Subscription selection with central cloud intelligence

Abstract

The present invention relates to a method for selecting suitable subscriptions to be loaded onto an eUICC of devices from an IoT fleet among candidate subscriptions, said method comprising: collecting teacher data related to MNO subscriptions and their associated favorite and forbidden PLMNs from field deployed devices and / or laboratory devices hosting samples for a central cloud intelligence, and associating said teacher data with the candidate subscriptions in a database, said method further comprising: receiving information about the radio environment of the device requiring a subscription, processing the received information about the radio environment regarding the teacher data available in the database and the candidate subscriptions, determining an adapted subscription among the candidate subscriptions according to this processing, and triggering a download of the determined subscription to the device requiring the subscription.

Description

[Technical Field]

[0001] FIELD OF THE INVENTION The present invention relates to a method for selecting subscriptions to be loaded onto IoT devices from among candidate subscriptions implemented by a central cloud intelligence.

[0002] The invention also relates to a central cloud intelligence for implementing the above method. [Background technology]

[0003] Background of the Invention The present invention addresses the situation where an enterprise customer wants to load subscriptions into each field device from a fleet of cellular IoT devices. This can be done using a bootstrap profile for initial provisioning or using an already preloaded operational profile in the case of a Mobile Network Operator (MNO) switch. This situation typically occurs when a service provider or original equipment manufacturer (OEM) is requested to select a field device, such as a smart meter, and download its subscriptions.

[0004] In such situations, customers may face subscription download errors, which can result in additional costs and service outages / degradation of associated devices. The larger the IoT device fleet, the greater the cost impact.

[0005] For fleets of metering or highly distributed, sometimes high-value devices, this situation can be very detrimental and you need to ensure the correct subscriptions are downloaded to each device in the fleet.

[0006] An existing solution would be to obtain information from operators in an exhaustive manner regarding their roaming partners and RATs. With all this information, a database can be established and maintained to record such information. However, such a solution presents the drawback of being static and dependent on contact with the operator and its files and contracts.

[0007] Therefore, further alternative and advantageous solutions would be desirable in the art. Summary of the Invention [Problem to be solved by the invention]

[0008] Summary of the Invention The present invention aims to reduce or eliminate the risk of loading subscriptions that will not work once deployed to individual devices. [Means for solving the problem]

[0009] The present invention is defined in its broadest sense as a method for selecting subscriptions to be loaded onto an embedded universal integrated circuit card of an Internet of Things device from among candidate subscriptions, said method comprising the steps of: collecting training data related to mobile network operator subscriptions and their associated favorite and / or barred PLMNs from field-deployed Internet of Things devices and / or from Internet of Things lab devices hosting a sample of subscriptions; -Associating those teacher data with candidate subscriptions in the database; and Including, The method comprises: receiving information about the wireless environment of an Internet of Things device that requires a subscription; - processing the received information about the wireless environment regarding teacher data and candidate subscriptions in the database; - determining a subscription from among the candidate subscriptions that is adapted in response to this process; triggering a download of the determined subscription to a device requiring the subscription; Further includes:

[0010] The principle of the present invention is to train a central cloud intelligence based on training data collected from field and / or in-lab devices. Such field devices are typically deployed for production purposes, while in-lab devices can be dedicated blueprint devices for running some experiments. It should be noted here that Mobile Network Operators (MNOs) are indiscriminately physical and therefore fully own the Public Land Mobile Networks (PLMNs) of their MNO partners, either partially or fully using their PLMN equipment or virtual, i.e., Mobile Virtual Network Operators (MVNOs).

[0011] After educating the cloud intelligence, as a prerequisite for determining subscriptions, the central cloud intelligence needs to know the radio environment of each individual device. The radio environment is typically a list of PLMNs known to the device, along with associated radio access technologies (RATs), power strength, and signal quality parameters. Such parameters are typically RSSI, RSCP, RSRP, ECN0, and RS-SINR, as defined in the 3GPP standard.

[0012] The present invention aids central cloud intelligence, responsible for more relevant and efficient dispatching of MNO subscriptions to fleets of cellular devices containing embedded Universal Integrated Circuit Cards (eUICCs). MNOs have a habit of pushing lists of favorite and forbidden PLMNs to SIM cards. Such a mechanism allows MNOs to influence the PLMNs selected by devices, as defined in 3GPP standards. The uniqueness of the present invention is that it leverages such information by pulling it into an MNO-independent central cloud intelligence, so that service providers using such central cloud intelligence can use it as training data to download the most suitable MNO subscriptions to each of their devices. In effect, it implements a kind of roaming partner map in a centralized, MNO-independent manner. Such a mechanism then allows service providers to maximize the relevance of MNO subscription selection for each of their devices.

[0013] The present invention, through its education mechanism, helps to create a complete fleet of devices that benefit from the experience of a small number of devices thanks to its adjustable training data collection methods.

[0014] According to an advantageous feature, the training data relating to MNO subscriptions and their associated favourite and forbidden PLMNs comprises some standard 3GPP information to be read from the MNO profile in the eUICC.

[0015] Standard 3GPP information includes: -Home PLMN, - equivalent PLMN, -Priority PLMN, -Ban PLMN.

[0016] These are the Home PLMN, PLMNs equivalent to the Home PLMN, Preferred PLMNs, and such a list includes PLMNs that should be used as PLMN selectors from the operator and user perspective, and forbidden PLMNs.

[0017] According to an advantageous feature, the collecting of training data is periodic. This allows the server to decide to read the training data at regular intervals, for example, once a month.

[0018] According to another feature, collecting the training data occurs after a given number of periodic subscriptions have been successfully downloaded to the device.

[0019] This feature allows you to collect, for example, every 1,000 devices that have successfully downloaded the same type of subscription.

[0020] According to another feature, collecting the training data is performed after a network registration failure experienced by the IoT device, and such network registration failure directs an update of a forbidden PLMN list for an associated subscription in an embedded universal integrated circuit card of the failed IoT device.

[0021] This feature allows for exemplary real-time updates of subscription and training data associations in the event of a device failure.

[0022] According to another feature, collecting training data is performed from devices registered with a given mobile network operator after an update of a subscriber identity module file triggered by the given mobile network operator.

[0023] This feature makes it possible to take into account any kind of update, including mass updates, of all subscription eUICCs with a given mobile network operator.

[0024] Collecting training data may be a combination of the above features. According to an advantageous feature, the method comprises, in advance, for collecting training data: -Access to hashes of training data; Comparing the received hash value with a previous hash value calculated on the most recently updated training data for the given subscription; further comprising Collecting training data is triggered only if the hash value differs from the previous hash value.

[0025] Due to this feature, the use of a hashing mechanism avoids the cloud intelligence server from reading the entire training data if the training data is identified as being the same as some previously downloaded training data.

[0026] Using the above features, the present invention proposes a dynamic method for creating and maintaining a frequently evolving MNO roaming partner map in a central cloud intelligence database, composed of training data. Because of its short cycle, it can actually move quasi-continuously. Therefore, the present invention provides a dynamic method for constantly selecting and downloading the optimal MNO subscription to each IoT fleet device.

[0027] According to a first embodiment, receiving information about the radio environment of a device requiring a subscription is preceded by scanning network frequencies for the device requiring a subscription to identify available PLMNs and associated RATs that make up the radio environment, and transmitting this information about the radio environment to a central cloud intelligence.

[0028] In this embodiment, a specific report from the individual device itself is generated with the help of a scan, and the information about the device's radio environment as contained in the report is then processed by cloud intelligence and combined with the MNO roaming partner map to determine the best subscription to download for this particular device.

[0029] Advantageously and classically, the radio environment may also include signal power and quality information associated with the discovered PLMNs and associated RATs.

[0030] According to a second embodiment, receiving information about the radio environment of a device requiring a subscription is preceded by scanning network frequencies for a device companion of the device requiring a subscription to identify available PLMNs, associated RATs that make up the radio environment, and transmitting this information about the radio environment to a central cloud intelligence.

[0031] In this embodiment, the specific report is generated by a device companion close to the target device, for example, it can be a smartphone with an application specific to the present invention that is close and paired with the IoT device for which the subscription is to be provisioned.

[0032] The present invention also relates to a cloud infrastructure having a central cloud intelligence for implementing the method of the present invention, said cloud infrastructure having connectivity to collect training data relating to MNO subscriptions of a mobile network operator and their associated favourite and forbidden PLMNs from field deployed devices and / or lab devices hosting a sample of subscriptions, The cloud infrastructure further has a processing module for processing the received information about the wireless environment regarding the training data and candidate subscriptions in the database when the cloud infrastructure receives information about the wireless environment from an Internet of Things device requiring a subscription, and associating the training data with the candidate subscriptions in the database to determine a subscription adapted from among the candidate subscriptions in accordance with this processing, and the cloud infrastructure then triggers download of the determined subscription to the device requiring the subscription.

[0033] Such a cloud infrastructure implements the central cloud intelligence of the present invention for managing subscription selection for a fleet of devices in a reliable and relevant manner, which allows the central cloud intelligence to filter out subscriptions that do not fit and advantageously also filter out subscriptions that are at risk, i.e., that do not have an explicit PLMN in the favorites list.

[0034] The present invention also relates to an Internet of Things device having an embedded universal integrated circuit card into which candidate subscriptions are downloaded, said device having a cellular module or chipset supporting network scanning or being connectable to a device companion, said cellular module or chipset supporting network scanning or said device companion being adapted to scan network frequencies to identify available PLMNs and associated RATs that make up the wireless environment, said device further having a transmitting / receiving cellular module or chipset for transmitting this information about its wireless environment to a central cloud intelligence of the present invention and for receiving the downloaded subscriptions.

[0035] The present invention therefore also relates to a "chip-on-board" solution, and therefore also to devices of the IoT that use the chipset directly without a cellular module.

[0036] Such devices are therefore provided with subscriptions that are adapted to their wireless environment, thus reducing the risk of downloading the wrong subscription.

[0037] Advantageously, the device further has agent software within the embedded universal integrated circuit card for reading training data relating to mobile network operator subscriptions and their associated favorite and forbidden PLMNs and transmitting the read training data to a central cloud intelligence.

[0038] Using such agent software, devices are adapted to feed further training data to a central cloud intelligence on a periodic or regular basis, which dynamically renders the database in the central cloud intelligence. This is particularly beneficial because MNO roaming partner maps are tedious to establish manually and nearly impossible to maintain due to their high update cadence.

[0039] To the accomplishment of the foregoing and related ends, the one or more embodiments comprise the features hereinafter fully described and particularly pointed out in the claims.

[0040] BRIEF DESCRIPTION OF THE DRAWINGS The following description and the annexed drawings set forth in detail certain illustrative aspects and illustrate some of the various ways in which the principles of the embodiments may be employed. Other advantages and novel features will become apparent from the following detailed description when considered in conjunction with the drawings, and the disclosed embodiments are intended to include all such aspects and their equivalents. [Brief explanation of the drawings]

[0041] [Figure 1] 1 shows a schematic representation of the situation of the present invention; [Figure 2] 1 shows a schematic diagram of the general functioning of the method of the present invention. [Figure 3] FIG. 1 is a time diagram illustrating a set of experiences according to the present invention. [Figure 4] 1 illustrates the use of the present invention to select a subscription. [Figure 5A] 3 illustrates implementation options for the subscription selection method of the present invention. [Figure 5B] 3 illustrates implementation options for the subscription selection method of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0042] Detailed Description of Embodiments of the Invention For a more complete understanding of the present invention, the present invention will now be described in detail with reference to the accompanying drawings. The detailed description illustrates and describes what are considered to be preferred embodiments of the present invention. It should, of course, be understood that various modifications and changes in form or detail can be readily made without departing from the scope of the present invention. Therefore, the present invention may not be limited to the exact forms and details shown and described herein, and it is intended that the present invention be limited to less than the entire invention disclosed herein and claimed below. Identical elements are designated by the same reference numerals in different drawings. For clarity, only those elements and steps useful for understanding the present invention are shown in the drawings and described.

[0043] 1 shows a schematic illustration of an environment in which the invention is exemplarily implemented. The invention relates to a central cloud intelligence CCI implemented on at least one server connected to several devices D1 to D1.

[0044] Of these devices, at least two, D1 and D2, are loaded with subscriptions from Mobile Network Operator A (MNO A) and Mobile Network Operator B (MNO B), respectively, and are therefore sending training data TDA and TDB to the central cloud intelligence CCI. The training data TDA and TDB are information related to subscriptions such as home PLMN, equivalent PLMN, and their associated RATs, if applicable. Such information is classically found in files on the SIM card.

[0045] This is also shown in Figure 2, where training data TDA and TDB from devices loaded with different subscriptions from MNOs A and B are sending their training data to the central cloud intelligence CCI. The training data TDA and TDB are stored in the central cloud intelligence CCI's central cloud intelligence database (CCI DB) associated with the candidate subscription. The CCI DB then contains some kind of up-to-date roaming partner map with a list of favorite and forbidden PLMNs for subscriptions from MNOs A and B.

[0046] The devices D3 to Di are devices located in various radio environments RE3 to REi that are scanned by each device D3 to Di. The scanned radio environments REi are then provided to a central cloud intelligence CCI, as shown schematically in Figures 1 and 2.

[0047] The radio environment REi of each device is then used by the central cloud intelligence CCI and combined with the database CCI DB to determine adapted subscriptions AS3 to ASi that are downloaded to the devices D3 to Di respectively, as shown in Figure 2.

[0048] In an advantageous embodiment of the invention, all devices D1 to Di are then used to supply the central cloud intelligence CCI with further updated data constituting further training data TD1 to TDi.

[0049] Figure 3 shows a timeline of the learning phase by the central cloud intelligence CCI. A device D participating in the provisioning of the central cloud intelligence CCI comprises at least a cellular module or chipset and an embedded universal integrated circuit card (eUICC) to which a subscription is to be provisioned. Based on such subscription, the device D selects a public land mobile network (PLMN) and performs an installation procedure on it. Here, a subscription already exists in the eUICC for the learning phase of the central cloud intelligence CCI. The device D in the field or laboratory can access information stored in the associated SIM profile installed on the eUICC. Agent software (AG) running on the device D can read the list of favorite and forbidden PLMNs associated with the subscription and transmit such training data to the central cloud intelligence CCI.

[0050] The device further comprises a Modem Stack (MS) as an interface between the eUICC and a PLMN, which may be a Home PLMN where the device D is connected to a network belonging to the issuer of the subscription, or a Visited PLMN if the device is roaming.

[0051] The agent software AG is further connected to a central cloud intelligence CCI and processes information relating to the present invention.

[0052] In a first step S1, the modem stack MS reads the subscriptions stored therein in the eUICC module and accesses the PLMN lists, i.e., the lists of forbidden, preferred, and home PLMNs. Using these lists, the modem stack MS selects the most suitable PLMN and proceeds to the attachment procedure in step S2. In the event of a failure due to a specific cause, the selected PLMN can be declared "forbidden," and the modem stack MS updates the list of PLMNs in the eUICC profile accordingly and restarts the PLMN selection and attachment procedure using another available PLMN. In such a case, the eUICC profile of this device D contains some new useful information that becomes new training data for the CCI in the next step. Such an update can also be performed using other lists of PLMNs known by the eUICC.

[0053] The successive attachment procedures of the device D to different PLMNs indeed result in an update of the SIM PLMN list that constitutes the training data TD of the invention. The dashed procedure is therefore an example of a procedure for performing an update of the PLMN list.

[0054] The agent AG then reads the updated PLMN lists in step S4, these updated PLMN lists being teaching data TD, and sends the updated PLMN lists associated with active eUICC subscriptions to the central cloud intelligence CCI in step S5, which then stores these lists in association with the type of subscription of the corresponding device in the database CCI DB in step S6.

[0055] According to the present invention, the central cloud intelligence CCI can read such information from the devices at different times depending on the implementation. Such reading can be performed when a new type of subscription is first installed on an eUICC in the device fleet. Optionally, a hashing mechanism can be deployed to allow the server to avoid reading the same training data TD several times during the device Di step 5, which will be published by the same device D in a further subsequent reading or by different devices D for the same type of subscription. In such cases, it is advantageous for the server to first systematically read the hash to determine whether it should read the complete training data TD associated with the subscription.

[0056] Thus, the server may decide to read the training data once at regular time intervals, for example monthly, or / and once on a regular subscription that has been successfully downloaded to the device, for example every 1000 successful downloads.

[0057] Thus, the present invention allows leveraging the same PLMN list as used by the modem stack of device D during 3GPP PLMN selection. This is shown in FIG.

[0058] According to the invention, such a device D performs a scan of the radio environment RE that results in a list of available PLMNs, RATs, power strengths and signal quality parameters. This scan allows the device D to know the active frequencies.

[0059] Such a scan of the radio environment RE can be performed according to two options shown in Figures 5A and 5B.

[0060] In the first case, the device D is independent and proceeds to scan the radio environment: According to the invention, the agent AG requests the modem stack MS to start an informal network scan in step S7.

[0061] The informal scan is indicated by step S8 where available PLMNs, RATs, power strengths and signal quality parameters are listed.

[0062] Then, in step S9, the agent AG reads the radio environment RE established in step S8. This is an example of a procedure for obtaining the radio environment RE of an individual device that needs to download a subscription. This allows the agent AG to send the radio environment RE to the central cloud intelligence CCI, as shown in Figure 4.

[0063] In step S11, the central cloud intelligence CCI then processes information about the radio environment RE using training data TD stored in the subscription database CCI DB associated with the candidate subscription. In this step S11, the central cloud intelligence CCI compares a list of available PLMNs / RATs recognized by the device with a list of PLMNs / RATs compatible with the candidate subscription. The central cloud intelligence CCI then filters out incompatible subscriptions, such as those classified as prohibited PLMNs, for the candidate subscription involving all PLMNs / RATs recognized by the device, according to the subscription database CCI DB. The central cloud intelligence CCI then filters out subscriptions that may pose a risk of incompatibility, such as those not explicitly listed in the list of favorite PLMNs for the candidate subscription involving all PLMNs / RATs recognized by the device, such as home PLMNs, equivalent PLMNs, and preferred PLMNs, according to the subscription database CCI DB. The central cloud intelligence CCI then selects the best subscription from the remaining subscriptions based on several criteria for selection, such as best signal strength and best price. Finally, in step S12, the selected subscription AS is downloaded to the module eUICC of the device D.

[0064] 5B illustrates a second embodiment of the present invention in which device D is not the most suitable device for performing a frequency scan and is not a candidate for determining the radio environment RE. For example, device D is a meter. Therefore, it relies on another device companion DC to perform network analysis via a network scan. Therefore, typically, a smartphone with an installed application APP is dedicated to managing device D or at least activating device D. This smartphone is a device companion DC with a telecommunications stack TS.

[0065] In step S7', the application APP requests the telecommunications stack TS to perform a network scan, which is executed by the telecommunications stack TS in step S8'. Then, in step S9', the application APP reads the radio environment in the telecommunications stack TS, which lists available PLMNs, RATs, power strength and signal quality parameters. This is an example of a procedure for obtaining the radio environment RE of an individual device by utilizing the device companion DC.

[0066] Next, the application APP can provide the wireless environment RE to the central cloud intelligence CCI in step S10, as shown in Figure 3, which triggers the determination of the best subscription AS for the wireless environment RE by the central cloud intelligence CCI in step S11 using the training data stored in the database CCI DB, and then triggers the download of the subscription AS in the device D in step S12.

[0067] Thus, according to the present invention, one or a small number of devices collect inputs from the eUICC hosting one sample of candidate subscriptions, including the home PLMN, PLMNs equivalent to the home PLMN, preferred PLMNs that allow PLMN selection, and forbidden PLMNs. This or these devices then transmit such inputs to the central cloud intelligence CCI, which is responsible for dispatching / delivering MNO subscriptions, regardless of whether this provisioning is performed in push mode or pull mode. The central cloud intelligence CCI thus educates a subscription selection algorithm that is more relevant to the subscription selection of the rest of the device fleet, as follows:

[0068] - Not selecting a subscription if any PLMN recognized by the device involved is on the forbidden PLMN list Select a subscription for which an equivalent PLMN is identified among the PLMNs known to the device involved -Select a subscription for which a preferred PLMN is identified among the PLMNs known to the device involved.

[0069] Such an algorithm retrieves training data TD in the database CCI DB as the basis for determining whether a subscription is adapted to a device having a particular radio environment RE or whether this subscription is not adapted. For example, the algorithm may filter out subscriptions of MNOs whose PLMNs do not appear in the radio environment RE of the device for which the subscription is provisioned, thereby avoiding erroneous subscription downloads in devices that do not have a network corresponding to that radio environment RE.

[0070] In the foregoing detailed description, reference is made to the accompanying drawings which show, by way of illustration, specific embodiments in which the invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. The foregoing detailed description, therefore, is not to be taken in a limiting sense, and the scope of the invention is defined only by the appended claims, appropriately interpreted.

Claims

1. 1. A method for selecting a subscription to be loaded onto an embedded universal integrated circuit card of an Internet of Things device from among candidate subscriptions, the method comprising: collecting training data relating to mobile network operator subscriptions and their associated favorite and forbidden PLMNs from each of a plurality of field-deployed Internet of Things devices and / or a plurality of Internet of Things lab devices hosting a sample of subscriptions; the training data relating to mobile network operator subscriptions and their associated favorite and forbidden PLMNs comprises some standard 3GPP information to be read from the mobile network operator profile in the embedded universal integrated circuit card; The standard 3GPP information includes at least: - Home PLMN, - equivalent PLMN, -Priority PLMN, - Ban PLMN, Including, - extracting candidate subscriptions corresponding to the training data from within a database; receiving information about the radio environment of an Internet of Things device requiring a subscription, the radio environment including a list of PLMNs recognized by the device; - filtering out from the candidate subscriptions candidate subscriptions of mobile network operators for which the favorite PLMN of the device requiring a subscription does not appear in the radio environment in order to process the received information about the radio environment; - determining from among said candidate subscriptions a subscription that is adapted in response to this processing; - triggering a download of the determined subscription to the device requiring the subscription; The method further comprises:

2. The method of claim 1 , wherein the collecting of training data is periodic.

3. The method of claim 1 , wherein the collecting of training data occurs after a given number of periodic subscriptions have been successfully downloaded to the device.

4. 2. The method of claim 1, wherein the collecting of training data occurs after a network registration failure experienced by an IoT device, and such network registration failure directs an update of the forbidden PLMN list for the associated subscription in the embedded universal integrated circuit card of the failed IoT device.

5. 2. The method of claim 1, wherein the collecting of training data is performed from devices registered with a given mobile network operator after an update of a subscriber identity module file triggered by the given mobile network operator.

6. Preliminary to collect teacher data, - accessing a hash of the training data; - comparing the received hash value with a previous hash value calculated on the last updated training data for the given subscription; further comprising The method of claim 2 , wherein the collecting of training data is triggered only if the hash value differs from the previous hash value.

7. 2. The method of claim 1, wherein the receiving of information about the radio environment of a device requiring a subscription is preceded by scanning network frequencies for the device requiring a subscription to identify available PLMNs and associated RATs that comprise the radio environment, and transmitting this information about the radio environment to the central cloud intelligence.

8. 2. The method of claim 1, wherein the receiving of information about the wireless environment of a device requiring a subscription is preceded by scanning network frequencies for a device companion of the device requiring a subscription to identify available PLMNs and associated RATs that comprise the wireless environment and transmitting this information about the wireless environment to the central cloud intelligence.

9. 10. A cloud infrastructure having a central cloud intelligence for implementing the method of claim 1, said cloud infrastructure having connectivity to collect supervision data relating to mobile network operator subscriptions and their associated favorite and forbidden PLMNs from each of a plurality of field-deployed Internet of Things devices and / or a plurality of Internet of Things lab devices hosting a sample of subscriptions; the training data relating to mobile network operator subscriptions and their associated favorite and forbidden PLMNs comprises some standard 3GPP information to be read from the mobile network operator profile in the embedded universal integrated circuit card; The standard 3GPP information includes at least: - Home PLMN, - equivalent PLMN, -Priority PLMN, - Forbidden PLMN Including, The cloud infrastructure further comprises a processing module, the processing module comprising: When the cloud infrastructure receives, from an Internet of Things device requiring a subscription, information about the wireless environment including a list of PLMNs recognized by the device, it extracts candidate subscriptions corresponding to the training data from within a database; filtering out from the candidate subscriptions candidate subscriptions of mobile network operators where the favorite PLMN of the device requiring a subscription does not appear in the wireless environment for processing the received information about the wireless environment; determining a subscription from among the candidate subscriptions that is adapted in response to this process; The cloud infrastructure then triggers the download of the determined subscription to the device requiring the subscription.

10. 10. An Internet of Things device having an embedded universal integrated circuit card into which candidate subscriptions are downloaded, the device having a cellular module or chipset supporting network scanning or being connectable to a device companion, the cellular module or chipset supporting network scanning or the device companion being adapted to scan network frequencies to identify available PLMNs and associated RATs that comprise the wireless environment, the device further having a transmit / receive cellular module or chipset for transmitting said information about the device's wireless environment to a central cloud intelligence as described in claim 9 and receiving the downloaded subscriptions.

11. 11. The Internet of Things device of claim 10, further comprising agent software within the embedded universal integrated circuit card for reading training data related to mobile network operator subscriptions and their associated favorite and forbidden PLMNs and transmitting the read training data to the central cloud intelligence.

Images