Method and system for geographic location of terminals deployed in groups - Patents.com

Abstract

The invention relates to a method for estimating the geographical position of a terminal (70i) of interest among a panel of terminals (70) of a wireless communication system (60), comprising a step (52) of determining, for each terminal of said panel, a temporal signature comprising values ​​measured for said terminal during a predetermined observation period, a step (54) of calculating a similarity value between the temporal signatures of the terminals of the panel, the similarity value calculated for two terminals representing the probability that said terminals have moved together or been co-located during said observation period, a step (56) of dividing the various terminals of the panel into different groups according to their similarity values, and a step (44) of estimating the geographical position of the terminal of interest as a function of available geolocation data for at least one other terminal of the group to which the terminal of interest belongs. [Selected figure] Figure 3

Description

[Technical field]

[0001] The present invention relates to the field of geolocation. In particular, the present invention relates to a method and a system for geolocating terminals of a wireless communication system. The present invention can be particularly well applied to the geolocation of smart objects of the "Internet of Things" (or IoT) type, which have a group-moving nature. [Background technology]

[0002] Over the past few years, the increased use of wireless communication systems has, naturally, led to the development of services based on the geographic location of objects and used, for example, to assist navigation, manage traffic, track goods deliveries, and the like.

[0003] Satellite positioning systems, such as the Global Positioning System (GPS), are one of the best known geolocation technologies. These systems are based on the use at the receiving end of radio signals transmitted from dedicated satellites. Although satellite geolocation is particularly precise, it has several drawbacks, including the cost and power consumption incurred by incorporating a GPS receiver into the object, as well as the lack of performance in closed zones.

[0004] Other techniques exist for geolocating a terminal on the basis of signals exchanged with base stations of the access network to which the terminal is connected. In cellular networks such as GSM ("Wide area digital system for mobile communications"), it is known to estimate the location of a terminal as being the location of the base station with which the terminal is currently associated. This method has mediocre geolocation accuracy, since the coverage zone of a base station can reach a radius of several kilometers, or even tens of kilometers.

[0005] Other methods involve estimating the distance separating a terminal from several base stations by calculating the time of arrival, the propagation angle or the frequency difference for the signals exchanged between the terminal and the base stations. However, all these various methods have the disadvantage of requiring specific hardware and software. In fact, they generally require synchronization of the terminal and / or of the various base stations acting as observation points. Furthermore, they are particularly sensitive to the phenomenon called multipath (the propagation of the same radio signal over several paths due to reflection, refraction and diffraction phenomena on the barriers encountered).

[0006] Other geolocation methods are based on the received signal strength indicator (RSSI) levels of the signals exchanged between the terminal and the base station. These methods are particularly well suited for wireless communication systems of the cellular network type (for example GSM), since the RSSI information is directly available to the communication system, since it is used by the communication system itself. These methods are based on the fact that radio signals are attenuated in the atmosphere, and therefore the RSSI level of the signal received by the receiver varies depending on the distance separating it from the signal transmitter. It is therefore possible to determine the geolocation of the terminal by trilateration, by estimating the distances separating it from the various base stations surrounding it, on the basis of the RSSI levels measured by the base stations. A drawback of such geolocation methods by trilateration, based on RSSI levels, is their lack of accuracy, due to the fact that the many parameters that affect the attenuation of the signal (obstacles, radio interference, terminal movement, etc.) make the function for determining the distance on the basis of the RSSI levels very complicated.

[0007] Therefore, new geographic location methods based on RSSI levels have been developed. These new methods are based on machine learning techniques. In particular, it involves creating a database that associates, during a calibration phase, radio signatures corresponding to all of the RSSI levels measured for a terminal at a known location relative to a set of base stations of the system with known geographic locations. Then, during a search phase, the radio signature observed for a terminal at an unknown location is compared with all the signatures in the database and the location of the terminal is estimated based on the location corresponding to the most similar signature.

[0008] To carry out the calibration phase, it is known to place devices adapted to provide precisely the geographical positions and RSSI levels at different locations to the base stations of the communication system on board a fleet of vehicles moving across the geographical zone to be covered (this phase is called "war-driving"). The higher the number of points, the better the performance of the geolocation method in terms of accuracy, but the longer the calibration phase and the higher the costs.

[0009] Such a method for geolocation by machine learning based on radio signatures has several disadvantages, especially when the covered geographical zone is vast, for example when an entire country or even a continent has to be covered. The first disadvantage concerns the geolocation accuracy. In fact, it happens that two radio signatures corresponding to two adjacent geographical locations are significantly different, or that two particularly similar radio signatures correspond to two distant geographical locations. Establishing a relationship between a radio signature and the associated geographical location is therefore made difficult by the noise generated by such a situation. Another disadvantage concerns the complexity of the machine learning algorithm used. In fact, in order to obtain a sufficient accuracy, it is necessary to insert a large number of elements into the database created during the calibration phase. Moreover, when the covered zone is vast, there is a large number of base stations to be considered. All this contributes to a significant increase in the data at the input of the machine learning algorithm, thus generating capacity and computation time constraints. Summary of the Invention

[0010] The object of the present invention is to overcome all or some of the disadvantages of the prior art, in particular those disclosed above, by proposing a method for geolocating terminals of a wireless communication system, the performance of which is improved in terms of accuracy and complexity by being based on the idea that the terminals of the system may move in groups.

[0011] Conventional geolocation methods, which use the radio signature of a terminal to estimate its geographical location, actually concentrate only on information about the relationship existing between a terminal and an access network of a communication system, such as the RSSI level or the propagation time of the signals exchanged between the terminal and the access network.

[0012] The geographical location method proposed by the invention differs from the prior art in that it additionally uses information coming from the terminals of the system making it possible to determine, for example, whether, during a particular observation period, the terminal whose geographical location is sought to be estimated has moved as a group together with other terminals, in which case it is possible to estimate the geographical location of said terminal, for example by using the information in comparison with other terminals of the group.

[0013] The invention is particularly advantageous for use in the field of goods transport, although it is in no way exclusive thereto. For example, pallets used for goods transport can be equipped with terminals of a wireless communication system to allow geographical location of the terminals. Such pallets have in fact the particular property of moving in groups.

[0014] Thus, according to a first aspect, the invention proposes a method for estimating the geographical position of a terminal, called "target terminal", from a panel of the terminal of a wireless communication system. Said method comprises: determining, for each terminal of said panel, a time signature comprising a value representative of a mobility stage of said terminal, measured independently by said terminal by at least one sensor of said terminal during a predetermined observation period; - calculating the value of a similarity measure between the time signatures of the terminals of the panel, the value of the similarity measure calculated for two terminals representing the probability that the terminals have moved together or been co-located during the observation period; - dividing the different terminals of the panel into different groups depending on the value of a similarity criterion; Estimating the geographical location of the target terminal as a function of geolocation data available to at least one other terminal of a group to which the target terminal belongs, called the "target group".

[0015] The panel of terminals can be determined, for example, by a selection of terminals of the system, in which it is known a priori that the terminal is close to the target terminal (for example, if the terminal is covered by at least one base station of the wireless communication system that also covers the target terminal), or that the terminal has certain properties that make it belong to a group of several terminals that have similarities.

[0016] A terminal's time signature corresponds to a set of values ​​that are independently measured by the terminal using its sensors, such as motion sensors.

[0017] "Independently" means that the measurements representative of the mobility phase of the terminal performed by the terminal to form the time signature do not depend on the behavior of another device. In particular, RSSI measurements are not measurements performed independently by the terminal, since such measurements depend on the transmission of signals by other devices, such as base stations. According to another embodiment, the geographical position measured on the basis of radio signals transmitted by one or more satellites of a satellite positioning system is also not a measurement performed independently by the terminal.

[0018] If the two time signatures for the two devices have significant similarities, it means that the two devices in question may have moved together or remained in the same location during the observation period.

[0019] The division of the various terminals of the panel into various groups makes it possible to improve the geographical location of the target terminals whose geographical location is sought to be estimated. In fact, it is possible to estimate the geographical location of said target terminals, for example if one of the terminals of the group is equipped with a GPS receiver, for example if the geographical location of at least one of the terminals of the group is precisely known. According to another embodiment, it is possible to estimate the geographical location of said terminals by correlating the estimated locations of several terminals of the group. In certain cases, it is possible to estimate the geographical location of said target terminals by correlating geographical location data (such as GPS locations and / or estimated locations) for several terminals of several different groups. It should be noted that this geographical location data does not directly correspond to a geographical location, but may be information that allows the geographical location to be determined or estimated (for example the propagation time or RSSI level of the signal, used by trilateration).

[0020] In particular embodiments, the invention may further comprise one or more of the following features, taken alone or in all technically possible combinations:

[0021] In a particular embodiment, the values ​​measured to determine the time signature of the terminal are: values ​​measured by at least one sensor of said terminal during the observation period, and / or A radio signature includes values ​​representative of the quality of radio links existing between a terminal and one or more base stations of the wireless communication system at various times during an observation period.

[0022] In a particular embodiment, the value representative of the mobility stage of the terminal, measured by at least one sensor of said terminal, is: Values ​​measured by a motion sensor, and / or the value measured by the temperature sensor, and / or The value measured by the atmospheric pressure sensor, and / or the value measured by a light intensity sensor, and / or The value measured by the magnetic field sensor.

[0023] Indeed, depending on the type of use considered, various environmental data can be used to determine whether the terminals moved together during the observation period. For example, if the environmental conditions observed for two separate terminals are similar, this means that said terminals are likely to be "co-located" during the observation period.

[0024] In a particular embodiment, the terminal's time signature further includes a radio signature including a value representative of the quality of a radio link existing between the terminal and one or more base stations of the wireless communication system at various times during the observation period, and the similarity criterion is determined according to a similarity criterion for the mobility phase and a similarity criterion for the radio signature.

[0025] For example, the time signature may include a value representative of the mobility phase of the terminal during the observation period, determined for example using a motion sensor such as an accelerometer or a gyroscope. In addition, the time signature may include the radio signature of the terminal determined at different times during the observation period. If the time signatures of two terminals are similar, in other words the values ​​and / or radio signatures forming them satisfy certain conditions of similarity, then the two terminals can probably be considered to belong to the same group of terminals that moved together during the observation period.

[0026] In a particular embodiment, the similarity criterion is determined depending on a weighting factor assigned to the measured values, the weighting factor representing the time elapsed between the time when said values ​​were measured and the time when the geographical position of the target terminal was estimated.

[0027] Such an arrangement makes it possible to give greater importance to recent data of the time signatures, indeed, if for example two terminals move together only during part of the observation period, more importance must be given to recent data in order to decide whether at the end of the observation period, i.e. when the estimation of the geographical location of the target terminals is performed, the terminals in question are together or not, i.e. whether they should be considered to belong to the same group.

[0028] In a particular embodiment, the division of the various terminals of the panel is performed such that each terminal belongs to at most one group.

[0029] Such a division is performed, for example, by a method for grouping elements together known as "hard clustering", in which each element either belongs to a single group or does not belong to any group (disjoint element).

[0030] In a particular embodiment, the division of the various terminals of the panel is performed such that each terminal belongs to each group depending on the value of the probability of belonging, the target group corresponding to the group to which the value of the probability of belonging of the target terminal is the largest.

[0031] Such a division can be performed, for example, by a more difficult method for grouping elements known as "soft clustering", in which each element belongs to several groups to a certain extent defined by probabilistic methods.

[0032] In a particular embodiment, the geographic location of the target terminal is estimated as a function of the geolocation data available for the various terminals belonging to the various groups and as a function of the probability values ​​that the target terminal belongs to the various groups.

[0033] In a particular embodiment, the geolocation data is provided by the terminal in the form of a geographic position obtained by a satellite positioning system of the terminal.

[0034] If a particular terminal of the system is equipped with a GPS receiver, it is possible to determine the geographical position of the target terminal using a GPS position provided by one of the terminals of the target group to which the target terminal belongs, or using several GPS positions provided by several terminals of the target group, or even using several GPS positions provided by several terminals of several different groups.

[0035] In a particular embodiment, the geolocation data is available to the terminal in the form of a radio signature that includes a value representative of the quality of radio links that exist between the terminal and one or more base stations of the wireless communication system.

[0036] In a particular embodiment, the geographic location of a terminal of the panel is estimated as a function of the terminal's radio signature and a reference set including radio signatures associated with known geographic locations. Thus, the geographic location of the target terminal is estimated as a function of the estimated geographic location of at least one other terminal of the target group.

[0037] This reference set is for example a database storing reference elements used by a machine learning algorithm. Each reference element is thus a pair of information consisting on the one hand of a radio signature and on the other hand of a known geographical location associated with this radio signature. The learning algorithm makes it possible to estimate the location of the target terminal on the basis of said radio signature of the target terminal and using the reference database. Thus, to reduce the complexity of the geographic location method, it may be sufficient to estimate the geographical location of a single terminal of a given group and to consider all terminals of the group to have the same geographical location.

[0038] According to another embodiment, in order to improve the accuracy of the geographic location method, it may be advantageous to correlate the estimated geographic positions of several terminals of the target group in order to refine the geographic position of each terminal of the group, in particular the geographic position of the target terminal.

[0039] In a particular embodiment, a virtual radio signature is determined for a group of interest as a function of radio signatures of terminals belonging to said group of interest, and the location of the terminal of interest is estimated as a function of this virtual radio signature.

[0040] In this case, this embodiment involves, for example, correlating the radio signatures of several terminals of the target group to determine a virtual radio signature that is used to estimate the geographical positions belonging to all terminals of the group, in particular the target terminal.

[0041] In a particular embodiment, the panel of terminals is determined by a step of selecting from among the terminals of the wireless communication system, the determination being made such that the terminals of the panel are covered by at least one base station of the wireless communication system which also covers the target terminal.

[0042] According to a second aspect, the invention relates to a wireless communication system comprising a terminal and an access network comprising a base station and a server connected to said base station. Said system implements a method for geolocation of a target terminal from a panel of said terminal according to any one of the preceding embodiments. Said server comprises: collecting values ​​representative of mobility stages of the terminals of the panel, measured independently by the terminals via at least one sensor of the terminals during a predetermined observation period; determining a time signature for the terminal based on the measured values; Calculating the value of a similarity metric between the time signatures of the terminals of the panel such that the value of the similarity metric calculated for two terminals represents the probability that the terminals moved together during the observation period; Dividing the various terminals of the panel into various groups according to the value of the similarity criterion, The target terminal is adapted to estimate the geographical location of the target terminal as a function of geolocation data available for at least one other terminal of a group to which the target terminal belongs, referred to as the "target group".

[0043] It is believed that the invention will be better understood on reading the following description, which is given by way of non-limiting example and is made with reference to figures 1 to 7. [Brief description of the drawings]

[0044] [Figure 1] 1 is a simplified diagram of a wireless communication system. [Diagram 2] This is a key step in traditional geolocation methods. [Diagram 3] 5-7 are the main steps of the geolocation method according to the invention; [Figure 4] 4 is a schematic diagram of mobility stages determined by a terminal; [Diagram 5] 1 is a schematic diagram of the atmospheric pressure measured by the terminal. [Figure 6] 1 is a schematic diagram of temperatures measured by the terminal. [Figure 7] 4 is a schematic illustration of a radio signature determined for a terminal; DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0045] In the drawings, the same reference numbers indicate the same or similar elements throughout the drawings, and for clarity, elements shown are not necessarily drawn to scale unless otherwise noted.

[0046] The following description describes, by way of example and not limitation, some embodiments of the present invention.

[0047] 1. Situation In the remainder of this document, we consider, without limitation, the case in which smart object type terminals of a wireless communication system are fixed to pallets of goods with the purpose of making it possible to determine their geographical location. Such pallets generally move in groups according to the destination of the goods they contain.

[0048] FIG. 1 shows diagrammatically a wireless communication system 60 comprising a terminal 70 and an access network 80 comprising, for example, a number of base stations 81 connected to a server 82 .

[0049] Terminal 70 and base station 81 exchange data in the form of radio signals. By "radio signals" we mean electromagnetic waves that propagate in free space and whose frequencies are within the spectrum of ordinary radio waves (from a few Hertz to a few hundred Gigahertz).

[0050] The data exchange between the terminal 70 and the base station 81 is for example bidirectional: in other words the terminal 70 is adapted to transmit data to the base station 81 via an upstream link and to receive data from this base station 81 to this terminal 70 via a downstream link.

[0051] However, according to other embodiments, it is not excluded to have a unidirectional exchange only via the upstream link. Many IoT type uses involve the collection of data transmitted by the terminals 70 and are entirely addressed by an exchange only via the upstream link between each terminal 70 and the base station 81.

[0052] According to a non-limiting example, the wireless communication system 60 is ultra-narrowband. "Ultra-narrowband" (UNB) means that the instantaneous frequency spectrum of the signals transmitted by the terminal 70 has a frequency width of less than 2 kilohertz, or even less than 1 kilohertz. Such an arrangement is particularly advantageous in that the transmission of such signals can be performed with a very reduced power consumption, which is particularly adapted for IoT type uses.

[0053] Each base station 81 is adapted to receive messages transmitted by terminals 70 within its range. Each message thus received is communicated, for example to a server 82 of the access network 80, optionally accompanied by other information, for example the identifier of the base station 81 which received the message, a value representative of the quality of the radio signal carrying the message, the center frequency at which the message was received, the date at which the message was received, etc. The server 82 for example processes all the messages received from the various base stations 81.

[0054] To estimate the geographical location of a terminal 70, called "target terminal 70i", it is known to use geographical location methods based, for example, on machine learning algorithms that seek to establish a relationship between the radio signature of the terminal and the geographical location of this terminal. The server 82 may in particular be used to implement such a geographical location method.

[0055] In the remainder of this specification, as a non-limiting example, the case is considered where the invention improves such existing geolocation methods that are based on machine learning algorithms that aim to establish a relationship between the radio signature of a terminal and its geographic location.

[0056] However, it should be noted that there are other methods for estimating the location of the target terminal 70i, in particular methods that are not necessarily based on machine learning algorithms, such as trilateration methods that allow to determine the geographical position on the basis of calculating the Time Difference of Arrival (TDOA) of the signals, and the method according to the invention can also be applied to such methods.

[0057] In the remainder of this specification, a "geographic location" refers to a set of two coordinates corresponding to latitude and longitude, respectively. It should be noted that alternative forms for defining a geographic location are possible. For example, a third coordinate corresponding to altitude relative to mean sea level could be considered.

[0058] The radio signature of the terminal 70 is determined, for example, during the transmission of a message from this terminal 70 to the access network 80. For example, the radio signature comprises a set of values ​​corresponding to the levels of received average power (RSSI) (for example in decibels) measured by the various base stations at the time of receiving said message. For example, for a base station 81 from which the message was not received because the terminal 70 is too far away and not in its radio coverage, a default value of -160 dB is used, for example.

[0059] It should be noted that other methods for determining the radio signature are possible. For example, other values ​​representative of the quality of the radio link existing between the terminal 70 and the base station 81 during the transmission of the message, such as the Signal-to-Noise Ratio (SNR) or the Channel Quality Indicator (CQI), can be used to determine the radio signature. According to other embodiments, the radio signature may include the propagation time of the signals carrying the messages exchanged between the terminal 70 and the base station 81. The choice of a particular method for determining the radio signature is merely an alternative embodiment of the invention.

[0060] To estimate the geographical position of the target terminal 70i, the geographical location method uses a set of reference elements, each of which comprises a radio signature associated with a geographical position. To build the set of reference elements, it is possible to use a particular terminal 70 of the wireless communication system 60, for example called "calibration terminal 70c", which has a location system, for example a GPS receiver, that allows to obtain the geographical position of the terminal 70c with high accuracy.

[0061] Typically, this reference set is, for example, a database stored in the server 82 and used by the machine learning algorithm. Each reference element stored in this database is a pair of information consisting, on the one hand, of a radio signature and, on the other hand, of a known geographical location associated with this radio signature. The radio signature and the known geographical location of the reference element are obtained on the basis of the calibration terminal 70c. The machine learning algorithm thus makes it possible to estimate the location of the target terminal 70i on the basis of the radio signature of said target terminal 70i and using the database.

[0062] FIG. 2 illustrates the main steps of such a geolocation method 10.

[0063] The purpose of the first calibration stage 20 is to generate a kind of radio map of the geographical zone to be covered. This stage: determining 22 radio signatures respectively associated with geographic locations; and storing 24 the obtained reference elements (related to the radio signature and the geographic location) in a set 26 of reference elements.

[0064] The step 22 of determining the radio signature involves, for example, sending, by the calibration terminal 70c, a message including its current geographical location to the access network 80 of the wireless communication system 60.

[0065] It should be noted that this message may be sent independently of the geo-location method 10. For example, the message may be a regular message for remotely reading information including the current geographic location and may not have the primary goal of participating in the calibration phase 20 of the geo-location method 10.

[0066] Thus, a base station 81 of an access network 80 receiving a signal containing the aforementioned message performs a measurement of the RSSI level at which the message was received and communicates the message and additional information (RSSI level, date of receipt of the message, base station identifier, etc.) to a server 82.

[0067] The server 82 collects the RSSI levels measured by the various base stations 81 and includes them in the radio signature thus determined for the calibration terminal 70c.

[0068] It should be noted that for the selection of base stations to take into account for the formation of the radio signature several methods are possible: According to a first embodiment, all base stations of the wireless communication system 60 are taken into account; According to other embodiments, the number of base stations to be taken into account can be limited, for example according to a limited geographical zone of interest.

[0069] The server 82 also performs the extraction of the current geographical position contained in the message sent by the calibration terminal 70c and communicated to it by the base station 81 which received it.

[0070] Finally, the pair of information formed by the geographical location of the calibration terminal 70c and the associated radio signature is added to the set of reference elements 26 by a storage step 24. The server 82 comprises, for example, a set of hardware and software means for storing the reference elements in a database.

[0071] In a particular embodiment, the steps of the calibration phase 20 described above are repeated for a predetermined duration, for example days, or even weeks or months, for the calibration terminal 70c of the wireless communication system 60, in order to obtain a set 26 of reference elements that includes a significant amount of information, i.e. a precise map of the covered zone. In one alternative, the steps of the calibration phase 20 are repeated until a predetermined number of reference elements are obtained in the set 26.

[0072] In the preferred embodiment, the reference data set 26 is continually enriched with new reference elements coming from the calibration terminal 70c.

[0073] Then, the search step 40: determining 42 a radio signature for a target terminal 70i located at an estimated geographic location; and a step 44 of estimating the geographical location of said target terminal 70 i on the basis of the determined radio signature and the set of reference elements 26 acquired during the calibration phase 20 .

[0074] The step 42 of determining the radio signature involves, for example, the transmission by the target terminal 70i of a message to the access network 80 of the wireless communication system 60. This message can be any message transmitted independently of the method 10 of geolocating a terminal. For example, this message can be transmitted for the purpose of remote reading of information unrelated to the geolocating of the terminal 70. Alternatively, this message can be transmitted purposefully with the goal of geolocating the terminal 70. In all cases, the content of the message is not necessarily important for the search stage 40.

[0075] Similar to the one performed for the calibration phase 20, a base station 81 of an access network 80 receiving the aforementioned message performs a measurement of the RSSI level of the signal carrying the message.

[0076] The server 82 therefore collects the RSSI levels measured by the various base stations 81 and includes them in the radio signature thus determined for the target terminal 70i.

[0077] Thus, the estimation 44 of the geographical location of the target terminal 70i which sent the message is carried out, for example, using a regression machine learning algorithm, on the one hand based on the set of reference elements 26 and, on the other hand, based on the radio signature determined for the target terminal 70i whose geographical location has to be estimated.

[0078] However, the above-described geographic location method 10 has mediocre performance in terms of geographic location accuracy, and the method is particularly complex (i.e., requires extremely large capacity and computation time), especially when the geographic zone to be covered is large, e.g., when an entire country or even a continent has to be covered.

[0079] 2. Principle of the present invention In order to improve the accuracy and / or simplify the complexity of the geolocation method 10, the invention is based on the idea that the terminals 70 of the considered system may move in groups. Thus, the invention proposes a method 10 for geolocating a target terminal 70i, which makes use of information provided by other terminals 70 that have moved together with the target terminal 70i during a certain period of time.

[0080] Figure 3 shows the main steps of a particular embodiment of a geolocation method 10 according to the invention. It should be noted that the order of the various steps shown in Figure 3 is given for informational purposes only, for clarity of the drawing, and should not be considered limiting for the invention.

[0081] For the geolocation method 10 in question, and now described by way of example, the calibration stage 20 is similar to that described above with reference to Fig. 2. However, the search stage 40 includes an additional step intended to form groups of terminals 70, each group including terminals 70 that are considered to have moved together during a certain current observation period up to the time at which the position of the target terminal 70i was estimated.

[0082] The search stage 40 includes a step 42 of determining a radio signature for a target terminal 70i whose geographical location it is desired to estimate. This step is similar to the step described above with reference to FIG.

[0083] The search stage 40 also includes a step 52 of determining, for each terminal 70 of the panel of terminals 70 of the wireless communication system 60, a time signature. The time signature of a terminal 70 comprises values ​​independently measured by the terminal 70 by at least one sensor of the terminal 70 during a past observation period, the duration of which is predetermined depending on the use considered. These values ​​correspond to data relating to the environment in which the terminal 70 has moved or to the behavior with which the terminal 70 has behaved during said observation period.

[0084] The search phase 40 also includes a step 54 of calculating the value of a similarity measure between the time signatures of the terminals 70 of the panel. If two time signatures for two terminals 70 have a significant similarity, it means that the two terminals 70 in question may have moved together or may have remained in the same place during the observation period.

[0085] The search stage 40 also comprises a step 56 of dividing the various terminals 70 of the panel into various groups depending on a similarity criterion.

[0086] Finally, the search stage 40 includes a step 44 of estimating the geographical location of the target terminal 70i depending on the geographical location data available for at least one other terminal 70 of a group, called the "target group", to which the target terminal 70i belongs.

[0087] The various steps of the search stage 40 are described in detail below.

[0088] 3. Determining the time signature Various methods are possible for the step 52 of determining the time signatures of the terminals 70 of the panel under consideration.

[0089] In a particular embodiment, the time signature of the terminal 70 comprises a value representative of a mobility stage of said terminal 70 during the observation period, the mobility stage being determined using measurements carried out by a motion sensor of the terminal 70, for example an accelerometer, or on the basis of a sensor allowing the measurement of values ​​of the terminal's environment (temperature, atmospheric pressure, light intensity, magnetic field, etc.).

[0090] A "mobility phase" means a time interval during which the terminal 70 is considered to be moving. Two successive mobility phases are separated by an immobility phase, which corresponds to a time interval during which the terminal is considered to be stationary.

[0091] Figure 4 shows the results for the observation period T obs 1 shows a schematic representation of a mobility phase of a terminal 70 of a wireless communication system 60 during an observation period T obsprecedes time t0, which corresponds, for example, to the time at which the geographical location of the target terminal 70i is estimated by the geolocation method 10. Time t0 generally corresponds to the time of receipt of a message transmitted by the target terminal 70i, on the basis of which the radio signature of the target terminal 70i is determined in step 42 of the search phase 40.

[0092] The curve 90 in Fig. 4 represents the mobility phase of the terminal 70. If the curve 90 assumes a non-zero constant value (e.g., value 1), this means that the terminal 70 is moving (the terminal 70 is in the mobility phase). If the curve assumes a null value (value 0), this means that the terminal 70 is stationary (the terminal 70 is in the immobility phase).

[0093] The curve 90 may be constructed using information conveyed by the terminal 70 in a message sent, for example, by the server 82 to the access network 80 .

[0094] For this purpose, the terminal 70 is equipped with a motion sensor, for example an accelerometer, and a set of usual means configured as software (computer program product) and / or hardware (one or more programmable logic circuits of FPGA type and / or one or more dedicated integrated circuits of ASIC type, etc.) to determine the corresponding start and end times of mobility phases of the terminal 70. For example, acceleration measurements are made repeatedly by the sensor (for example with periods of tens to hundreds of milliseconds) and, if the measured value is greater than a predefined motion detection threshold, the current detection time is stored and compared with the previous detection time. If the two detection times are separated by a duration smaller than the predefined detection threshold of the new mobility phase, the two detection times are considered to belong to the same mobility phase. However, if the two detection times are separated by a duration larger than the detection threshold of the new mobility phase, the previous detection time is considered to correspond to the end time of the mobility phase and the current detection time to the start time of the new mobility phase.

[0095] The start and end times of the mobility phase of the terminal 70 are, for example, stored and then sent to the server 82 in a message that is repeatedly transmitted by the terminal 70 to the access network 80. Such a message can, for example, be transmitted periodically or upon detection of a new mobility phase.

[0096] Based on the curve 90, the server 82 may, for example, determine a set of values ​​(M1, M2, M3, ..., M L ) over the observation period T obs A time signature for the terminal 70 in can be determined. The sampling period is selected in a suitable manner depending on the typical duration of the mobility phase and of the immobility phase of the terminal 70 for the use considered. For a fixed observation period, the higher the number of values ​​L, the more information there is about the movements performed by the terminal 70 and the more precise the time signature.

[0097] By using the same sampling over the same observation period for all terminals 70 of the considered panel, it is therefore possible to compare two-by-two the time signatures of the various terminals 70. If two terminals 70 have similar time signatures, this means that the two terminals had similar mobility phases during the observation period and, consequently, this means that the two terminals may have moved together.

[0098] In a particular embodiment, the time signature of a terminal 70 comprises values ​​representative of the atmospheric pressure to which said terminal 70 was exposed. For example, these values ​​are measured by an atmospheric pressure sensor of the terminal 70.

[0099] FIG. 5 shows the time period T obs 3 shows a curve 91 of the change in atmospheric pressure to which the terminal 70 was exposed during the test.

[0100] The curve 91 is for example constructed by the server 82 using atmospheric pressure measurements carried out by the terminal 70 and transmitted to the server 82 in messages which are repeatedly transmitted to the access network 80 .

[0101] Based on the curve 91, the server 82 may, for example, determine a set of values ​​(P1, P2, P3, . . . , P K ) over the observation period T obs A time signature for terminal 70 in the sampling period (and thus the observation period T obs The number of sampled values ​​in (K) is suitably selected depending on the use envisaged.

[0102] For example, if a pallet of goods equipped with the terminal 70 is transported by airplane, it is of interest to distinguish between an airborne phase, which corresponds to a period when the atmospheric pressure is low (due to the high altitude of the airplane), and a ground phase, which corresponds to a period when the atmospheric pressure is higher. The sampling period may therefore be selected depending on the typical duration of the airborne phase and of the ground phase.

[0103] By using the same sampling over the same observation period for all terminals 70 of the considered panel, it is therefore possible to compare two-by-two the time signatures of the various terminals 70. If two terminals 70 have similar time signatures, this means that the two terminals had similar airborne transport phases during the observation period and, consequently, this means that the two terminals may have traveled together.

[0104] It should be noted that other methods for determining the time signature on the basis of the curve 91 are possible. For example, obs into various time intervals and for each interval it is possible to determine a value, for example the average value, or the minimum or maximum value taken by the curve 91 over the time interval.

[0105] In a particular embodiment, the time signature of a terminal 70 includes values ​​representative of the temperature to which this terminal 70 is exposed. For example, these values ​​are measured by a temperature sensor of the terminal 70.

[0106] FIG. 6 shows the time period T obs 3 shows a curve 92 of the change in temperature during which the terminal 70 was exposed.

[0107] The curve 92 is for example constructed by the server 82 using temperature measurements performed by the terminal 70 and sent to the server 82 in messages transmitted to the access network 80 .

[0108] Based on the curve 92, the server 82 may, for example, determine a set of values ​​(T1, T2, T3, ..., T M ) over the observation period T obs It is possible to determine a time signature for the terminal 70 at . Here again, the sampling period is suitably selected depending on the use envisaged.

[0109] For example, if a pallet of goods equipped with the terminal 70 is transported in a refrigerated vehicle, it is of interest to distinguish between phases where the temperature is low (the pallet is in the refrigerator or in the refrigerated vehicle) and phases where the temperature is higher (the pallet is being loaded into or unloaded from the refrigerated vehicle).The sampling period is therefore selected, for example, according to the typical duration of the phase where the temperature is higher.

[0110] By using the same sampling over the same observation period for all terminals 70 of the considered panel, it is therefore possible to compare the time signatures of the various terminals two-by-two: if two terminals have similar time signatures, this means that they have been exposed to similar temperature changes during the observation period, and consequently, this means that they may have moved together.

[0111] In a particular embodiment, the time signature of a terminal 70 includes radio signatures determined for this terminal 70 at various times during an observation period.

[0112] FIG. 7 shows the time period Tobs The radio signature determined during A ,S B ,S C ,…,S L Each radio signature is, for example, obs The time series is determined on the basis of RSSI measurements performed by the base station 81 of the wireless communication system 60 upon receipt of messages transmitted by the terminal 70 at various times during the period.

[0113] For example, observation period T obs By dividing N into N time intervals with the same duration, and for each of the time intervals, we obtain the wireless signatures S A ,S B ,…,S L Depending on the N By defining the radio signatures S1, S2, ..., S N Each component of, for example,,is a radio signature S,belonging to the corresponding time interval. A ,S B ,…,S L For example, in FIG. 7, the observation period T obs The radio signature S3 corresponding to the third time interval is a set of values ​​(RSSI 3,1 ,RSSI 3,2 ,…,RSSI 3,Q ) and each component RSSI 3,i Wireless Signature S D , S E and S F The corresponding components of RSSI D,i , RSSI E,i and RSSI F,i Therefore, the observation period T obs The time signature for a terminal 70 in N ) S i The measurement time t corresponds to the time located in the middle of the time interval used to determine i For example, each wireless signature S i,i∈{1…N}.

[0114] By doing the same for all terminals 70 of the panel considered, it is therefore possible to compare the time signatures of the various terminals two-by-two: if two terminals have similar time signatures, this means that they faced similar radio conditions during the observation period, and consequently, this means that they may have traveled together.

[0115] Radio signature S1,S2,…,S N Instead of using a discrete time interval to define, we use a sliding time window mechanism to calculate the radio signature S for several positions in this sliding window. A ,S B ,…,S L It should be noted that it is possible to make each of the radio signatures S1, S2, ..., S N is the position of the sliding window and the radio signature S A ,S B ,…,S L This corresponds to the convolution between

[0116] It should also be noted that other environmental measurements can be used to define the time signature of the terminal 70, such as light intensity levels, magnetic field strength, etc. The selection of a particular type of measurement to define the time signature is merely an alternative embodiment of the invention.

[0117] In a particular embodiment, the time signature of the terminal 70 is a combination of several sets of values ​​corresponding to various environmental measurements. For example, the time signature of the terminal 70 may be a combination of the values ​​(M1, M2, M3, ..., M L ), and the values ​​(S1, S2, S3, ..., S N )

[0118] 4. Determining Similarity Criteria As regards the step 54 of calculating the value of the similarity measure between the temporal signatures of the terminals 70 of the panel considered, various methods are possible.

[0119] The time signature of the terminal 70 is obs 4, which represent the mobility stage of the terminal 70. L ) as a note. i Each of the terminals 70 is i is equal to 1 or 0 depending on whether or not the migratory phase is in progress at

[0120] A and B, respectively, are time signatures (M1, M2, …, M L ) and (M1',M2',…,M L '), for example, between the two time signatures, M A similarity measure, denoted as (A,B), can be defined by the following formula:

[0121]

number

[0122] D -1 (M i ,M i ')=C·M i M i '(Formula 3)

[0123] In formula (3), C is, for example, SIM M It is a constant defined according to the number L relating to (A, B) and the desired degree of magnitude (in this case, taking a value between 0 and L×C).

[0124] In equation (2), t i is the measured value M i and M. i ' corresponds to the time associated with the iis a weighting factor that gives more importance to recent measurements. This is advantageous if two terminals move together only for a part of the observation period. In this case, more importance must be given to recent data in order to decide whether or not the terminals in question are together at the end of the observation period, i.e. at the time t0 when the estimation of the geographical position of the target terminal 70i is performed, i.e. whether or not they should be considered to belong to the same group.

[0125] Therefore, SIM M The larger the values ​​for (A,B), the greater the observation period T obs This means that the time signatures of terminals A and B over the observation period are similar, which means that there is a high probability that terminals A and B moved together during the observation period, and more specifically towards the end of the observation period.

[0126] Now consider the case described with reference to FIG. 6, where the time signature of terminal 70 contains a temperature value. For example, if there is a SIM T A similarity measure, denoted as (A,B), can be defined using the following formula:

[0127]

number

[0128] The term sat in equation (6) represents the time t i is a constant having a small value (for example equal to 0.5° C. for the example in question) intended to avoid division by 0 when the same temperature is measured at

[0129] Now consider the case depicted in FIG. 7 where the time signature of terminal 70 includes a radio signature. For example, if two terminals, denoted as A and B, each have a time signature (S1, S2, ..., S N ) and (S1',S2',…,S N Consider two different terminals 70 with differenti Each wireless signature S determined in i is the set of RSSI levels measured by base station number Q (RSSI i,1 ,RSSI i,2 ,…,RSSI i,Q )

[0130] For example, between two time signatures, SIM S A similarity measure, denoted as (A,B), can be defined by the following formula:

[0131]

number

[0132] In equation (9), the term sat is used to avoid division by 0, and therefore SI S A constant that limits the value of (A,B).

[0133] In equations (3), (6) and (9), the function D -1 represents the inverse of the distance separating two components of the time signature. It should be noted that other functions can be chosen to define such a distance. Similarly, the weighting factor W i Other functions can be chosen to define . Such choices are merely alternative embodiments of the invention.

[0134] In a particular embodiment, the time signature of the terminal 70 comprises both a value representative of the mobility stage of the terminal 70 (e.g., the value described with reference to FIG. 4) and a radio signature of the terminal determined at various times during the observation period (e.g., the value described with reference to FIG. 7). Thus, the similarity measure SIM(A,B) may be, for example, a similarity measure SIM(A,B) for the mobility stage. M As a convex combination of (A,B), and the similarity criterion SIM for radio signatures S It can be defined as a convex combination of (A,B). SIM(A,B) = β SIM M (A,B)+(1-β)·SIM S(A,B)(Equation 10)

[0135]

number

[0136] In equation (11), M is the value M in the time signature representing the mobility stage of terminals A and B. i and M. i ' and b is a positive constant value. Note that the value M i Each of the terminals 70 is i is equal to 1 or 0 depending on whether or not the migratory phase is in progress at

[0137] Therefore, if terminals A and B have similar mobility stages at times close to t0, the coefficient β takes a larger value, and vice versa. Therefore, if terminals A and B move together at times close to t0, β is the SIM S Compare (A,B) with SIM M (A,B) serves as a weighting factor in the calculation of SIM(A,B). Conversely, if terminals A and B did not have similar mobility phases at times close to t0, then the factor β would be M Compare (A,B) with SIM S More importance is given to (A, B). Thus, it is possible to give more or less importance to the similarity with respect to the mobility phase compared to the similarity with respect to the radio environment, depending on whether the terminal has had a similar mobility phase recently with respect to the time t0 at which the position of the target terminal 70i was estimated.

[0138] 5. Split your device As regards the step 56 of dividing the terminals 70 of the panel considered into various groups according to a similarity criterion or a previously determined criterion, various methods are possible.

[0139] The purpose of this step is to estimate the geographic location of the target terminal 70i during the observation period Tobs The aim of the present invention is to determine groups of terminals that are likely to have moved together during a given period of time. This therefore entails grouping terminals 70 that have particularly high values ​​of similarity between them.

[0140] There are several known methods for dividing the terminals 70 of the panel under consideration into various homogeneous groups, in the sense that the terminals 70 of each group share similarities (i.e., the value of the similarity criterion between two terminals of the same group is relatively large).

[0141] So-called "centroid" methods, such as the "k-means" or "k-medoids" algorithms, for example, define a particular point in the space of the measurements considered, called the "centroid", and maximize the similarity between the centroid and the terminals 70 of the group. Such methods generally associate a terminal 70 with at most one group (hard clustering methods). If a terminal 70 does not have sufficient similarity with other terminals 70 of the panel, it will not be associated with any group and will remain isolated.

[0142] In one alternative, a particular partitioning method associates a terminal 70 with several groups according to a certain degree corresponding to the probability that the terminal 70 belongs to a group (soft clustering method).

[0143] In an embodiment in which the time signature of a terminal 70 is a combination of two independent sets of values, such as a combination of values ​​representing mobility stages and a combination of radio signatures, several ways are possible to implement the division.

[0144] According to a first embodiment, the partitioning is performed according to a single similarity criterion taking into account the two sets, as defined above with respect to equation (10).

[0145] According to a second embodiment, for each independent set a partition can be performed according to a similarity criterion specific to it, and a consensus is reached between the various partitions obtained.

[0146] The third embodiment is based on "conditional independence" between the partitions, which is performed according to a similarity criterion specific to each set. For example, the partition is performed for one of the two sets based on a similarity criterion SIM1 specific to it. For example, for terminal A, vector P A =(p A1 ,p A2 ,p A3 ,…,p AK ), where p Ai denotes the probability that terminal A belongs to the group with index i. The value taken by the similarity criterion SIM2 specific to the other set is then modified according to the partition obtained, for example:

[0147]

number

[0148] In the formula, p c is a predefined threshold. The final segmentation is then performed based on this modified similarity value to the reference SIM2.

[0149] The methods presented above for performing division 56 of the terminals 70 of the panel are given as examples and are in no way limiting. Other methods are possible and the selection of a particular method is merely an alternative embodiment of the present invention.

[0150] 6.Determine the terminal panel The panel considered in steps 52, 54 and 56 can be determined, for example, by the selection of a terminal 70 of the wireless communication system 60, in which it is known a priori that it is close (geographically speaking) to the target terminal 70i, or that it has certain properties that make it belong to a group of several terminals 70.

[0151] In a particular embodiment, only terminals 70 that were covered by at least one base station 81 of the wireless communication system 60 during the transmission of the terminal 70's last message prior to time t0, and whose base station 81 also covers the target terminal 70i at time t0, are selected.

[0152] In a particular embodiment, the BS receives the message transmitted by the target terminal 70i at time t0. bs Denotes a set of base stations 81, N mvt (A) is the observation period T obs This indicates the number of mobility stages of terminal A in the , for example, the following is defined:

[0153]

number

[0154] Therefore, P tot Only terminals for which (A) is greater than a predefined threshold are selected in the panel.

[0155] In equation (13), RSSI X,bs bs∈BS, is the RSSI level measured by the base station bs for the last message sent by terminal X before t0. Therefore, P radio (A) is a value representing the probability of finding a terminal 70 having a radio environment similar to that of terminal A among other terminals 70 in the system 60 .

[0156] In formula (14), σ is, for example, a value N mvt It corresponds to the standard deviation of (X). Therefore, P mvt (A) is for period T obs , which represents the probability of finding a terminal 70 in the system 60 that has a mobility stage number close to the mobility stage number of terminal A.

[0157] Therefore, the term P defined by equation (15) tot(A) represents the property of terminal A discovering terminals 70 that are exposed to similar conditions in terms of radio environment and movement among other terminals 70 in the system 60. In other words, P tot The larger (A), the greater the probability that terminal A can be grouped with other terminals 70 in the system 60.

[0158] It should be noted that the methods presented above for determining the panel of the terminal 70 are given as examples and are in no way limiting. Thus, the selection of a particular method is merely an alternative embodiment of the present invention.

[0159] 7. Estimating the geographic location of the target device The search stage 40 of the geographical location method 10 according to the invention comprises a step 44 of estimating the geographical location of the target terminal 70i as a function of the geographical location data available to at least one other terminal 70 of a group, called the "target group", to which the target terminal 70i belongs.

[0160] In the remainder of this document, by way of example and in no way limiting, the time signature of the terminal 70 is taken to be the time signature of the terminal 70i during the observation period T obs The time t0 is considered to be a combination of a value representative of a mobility phase during the immobility phase and the radio signature of the terminal 70. Advantageously, the time t0 is considered to correspond to the time at which a message sent by said target terminal 70i during the immobility phase was received by the server 82.

[0161] It is contemplated that the panel of the terminal 70 is determined according to any one of the methods described above.

[0162] It is also conceivable to perform the division of the terminals 70 of the panel using a similarity criterion as described above by equation (10).

[0163] It is also conceivable that the geographical location of the terminals 70 of the panel may be estimated using conventional geolocation algorithms, for example as described with reference to Fig. 2. As already mentioned above, such estimation of the geographical location may be complex and relatively inaccurate, especially when the geographical zone to be covered is large and / or the number of terminals and / or base stations is large.

[0164] The remainder of the specification presents, by way of non-limiting examples, several ways of using the resulting partitioning and geolocation data provided by one or more terminals 70 of the target group, or even other groups, to estimate the geographic location of the target terminal 70i with greater accuracy and / or reduced complexity.

[0165] In a particular embodiment, the geographic location data is a GPS position provided by a calibration terminal 70c that is equipped with a GPS receiver.

[0166] According to a first embodiment, the target terminal 70i is associated with a single group ("hard clustering" method) and at least one calibration terminal 70c belonging to this group has provided its GPS position at a time preceding t0, a time since which it has not moved, via a message sent to the access network 80. The geographical location of the target terminal 70i is therefore associated with its GPS position (or with the average of several GPS positions available to the group).

[0167] According to a second embodiment, the target terminal 70i is associated with several groups, with a certain probability of belonging to each group ("soft clustering" method), and a geographical location is assigned to each group on the basis of at least one GPS position provided by the calibration terminal 70c of said group (or optionally by averaging several GPS positions available to the group). k is the geographic location belonging to group k, and x j kis the GPS position of the calibration terminal 70c denoted by j in group k, and p jk is the probability that terminal j belongs to group k. Therefore, the position Y i is determined according to the geographical location belonging to the group and the probability that the target terminal 70i belongs to various groups, according to the following formula:

[0168]

number

[0169] In a particular embodiment, the geolocation data is the geographical location of the terminal 70 estimated by a conventional geolocation algorithm, such as that described with reference to Fig. 2. Preferably, said geographical location is estimated on the basis of messages transmitted relatively close to a time t0, after which the terminal 70 is no longer moving.

[0170] For example, a target terminal 70i is associated with a single group ("hard clustering") and the geographic location of the target terminal 70i is determined depending on one or more estimated geographic locations of other terminals 70 of this group, for example by averaging the estimated geographic locations of the various terminals 70 of the group, optionally while eliminating a particular estimated location if it is determined to be anomalous with respect to the other locations, or while prioritizing a particular estimated geographic location with respect to other locations depending on the RSSI level of the associated radio signature.

[0171] According to another embodiment, the target terminal 70i is associated with several groups and has a certain probability of belonging to each group ("soft clustering"), a geographical location is assigned to each group based on the estimated geographical locations of the terminals 70 of said groups, and the geographical location of the target terminal 70i is then estimated by a method similar to that described above using equation (17).

[0172] In a particular embodiment, the target terminal 70i is associated with a single group ("hard clustering") and the geolocation data is the radio signature of the terminal 70. Preferably, said radio signature is determined on the basis of messages transmitted relatively close to t0, a time after which the terminal 70 no longer moves. A virtual radio signature is then determined for the group to which the target terminal 70i belongs, for example by averaging each component of the radio signature or by taking the maximum value of the radio signature for each component. The geographic location of the target terminal 70i is then conventionally estimated on the basis of the virtual radio signature and the set of reference elements 26.

[0173] It should be noted that in the very special case where the target terminal 70i does not belong to any group, for example because it does not have sufficient similarity to the other terminals 70 of the panel, a mediocre estimate of the geographical location without utilizing information about the division should be sufficient.

[0174] 8. Conclusion The above description clearly shows that the invention achieves the set objects by means of its various features and their advantages.

[0175] In particular, the geolocation method according to the invention provides advantages in terms of geolocation accuracy compared to methods according to the prior art.

[0176] Indeed, in certain embodiments, the location of the target terminal 70i may be determined directly based on the GPS locations of one or more of the calibration terminals 70c that belong to the group to which the target terminal 70i belongs.

[0177] In other embodiments, GPS data is not available to estimate the location of the target terminal and therefore the geographical locations of other terminals 70 of the system, estimated by conventional geolocation methods based on radio signatures, must be used as geolocation data, for example. Even in this case, the accuracy of the geolocation of the target terminal 70i is significantly improved compared to conventional methods, due to the correlation of the various information available to the terminals 70 belonging to the group to which the target terminal 70i belongs.

[0178] The invention is particularly well suited when the geographical zone to be covered is large (e.g. when a city, a country or even a continent has to be covered) and / or when the number of terminals is large (e.g. thousands to hundreds of thousands of terminals) and / or when the number of base stations is large (e.g. tens, or even hundreds, or even thousands of base stations).

[0179] The complexity of the geographic location method is also improved in terms of time and computational power: for example, it is possible to determine the geographical positions of all terminals 70 of a group by attributing to them the same geographical position (corresponding to the geographical position of the group) estimated on the basis of the geographical positions of only one terminal or of several terminals belonging to said group.

[0180] In general, it should be noted that the embodiments considered above are described as non-limiting examples, so that other alternatives are possible.

[0181] In particular, there are several ways to determine the radio signature for the terminal 70: based on the RSSI level of the signals exchanged between the terminal 70 and the base station 81, based on the propagation time of the signals exchanged between the terminal 70 and the base station 81, etc. The selection of a particular method is merely an alternative embodiment of the present invention.

[0182] Also, there are several ways to determine the time signature for the terminal 70. The choice of a particular type of measurement (detection of mobility phases, temperature values, pressure values, etc.) is merely an alternative embodiment of the invention.

[0183] Similarly, there are several ways to determine the similarity measure between two temporal signatures, and the selection of a particular method is merely an alternative embodiment of the present invention.

[0184] Several methods are possible for dividing the terminals 70 into groups according to similarity criteria, or for determining the panel of terminals 70 to be divided into. Again, the choice of a particular method is merely an alternative embodiment of the present invention.

[0185] Also, different types of geolocation data provided by a particular terminal 70 of the system are possible. Although the GPS location of the terminal 70 is given as an example of geolocation data, other geolocation data are possible. For example, the MAC ("Medium Access Control") address of a Wi-Fi access point to which the terminal 70 is connected can serve as one of the geolocation data, if the geographic location of said access point is known.

[0186] The concept of the present invention can also be applied to the geolocation of a single target terminal 70i over time, independent of other terminals 70 of the system 60. For example, in the case where a mobility phase of the target terminal 70i is not detected during two times t1 and t2 corresponding to the radio signatures S1 and S2, respectively, both radio signatures S1 and S2 can be used to estimate the same location of the target terminal 70i if they are similar. As an extension, if the radio signatures S1 and S2 are N≧2, then i groups are similar and at time t i(index i varies from 1 to N) and this group can be used if the target terminal 70i has not moved in the meantime. Such a method may make it possible to improve the accuracy of the estimation of the geographical position of the target terminal 70i as well as to correlate the estimations performed in the past. However, if the radio signatures are not similar, it can be assumed that a mobility phase has been missed and a different geographical position must be estimated for each radio signature.

[0187] The invention has been described by way of example, while also considering an IoT type wireless communication system 60 for use in the field of logistics (transportation of pallets equipped with terminals of the IoT communication system 60), but this does not exclude considering, according to other embodiments, other wireless communication systems such as mobile telephone communication networks and, optionally, other uses in other industrial fields.

Claims

1. A method (10) for estimating the geographical position of a terminal (70) called a "target terminal (70i)" from a panel of the terminal (70) of a wireless communication system (60), said method comprising the steps of: determining (52) for each of the terminals (70) of the panel, data including a value representative of the mobility stage of said terminal (70), called a "time signature", measured independently by said terminal (70) by at least one sensor of said terminal (70) during a predetermined observation period; - calculating (54) a value of a similarity measure between the temporal signatures of the terminals (70) of the panel, the value of the similarity measure calculated for two terminals (70) representing the probability that the terminals (70) have moved together during the observation period; A step (56) of dividing the various terminals (70) of the panel into various groups depending on the value of the similarity criterion; a step (44) of estimating, after said dividing step (56), said geographical location of said target terminal (70i) depending on geolocation data available for at least one other terminal (70) of said group, called "target group", to which said target terminal (70i) belongs; A method (10), comprising:

2. The value measured by at least one sensor of the terminal (70) is Values ​​measured by a motion sensor, and / or a value measured by a temperature sensor, and / or a value measured by an atmospheric pressure sensor, and / or a value measured by a light intensity sensor, and / or The method of claim 1 , comprising a value measured by a magnetic field sensor.

3. The method (10) of claim 2, wherein the time signature of the terminal (70) further comprises data including a value representative of the quality of a radio link, called a "radio signature", existing between the terminal (70) and one or more base stations (81) of the wireless communication system (60) at various times during the observation period, and the similarity criterion is determined according to a similarity criterion for the mobility phase and a similarity criterion for the radio signature.

4. 4. The method (10) according to claim 1, wherein the similarity criterion is determined as a function of a weighting factor assigned to the measured value, said weighting factor being representative of the time elapsed between the time when said value was measured and the time when the geographical position of said target terminal (70i) was estimated (44).

5. The method (10) according to any one of claims 1 to 4, wherein the step (56) of dividing the various terminals (70) of the panel is performed in such a way that the terminals (70) belong to at most one group.

6. 5. A method (10) according to claim 1, wherein the step (56) of dividing the various terminals (70) of the panel is performed such that each of the terminals (70) belongs to each of the groups depending on the value of its probability of belonging, the target group corresponding to the group in which the value of the probability of belonging of the target terminal (70i) is the largest.

7. 7. The method (10) of claim 6, wherein the geographical location of the target terminal (70i) is estimated depending on geolocation data available for different terminals (70) belonging to different groups and depending on the values ​​of the probabilities of belonging of the target terminal (70i) to the different groups.

8. The method (10) according to any one of claims 1 to 7, wherein the geographic location data is provided by a terminal (70) in the form of a geographic position obtained by a satellite positioning system of the terminal.

9. The method (10) according to any one of claims 1 to 7, wherein the geographic location data is available to the terminal (70) in the form of data called a "radio signature", which includes a value representative of the quality of a radio link existing between the terminal (70) and one or more base stations (81) of the wireless communication system (60).

10. 10. The method of claim 9, wherein the geographic location of a terminal (70) is estimated in response to the radio signature of the terminal (70) and a reference set (26) including radio signatures associated with known geographic locations, and the geographic location of the target terminal (70i) is estimated in response to the estimated geographic location of at least one other terminal (70) of the target group.

11. The method (10) according to claim 9, wherein, in response to the radio signatures of the terminals (70) belonging to the target group, data called a "virtual radio signature" is determined for the target group, and in response to the virtual radio signature, the position of the target terminal (70i) is estimated.

12. The method (10) according to any one of claims 1 to 11, wherein the panel of terminals (70) is determined by a step of selecting from among the terminals (70) of the wireless communication system (60), whereby the terminals (70) of the panel are covered by at least one base station (81) of the wireless communication system (60) which also covers the target terminal (70i).

13. A wireless communication system (60) comprising a terminal (70) and an access network (80) comprising a base station (81) and a server (82) connected to said base station (81), said system (60) being adapted to implement a method for geolocating a target terminal (70i) from a panel of said terminal (70) according to the method of any one of claims 1 to 12, said server (82) comprising: collecting values ​​representative of the mobility stages of the terminals (70) of the panel, measured independently by at least one sensor of the terminals (70) during a predetermined observation period; determining said data, called a "time signature" for said terminal (70), on the basis of said measured values, calculating values ​​of a similarity measure between the temporal signatures of the terminals (70) of the panel such that the value of the similarity measure calculated for two terminals (70) represents the probability that the terminals (70) have moved together during the observation period; Dividing the various terminals (70) of the panel into various groups depending on the values ​​of the similarity criterion, after the division of the terminals, estimating the geographical location of the target terminal (70i) as a function of geolocation data available for at least one other terminal (70) of the group to which the target terminal (70i) belongs, called the "target group" - A wireless communication system (60) configured as follows.

Images