[0134] In order to improve the accuracy of the value of the direction of movement of a mobile terminal, steps S1 and S2 may also be repeated for a plurality of pairs of events provided that the instants youassociated with each of the events are included in the 15-minute time window defined above.

[0135] Thus, once all the pairs of events ER i , ER j taken within the time window have been formed, steps S1 and S2 can be implemented for each of these pairs of events ER i , ER j . At the end of these different iterations of steps S1 and S2, we obtain for example as many densities of direction of movement of the mobile terminal as there are pairs of events ER i , ER j .

[0136] In a step S3, the different densities of direction of movement of the mobile terminal obtained can be combined with each other. The result of such a combination gives a probability density of average direction of movement of the mobile terminal over the duration of the time window considered.

[0137] For this purpose, an additional hypothesis must be added. Thus, the existence of a direction law relating to the displacement of the mobile terminal is assumed. It is assumed that this direction law relating to the displacement of the mobile terminal can be obtained from the different densities of directions of displacement of the mobile terminal corresponding to the pairs of events ER i , ER j .

[0138] In order to obtain the probability density of the average direction of movement of the mobile terminal over the duration of the time window considered, two pairs of events are considered, a first pair of events C 1 consisting of events ER1 and ER2 and a second pair of events C 2 consisting of events ER3 and ER4, as well as the corresponding densities of direction of movement of the mobile terminal obtained at the end of the implementation of steps S1 and S2, and noted respectively i 1 and i 2 .

[0139] We note i the random variable representing the average direction density of movement of the mobile terminal, which we seek to obtain from the direction densities of movement of the mobile terminal i 1 and i 2 .

[0140] Knowing the following property: ∀ θ , θ ′ ∈ 0 , 2 π 2 , P θ ∈ θ , θ ′ > 0 ⇔ P θ 1 ∈ θ , θ ′ > 0 et P θ 2 ∈ θ , θ ′ > 0 which says that for the event, in the sense of probabilities, " i ∈ [ i , I will ] » is realized with a non-zero probability, it is necessary and sufficient that the events, in the sense of probabilities, « i i ∈ [ i , I will ] » are realized with a non-zero probability.

[0141] Noting that the mobile terminal's direction of movement densities i 1 and i 2 are independent of each other by construction, property (3) is then rewritten: ∀ θ , θ ′ ∈ 0 , 2 π 2 , P θ ∈ θ , θ ′ > 0 ⇔ P θ 1 > 0 ∩ θ 2 > 0 > 0

[0142] Considering the average direction density of the mobile terminal i verifying equation (3') as the result of two random experiments whose order does not matter, it is then possible to write, noting I i , i ' = [ i , I will ] for everything i , I will : P θ ∈ I θ , θ ′ = P θ 1 , θ 2 ∈ I θ , θ ′ 2 + P θ 1 ∉ I θ , θ ′ ∩ θ 2 ∈ I θ , θ ′ 1 P θ 1 ∈ I θ , θ ′ > 0 θ , θ ′ + P θ 1 ∈ I θ , θ ′ ∩ θ 2 ∉ I θ , θ ′ 1 P θ 2 ∈ I θ , θ ′ > 0 θ , θ ′ which can be rewritten, thanks to the independence of the two densities of direction of movement of the mobile terminal i 1 and i 2, in the form: P θ ∈ I θ , θ ′ = 1 − P θ 1 ∉ I θ , θ ′ P θ 2 ∉ I θ , θ ′ 1 P θ 1 ∈ I θ , θ ′ ∩ θ 2 ∈ I θ , θ ′ > 0 θ , θ ′

[0143] Such an expression is easily generalizable to n independent mobile terminal displacement direction densities where n corresponds to the number of pairs C i of events constituted for a given time window.

[0144] In some embodiments, equation (4) may then be normalized to verify the following property: P( i ∈ [0, 2π[) = 1, which reflects the fact that the average direction of movement of the mobile terminal sought is in the interval [0, 2π[.

[0145] In certain embodiments, from the average direction density of movement of the mobile terminal thus determined, it is possible to obtain, in a step S4, a value of an average direction of movement of the mobile terminal by calculating the expectation of a probability law associated with the average direction density of movement of the mobile terminal. i .

[0146] Since the direction densities of the mobile terminal are circular densities, the calculation of the expectation and the standard deviation should be adapted.

[0147] To do this, it is necessary to ask: m 1 = ∫ Γ P (θ) e iθ< dθ, where Γ is any interval of range 2π.

[0148] The average direction value of movement of the mobile terminal is then expressed as i = arg ( m 1).

[0149] For the calculation of the standard deviation, several methods are possible and most often use the modulus of m 1 as well as an analogy with a circular normal distribution. In an example, the following estimator of the standard deviation is given by: σ θ = arcsin ε 1 + 2 3 − 1 ε 3 , Or ϵ = 1 − Re m 1 2 + Im m 1 2 where Re(m1) and Im(m1) denote respectively the real part and the imaginary part of m1.

[0150] This standard deviation can then be used to determine a 95% confidence interval.

[0151] In this third implementation, a first angular sector is defined as extending between a first direction θ model max associated with one of the candidate segments associated with event E2 and a direction θ model min associated with another of the candidate segments associated with event E2, the directions of the other candidate road segments being between θ model max And θ model min .

[0152] A second angular sector is defined as extending between a first direction θ trans max associated with a segment connecting a first end of a candidate segment associated with event E1 and a first end of a candidate segment associated with event E2, and a direction θ trrans min associated with another segment connecting a second end of a candidate segment associated with event E1 and a second end of a candidate segment associated with event E2. The directions of the other segments connecting the ends of the two candidate segments associated respectively with event E1 and event E2 are between θ trans max And θ trans min .

[0153] Of course, the transition probability P ( s 2 | s1) can be determined as a function of at least two densities representative of a movement of the mobile terminal among the density of distribution of distances between the coverage areas of the first base station and the second base station f ( u ), the average speed density of movement of the mobile terminal f ( v ) and the average direction density of movement of the mobile terminal between the coverage areas of the first base station and the second base station f ( i ).

[0154] In order to take into account the behavior of the user of the mobile terminal, the transition probability can be weighted, in certain embodiments, by means of a penalty coefficient representative of the number of nodes of a transport network included in a candidate route. A node of a transport network is for example a crossroads, a metro station, a train station, etc. it is assumed that a user seeks to limit (for example minimize) changes of direction or mode of transport and therefore the number of nodes of the candidate route.

[0155] For example, the mathematical law linking the value of the penalty coefficient with the number of nodes is given by the following formula: y = λ e − λx where y corresponds to the value of the penalty coefficient, x to the number of nodes - 1 and l a constant. In the remainder of this document it is considered that l is worth 1. Of course, lmay take other values depending on the specific case.

[0156] At the end of step G4, a transition probability was calculated for all candidate road segments for a given pair of network events.

[0157] Steps G3 and G4 are for example implemented for all network events considered for determining the route taken by the mobile terminal.

[0158] Thus, once steps G3 and G4 have been implemented for all the network events considered for determining the route taken by the mobile terminal, a plurality of candidate routes are available for the mobile terminal.

[0159] In a step G5, the route taken by the mobile terminal is selected from the set of candidate routes taking into account the connection (or transmission) probabilities and the transition probabilities of the plurality of network events considered. For example, the selected route taken may be the one for which a product of the set of connection (or transmission) probabilities with the set of transition probabilities determined for the plurality of network events considered is the highest. For this, it is possible, for example, to apply a Viterbi algorithm which makes it possible to choose the most probable route from a set of candidate routes.

[0160] In some embodiments, a mobile terminal mobility detection algorithm may be used to obtain only data relating to mobile terminals that are actually moving. Indeed, determining the route of a mobile terminal that is not actually moving may, in some embodiments, be of little interest (or even no interest) and not taking into account stationary terminals may make it possible to limit the load in terms of memory resources and processing necessary for executing the method.

[0161] In such an embodiment, to estimate the effective mobility of a mobile terminal, it is possible, for example, to use the distance traveled, or the direction of a mobile terminal.

[0162] To do this, we can, for example, apply the same operations as explained previously with reference to the figure 4to a second pair of network events ER3 and ER4 belonging to the second set of network events JER2 in order to obtain a distance density f D 34 traveled for the pair of network events ER3, ER4. The network events ER3, ER4 occur within the same cell.

[0163] Thus, in certain embodiments, at the end of a step E2', a first distance density can be obtained f D 12 traveled for the pair of network events ER1, ER2 and a second a distance density f D 34 covered for the pair of network events ER3, ER4.

[0164] This second density of distance f D 34 traveled represents an uncertainty of movement of the mobile terminal inherent in the radio communication network which will make it possible to determine whether the first distance density f D12 traveled corresponds to an (effective) movement of the mobile terminal or corresponds to a mobile terminal which is stationary from the point of view of the radio communications network, i.e. a mobile terminal which has remained in the coverage area of the same base station.

[0165] In order to improve the accuracy of the value of the distance traveled by the mobile terminal, and / or the accuracy of the value of the uncertainty of movement of the mobile terminal inherent in the radio communication network, steps E1' and E2' (which correspond to steps E1 and E2 described with reference to the figure 4 ) may be repeated, in certain embodiments, for a plurality of pairs of network events belonging to the first set of network events JER1 and / or for a plurality of pairs of network events belonging to the second set of network events JER2 (on the condition, for example, that the instants youassociated with each of the events are included in the time window (15 minutes for example) defined above). Of course, the duration of the time window can take any other value depending on the methods of implementation (for example depending on needs, legislation, etc.)

[0166] Thus, once all the pairs of events ER i , ER j taken within the time window have been constituted for the two sets of network events JER1 and JER2, the steps E1' and E2' can be implemented for pairs of events ER i , ER j of one of these two sets of network events JER1 and JER2 (for each of these pairs of these two sets for example). At the end of these different iterations of steps E1' and E2,' it is possible to obtain, for the first set of network events JER1, as many densities of distances traveled by the mobile terminal as there are pairs of events ER i , ER j coming from this set of network events JER1 and for the second set of network events JER2, as many values of an uncertainty of movement of the mobile terminal inherent in the radio communication network as there are pairs of events ER i , ER j coming from this set of network events JER2.

[0167] In certain embodiments, a step E3' (which corresponds to step E3 and E2 described with reference to the figure 4 ) during which a combination between them of the different densities of distance traveled by the mobile terminal obtained for the pairs of network events belonging to the first set of network events JER1 can be implemented. Such a combination can for example result in an average density of distance traveled by the mobile terminal for a duration less than or equal to that of the time window considered.

[0168] Such embodiments may, for example, be based on an additional hypothesis. Thus, the existence of a law of distance traveled relating to the movement of the mobile terminal may be assumed. For example, it may be assumed that this law of distance traveled relating to the movement of the mobile terminal may be obtained from the different densities of distance traveled by the mobile terminal corresponding to the pairs of events ER i , ER j .

[0169] In the detailed embodiments, in order to obtain the average travel distance density of the mobile terminal over the duration of the time window considered, it is possible, for example, to consider two pairs of events belonging to the first set of network events JER1, a first pair of events C 1 consisting of events ER1 and ER2 and a second pair of events C 2 consisting of events ER5 and ER6, as well as the corresponding travel distance densities of the mobile terminal obtained at the end of the implementation of steps E1' and E2', and noted respectively D 1 and D 2 .

[0170] We denote by D the random variable representing the average travel distance density of the mobile terminal, which we seek to obtain from the travel distance densities of the mobile terminal D 1 and D 2 .

[0171] Knowing the following property: ∀ d , d ′ ∈ ℝ + 2 , P D ∈ d , d ′ > 0 ⇔ P D 1 ∈ d , d ′ > 0 et P D 2 ∈ d , d ′ > 0 which says that for the event, in the sense of probabilities, "D ∈ [d, d']" to occur with a non-zero probability, it is necessary and sufficient that the events, in the sense of probabilities, "D i ∈ [d, d']" occur with a non-zero probability.

[0172] Noting that the densities of distance traveled by the mobile terminal VD 1 and D 2 are independent of each other by construction, property (3) is then rewritten: ∀ d , d ′ ∈ ℝ + 2 , P D ∈ d , d ′ > 0 ⇔ P D 1 > 0 ∩ D 2 > 0 > 0

[0173] Considering the average traveled distance density of the mobile terminal D verifying equation (3') as the result of two random experiments whose order does not matter, it is then possible to write, noting I d,d' , = [d, d'] for all d, d' : D ∈ I d , d ′ = P D 1 , D 2 ∈ I d , d ′ 2 + P D 1 ∉ I d , d ′ ∩ D 2 ∈ I d , d ′ 1 P D 1 ∈ I d , d ′ > 0 d , d ′ + P D 1 ∈ I d , d ′ ∩ D 2 ∉ I d , d ′ 1 P D 2 ∈ I d , d ′ > 0 d , d ′ which can be rewritten, thanks to the independence of the two densities of distance traveled by the mobile terminal D 1 and D 2 , in the form: P D ∈ I d , d ′ = 1 − P D 1 ∉ I d , d ′ P D 2 ∉ I d , d ′ 1 P D 1 ∈ I d , d ′ ∩ D 2 ∈ I d , d ′ > 0 d , d ′

[0174] Such an expression is easily generalizable to n independent densities of distance traveled by the mobile terminal where n corresponds to the number of pairs C i of events constituted for a given time window.

[0175] Equation (4) can then be normalized to verify the following property: P V ∈ ℝ + = 1 , which reflects the fact that the average distance traveled by the mobile terminal sought is positive or zero.

[0176] The same reasoning can be applied in order to obtain a combination between them of the different values of an uncertainty of movement of the mobile terminal inherent in the radio communication network obtained for the pairs of network events belonging to the second set of network events JER2.

[0177] As previously, we consider two pairs of events belonging to the second set of network events JER2, a first pair of events C 3 consisting of events ER3 and ER4 and a second pair of events C 4 consisting of events ER7 and ER8, as well as the corresponding densities of distance traveled for the movement of the mobile terminal obtained at the end of the implementation of steps E1' and E2', and noted respectively D 3 and D 4 . We then obtain a random variable D' representing an uncertainty of movement of the mobile terminal inherent in the average radio communication network.

[0178] In certain embodiments, it is possible, for example, to compare the first density of distance traveled by the mobile terminal, obtained from pairs of events belonging to the first set of network events JER1, with this noise information, obtained from pairs of events belonging to the second set of network events JER2, in order to determine whether the mobile terminal has actually moved, i.e. has changed base station of attachment following an (effective) movement or whether the mobile terminal has not left the coverage area of a base station and is therefore immobile from the point of view of the radio communications network.

[0179] In some embodiments, to compare the first traveled distance density and the second traveled distance density, a Hellinger distance is determined, for example, during a step E4. Such a Hellinger distance may be constructed from a Bhattacharyya coefficient.

[0180] As explained in more detail below, the value of the Hellinger distance thus constructed can help determine whether the mobile terminal has actually moved. When the Hellinger distance has a high value, i.e., when it is close to 1, the mobile terminal is considered to have actually moved. When the Hellinger distance has a low value, i.e., when it is close to 0, the mobile terminal is considered to be in a state of immobility. Thus, for continuous probability densities p and q, the Bhattacharyya coefficient is defined as: BC p q = ∫ p x q x dx .

[0181] Such a Bhattacharyya coefficient lies between 0, the case where the continuous probability densities p and q do not overlap, and 1, the case where the continuous probability densities p and q are equal.

[0182] The Hellinger distance is then defined as: d H p q = 1 − BC p q .

[0183] This distance is also between 0, the case where the continuous probability densities p and q do not overlap, and 1, the case where the continuous probability densities p and q are equal.

[0184] THE [ FIG. 7A ] and [ FIG. 7B ] respectively represent a situation in which the mobile terminal is stationary from the point of view of the radio communications network and a situation in which the mobile terminal is mobile from the point of view of the radio communications network.

[0185] In fact, on the figure 7A ,we see that the two densities of distance traveled almost completely overlap, this corresponding to a Hellinger distance close to 0. In figure 4B, on the contrary, we see that the two densities of distance traveled do not overlap, this corresponding to a Hellinger distance close to 1.

[0186] In some embodiments, to mitigate the network oscillation phenomenon, the instants you And tjcorresponding respectively to a network event ER i and to an event ER j constituting a pair C i of network events can be chosen to be separated by a minimum duration. Thus, the duration elapsed between the occurrence of the event ER i and the event ER j is greater than or equal to a first duration. Such a duration, acting as a threshold, can for example when the time window considered lasts 15 minutes, be set in the order of a few minutes (for example from 3 to 9 minutes), such as a duration of 6 minutes. Of course other values of this first duration can be envisaged.

[0187] There [ fig.8 ] represents a device 10 capable of implementing certain steps of the solution previously described.

[0188] A device 10 may comprise at least one hardware processor 1001, a storage unit 1002, an interface 1003, which are connected to each other via a bus 1004. Of course, the constituent elements of the device 10 may be connected by means of a connection other than a bus.

[0189] The processor 1001 controls the operations of the device 10. The storage unit 1002 stores at least one program for implementing the method that is the subject of the invention to be executed by the processor 1001, and various data, such as parameters used for calculations performed by the processor 1001, intermediate data of calculations performed by the processor 1001, etc. The processor 1001 may be formed by any known and suitable hardware or software, or by a combination of hardware and software. For example, the processor 1001 may be formed by dedicated hardware such as a processing circuit, or by a programmable processing unit such as a central processing unit (Central Processing Unit) that executes a program stored in a memory thereof.

[0190] The storage unit 1002 may be formed by any suitable means capable of storing the program(s) and data in a computer-readable manner. Examples of the storage unit 1002 include computer-readable non-transitory storage media such as semiconductor memory devices, and magnetic, optical, or magneto-optical recording media loaded into a read-write unit.

[0191] Interface 1003 provides an interface between device 10 and other equipment in the radio communications network.

Claims

1. Method for determining a route of a mobile terminal on the basis of data relating to a plurality of network events involving said mobile terminal, said route consisting of at least one route segment of a transport network, said method comprising: - determining at least one candidate route, during which the following are determined for at least one first network event from among the plurality of network events: a probability of said mobile terminal connecting to a first base station involved in the first network event, in the knowledge that said mobile terminal is located, at a time t1 at which the first network event occurred, on a first candidate route segment located within a zone of coverage of said first base station, and a probability of said terminal transitioning, at a time t2 at which a second network event involving a second base station occurred, to at least one second candidate route segment located within a zone of coverage of said second base station, in the knowledge that the mobile terminal is located on the first candidate route segment at the time t1; - selecting said route of said mobile terminal from among the set of candidate routes, taking into account the probabilities of connection and the probabilities of transition determined for the plurality of network events.

2. Device for determining a route of a mobile terminal on the basis of data relating to a plurality of network events involving said mobile terminal, said route consisting of at least one route segment of a transport network, said device comprising a processor configured for the purpose of: - determining at least one candidate route by determining the following for at least one first network event from among the plurality of network events: a probability of said mobile terminal connecting to a first base station involved in the first network event, in the knowledge that said mobile terminal is located, at a time t1 at which the first network event occurred, on a first candidate route segment located within a zone of coverage of said first base station, referred to as probability of transmission, and a probability of said terminal transitioning, at a time t2 at which a second network event involving a second base station occurred, to at least one second candidate route segment located within a zone of coverage of said second base station, in the knowledge that the mobile terminal is located on the first candidate route segment at the time t1; - selecting said taken route from among the set of candidate routes, taking into account the probabilities of connection and the probabilities of transition determined for the plurality of network events.

3. Method for determining a route according to Claim 1, or device for determining a route according to Claim 2, where said selected route is the one for which a product of the set of probabilities of connection and the set of probabilities of transition determined for the plurality of network events is the highest.

4. Method for determining a route according to Claim 1 or Claim 3, or device for determining a route according to Claim 2 or Claim 3, where the probability of connection P(A | s1) for a given network event is determined using the following formula: P A s 1 = ∑ j = 1 n P A p j n wherein A represents the first base station involved in the first event, s1 represents the first candidate route segment located within the zone of coverage of said first base station, pj represents a pixel of a likelihood map representing a probability of said mobile terminal connecting to the first base station when the terminal is located on a pixel of said likelihood map, and n represents the number of pixels of said likelihood map of non-zero weight passed through by the first candidate route segment s1.

5. Method for determining a route according to Claim 4, comprising, or device for determining a route according to Claim 4, where said processor is configured for the purpose of, selecting said at least one first candidate route segment taking into account a distance between said at least one first candidate route segment and at least one candidate route segment at a time t0 prior to the time t1.

6. Method or device for determining a route according to Claim 5, where said at least one first candidate route segment is selected from among route segments located at a distance less than or equal to a first distance, referred to as connectivity tolerance distance, of the set of candidate route segments at a time t0 prior to the time t1.

7. Method or device for determining a route according to Claim 5 or 6, where the value of the first connectivity tolerance distance takes into account a density of the geographical distribution of the base stations and a median radius of a zone of coverage of base stations in the proximity of said first base station.

8. Method for determining a route according to any one of Claims 1 and 3 to 7, or device for determining a route according to any one of Claims 2 to 7, where the probability of connection is weighted as a function of a deviation of a direction of the first candidate route segment from a reference direction.

9. Method for determining a route according to any one of Claims 1 and 3 to 8, or device for determining a route according to any one of Claims 2 to 8, where the probability of transition P(s2 | s1) for a given network event is determined using the following formula: P s 2 s 1 = ∫ d min d max f u du wherein: s1 represents the first candidate route segment located within the zone of coverage of said first base station, s2 represents the second candidate route segment located within the zone of coverage of said second base station, f(u) represents a density of the distribution of distances between the zones of coverage of the first base station and of the second base station, dmin represents the minimum distance between the first candidate route segment and the second candidate route segment on a graph representing said transport network and dmax represents the sum of dmin with the length of the first candidate route segment and the length of the second candidate route segment.

10. Method or device for determining a route according to Claim 9, where the probability of transition is determined taking into account a density of the distribution of distances between the zones of coverage of the first base station and of the second base station.

11. Method for determining a route according to Claim 1 or any one of Claims 3 to 6, or device for determining a route according to any one of Claims 2 to 6, where the probability of transition P(s2 | s1) for a given network event is determined using the following formula: P s 2 s 1 = ∫ v min v max f v dv wherein s1 represents the first candidate route segment located within the zone of coverage of said first base station, s2 represents the second candidate route segment located within the zone of coverage of said second base station, f(v) represents an average speed of movement density of the mobile terminal, νmin = dmin / Δt where dmin represents the minimum distance between the first candidate route segment and the second candidate route segment on a graph representing said transport network and νmax = dmax / Δt where dmax represents the sum of dmin with the length of the first candidate route segment and the length of the second candidate route segment and where Δt = t1-t2.

12. Method for determining a route according to Claim 1 or any one of Claims 3 to 6, or device for determining a route according to any one of Claims 2 to 6, where the probability of transition P(s2 | s1) for a given network event is determined using the following formula: P s 2 s 1 = ∫ θ min θ max f θ dθ wherein s1 represents the first candidate route segment located within the zone of coverage of said first base station, s2 represents the second candidate route segment located within the zone of coverage of said second base station, f(θ) represents an average direction of movement density of the mobile terminal between the zones of coverage of the first base station and of the second base station, θmin represents the lower bound of an intersection of a first angular sector defining possible directions of movement for the mobile terminal with a second angular sector defining a maximum angle between a first end of the first candidate route segment and a second end of the second candidate route segment and θmax represents the upper bound of the intersection of the first angular sector and of the second angular sector.

13. Method or device for determining a route according to either one of Claims 7 and 8, where the probability of transition is determined as a function of at least two densities representative of a movement of the mobile terminal from among the density of the distribution of distances between the zones of coverage of the first station and of the second base station f(u), the average speed of movement density of the mobile terminal f(v) and the average speed of movement density of the mobile terminal between the zones of coverage of the first base station and of the second base station f(θ).

14. Method or device for determining a route according to any one of Claims 7 to 9, where the probability of transition is weighted as a function of the number of intersections of route segments of the transport network that are encountered along the candidate route.

15. Method or device for determining a route according to any one of Claims 7 to 9, where the probability of transition is weighted by way of a penalty coefficient representative of a number of intersections of route segments of the transport network that are encountered along the candidate route, the penalty coefficient tending towards zero as the number of intersections increases.