阅读说明：本技术 针对位置感测对移动装置分组 (Grouping mobile devices for location sensing ) 是由 A·梅尔奎斯特 B·帕利延多 H·森德斯特伦 L·诺尔德 A·伊斯贝里 A·佩特夫 于 2018-01-12 设计创作，主要内容包括：提供了用于移动装置分组方法(10A、10B)。一种方法(10B),包括从多个移动装置(40A-40C)中的每个移动装置(40A-40C)接收(15B)控制数据(30A、30B),所述控制数据(30A、30B)指示在由相应移动装置(40A-40C)的传感器(43)监测的物理可观测量的测量值的时间序列中检测到的至少一个异常；基于对来自多个移动装置(40A-40C)的控制数据(30A、30B)指示的异常的比较,确定(17)多个移动装置(40A-40C)到至少一个位置感测组中的指派；根据所述至少一个位置感测组实施(20B)组传感器报告。(A grouping method (10A, 10B) for mobile devices is provided. A method (10B) comprising receiving (15B), from each mobile device (40A-40C) of a plurality of mobile devices (40A-40C), control data (30A, 30B), the control data (30A, 30B) being indicative of at least one anomaly detected in a time series of measured values of physical observables monitored by sensors (43) of the respective mobile device (40A-40C); determining (17) an assignment of the plurality of mobile devices (40A-40C) into at least one location sensing group based on a comparison of anomalies indicated by control data (30A, 30B) from the plurality of mobile devices (40A-40C); performing (20B) group sensor reporting according to the at least one position sensing group.)

1. A method, comprising the steps of:

-receiving (15B) control data (30A, 30B) from each mobile device (40A-40C) of a plurality of mobile devices (40A-40C), the control data (30A, 30B) being indicative of at least one anomaly detected in a time series of physically observable measurement values monitored by a sensor (43) of the respective mobile device (40A-40C),

-determining (17) an assignment of the plurality of mobile devices (40A-40C) into at least one location sensing group based on a comparison of anomalies indicated by the control data (30A, 30B) from the plurality of mobile devices (40A-40C), and

-performing (20B) group sensor reporting according to the at least one position sensing group.

2. The method of claim 1, wherein the first and second regions are selected from the group consisting of,

wherein the control data (30B) is indicative of at least one of:

-a timestamp (31) of the at least one anomaly, and

-a marker (34) associated with the at least one anomaly, the marker (34) being identified according to a respective detector model used by a respective mobile device (40A-40C) of the plurality of mobile devices (40A-40C) to detect anomalies in the time series of measurement values.

3. The method according to claim 1 or 2,

wherein the control data (30A) is indicative of at least one of:

-a portion (32) of the time series of measurement values comprising the at least one anomaly, and

-position information (33) of the respective mobile device (40A-40C) when the at least one anomaly occurs.

4. The method according to any one of the preceding claims,

wherein the physical observable is selected from the group consisting of: acceleration, position, rotation, sound pressure, temperature, pressure, brightness.

5. The method according to any one of the preceding claims, further comprising:

-comparing (16) anomalies of the plurality of mobile devices (40A-40C) based on a correlation model,

wherein at least one parameter of the correlation model is configured by a machine learning technique.

6. The method of claim 5, wherein the first and second light sources are selected from the group consisting of,

wherein the machine learning technique operates based on a time series of measurement values.

7. The method according to any one of the preceding claims, further comprising:

-verifying (18) the determined assignment based on reference control data not originating from sensors (43) of the plurality of mobile devices (40A-40C).

8. The method according to any one of claims 5 to 7,

wherein the machine learning technique further operates based on the baseline control data.

9. The method according to any one of the preceding claims, further comprising:

-receiving (15B) uplink training control data (30A) indicative of a time sequence of measurement values from at least one mobile device (40A-40C) of the plurality of mobile devices (40A-40C),

-based on the uplink training control data (30A): configuring (19) at least one parameter of a respective detector model used by the at least one of the plurality of mobile devices (40A-40C) to detect anomalies, and

-transmitting (11B) downlink control data comprising at least one parameter of the respective detector model to the at least one mobile device (40A-40C) of the plurality of mobile devices (40A-40C).

10. The method of claim 9, wherein configuring the at least one parameter of the respective detector model comprises:

-training a respective detector model used by the at least one mobile device (40A-40C) of the plurality of mobile devices (40A-40C) to detect anomalies.

11. A method of operating a mobile device (40A-40C), the method comprising the steps of:

-receiving (11A) downlink control data comprising at least one parameter of a detector model from a network node (50) of the network,

-detecting (12) at least one anomaly in a time series of physically observable measurement values monitored by sensors (43) of the mobile device (40A-40C) based on the detector model configured according to the at least one parameter, and

-sending (15A) control data (30A, 30B) indicative of the at least one anomaly to the network node (50).

12. The method of claim 11, further comprising

-implementing (20A) a group sensor reporting according to at least one position sensing group established by the control data (30A, 30B).

13. The method of claim 11 or 12, further comprising:

-selecting (13) between periodic reporting and aperiodic reporting for sending the control data (30A, 30B) depending on the significance of identifying the at least one anomaly.

14. The method according to any one of claims 11-13, further comprising:

-aggregating (14) a plurality of anomalies into a message of the control data (30A, 30B) according to a periodic reporting schedule.

15. A mobile device (40A-40C) comprising

-a sensor (43); and

-a processor (41), said sensor (41) being adapted to

-receiving (11A) downlink control data comprising at least one parameter of a detector model from a network node (50) of the network,

-detecting (12) at least one anomaly in a time series of physically observable measurements monitored by sensors (43) of the mobile device (40A-40C) based on a detector model configured according to the at least one parameter, and

-sending (15A) control data (30A, 30B) indicative of the at least one anomaly to the network node (50).

16. The mobile device (40A-40C) of claim 15,

wherein the processor (41) is further adapted to perform the method of any one of claims 11-14.

17. A network node (50) comprising

-a processor (51) adapted to

-receiving (15B) control data (30A, 30B) from each mobile device (40A-40C) of a plurality of mobile devices (40A-40C), the control data (30A, 30B) being indicative of at least one anomaly detected in a time series of physically observable measurement values monitored by a sensor (43) of the respective mobile device (40A-40C),

-determining (17) an assignment of the plurality of mobile devices (40A-40C) into at least one location sensing group based on a comparison of anomalies indicated by control data (30A, 30B) from the plurality of mobile devices (40A-40C), and

-performing (20B) group sensor reporting according to the at least one position sensing group.

18. The network node (50) of claim 17,

wherein the processor (51) is further adapted to perform the method of any one of claims 1 to 10.

19. A system, said system comprising

-a mobile device (40A-40C) according to claim 15 or claim 16, and

-a network node (50) according to claim 17 or claim 18.

Technical Field

Various embodiments of the present invention relate to methods for group sensor reporting and corresponding mobile device grouping, and to devices operating according to these methods. Various embodiments are particularly directed to methods and apparatus operable in a cellular network and related to the context of the internet of things.

Background

The cost and size of internet of things devices is rapidly decreasing. More items will be equipped with communication technology such as low power wide area network, LPWAN, wide area network, WAN or bluetooth low energy BLE. This will support new types of applications; for example, in the logistics industry, it would be possible to monitor that a single item, rather than a group of items, is in a container or loaded onto a truck.

However, as IoT devices become smaller and the size of the battery also becomes smaller, battery power will still be a limited resource. While WAN wireless communications such as cellular technology will continue to require a significant amount of energy in such devices, one solution to reduce battery consumption is to group clusters of closely adjacent IoT devices and treat these IoT devices as entities, so the burden of reporting sensed data over the network can be distributed among the devices in the cluster.

For example, in a mobile tracking application, the location is the same for all devices in close proximity. The grouping of mobile devices and associated group sensor reports may be used to share the reporting burden between mobile devices, or to increase the reporting frequency of the cluster as a whole to achieve better location granularity. Upon detecting that the mobile device has left the group, the device will revert back to reporting the sensory data as a separate unit.

Identifying groups or clusters of devices for group sensor reports is a known problem, and several solutions have been proposed.

For example, short-range communication techniques may be used to detect that devices are in close proximity. One drawback of this solution is the need to have devices communicate with each other.

Alternatively, statistical methods may be applied to the reported sensed data to infer that the devices are in close proximity, for example by comparing positioning information. This solution takes a long time if the devices report data independently of each other at a low frequency. This requires that many data points be collected before forming a cluster.

Disclosure of Invention

In view of the foregoing, there is a continuing need in the art for methods and apparatus that address some of the above-mentioned needs.

These basic objects of the invention are solved by the independent claims. Preferred embodiments of the invention are set forth in the dependent claims.

According to a first aspect, a method is provided. The method comprises the following steps: the method includes receiving control data from each of a plurality of mobile devices, the control data indicating at least one anomaly detected in a time series of physically observable measurements monitored by sensors of the respective mobile device, determining an assignment of the plurality of mobile devices into at least one location sensing group based on a comparison of anomalies indicated by the control data from the plurality of mobile devices, and implementing group sensor reporting in accordance with the at least one location sensing group.

Advantageously, grouping of mobile devices can be facilitated based on sensor data originating from any sensor, such as an accelerometer, pressure sensor, gyroscope, photodiode, temperature sensor, or microphone. Different sensors measure different physical observables.

Advantageously, the grouping of mobile devices may be based on events occurring as at least one anomaly indicated in respective control data received from different mobile devices without the need to receive many data points.

Advantageously, grouping of mobile devices can be facilitated even if a dedicated positioning sensor (e.g., global positioning system, GPS, sensor, etc.) is unavailable or temporarily not received. Accordingly, the device grouping based on the abnormal comparison may be more accurate and robust than the conventional device grouping, and the accuracy and robustness of the conventional device grouping may be improved.

Advantageously, implementing group sensor reporting in accordance with a determined location sensing group may reduce battery consumption of multiple mobile devices of the respective location sensing group, as these mobile devices may be considered entities.

The term "mobile device" as used herein may refer to a device that is capable of moving or being moved and that includes a wireless interface through which a communication technology, such as LPWA, WAN, or BLE, establishes and maintains connectivity to a wireless network, particularly to a cellular network. Examples of such mobile devices include smart phones, computers, Machine Type Communication (MTC) devices, and narrowband internet of things (NB-IOT) devices.

The term "wireless network" as used herein may refer to a communication network comprising wireless/radio links between network nodes in addition to fixed network links interconnecting functional entities of the wireless network infrastructure. Examples of such networks include universal mobile telecommunications system UMTS and third generation partnership project 3GPP long term evolution LTE cellular networks, new wireless NR 5G networks, long range wireless Lora, and the like. In general, various techniques of wireless networking may be applicable, and (LP) WAN connectivity may be provided.

The term "anomaly" as used herein is based on anomaly detection, a technique for identifying unusual patterns called anomalies or outliers that do not conform to the baseline behavior. For example, an anomaly may refer to an observation or event in a given dataset that does not conform to an expected pattern. It will be possible that the measurements associated with a given anomaly are significantly different from other measurements not associated with the given anomaly. For example, an anomaly may be a peak or a trough of measurements, e.g., of some statistical significance. In other examples, the anomaly may be defined by a particular pattern of peaks and/or troughs in the measurement, such as three consecutive peaks spaced no more than 100ms apart, or the like. It will be appreciated that the specific characteristics of the anomaly may vary from sensor to sensor. For example, it is contemplated that the pressure sensor may exhibit anomalies in the time series of measurements that are different from the gyroscope.

Different anomalies may show different characteristic behaviors, sometimes referred to as anomalous fingerprints. For example, the measurements may show different time dependencies for different anomalies. For example, a first anomaly may be associated with a fingerprint indicating "three consecutive peaks in measurement"; while a second anomaly may be associated with a fingerprint indicating "three consecutive valleys in measurement". Different exceptions may be flagged.

The term "time series" as used herein may refer to a series of measurements indexed in time order, and in particular measured at consecutive and equally spaced time instants, referred to as samples.

The term "physical observable" as used herein may refer to a physical quantity whose instantaneous value may be determined by measurement. Examples include: pressure, sound, brightness, acceleration, temperature, etc.

The term "sensor" as used herein may refer to a functional entity of a device for detecting an event or change in the environment of the device. The sensor may comprise an analog to digital converter.

For example, an accelerometer is a physically observable sensor that can be used to detect acceleration of a sensor and its device host relative to the device environment, in m/s²。

The term "location sensing group" as used herein may refer to a plurality of mobile devices that move or are moved together without being aware of each other and that may be collectively managed by a network due to their proximity to each other.

The term "group sensor reporting" as used herein may refer to techniques that allow multiple mobile devices of a location sensing group to report anomalies in their respective sensor data to infer a common location of the location sensing group. This may be achieved, for example, by coordinating sensor reporting of individual mobile devices of a location sensing group, to share the reporting burden among multiple mobile devices, or to increase the reporting frequency of the group as a whole to achieve better location granularity. It should be understood that various group sensor reporting assignments may be assigned to mobile devices in the location sensing group, such as temperature, humidity, location, and the like. A cluster head may be provided which may control or implement the sensor reporting. The cluster head function may be assigned to one mobile device or implemented in an application server.

According to some embodiments, the control data indicates at least one of a timestamp of the at least one anomaly and a marker associated with the at least one anomaly, the marker being identified according to a respective detector model used by a respective one of the plurality of mobile devices to detect the anomaly in the time series of measurement values.

Advantageously, comparing anomalies indicated by respective association flags may reduce the battery consumption of the respective mobile devices by transmitting only basic control data, and may reduce the power consumption of the receiving and data processing network nodes by simplifying the comparison itself.

The term "marker" as used herein may refer to an identifier representing at least one anomaly when detected using a detector model that may be preconfigured by the network node.

In particular, a flag may be assigned to the at least one anomaly if the at least one anomaly is detectable using a detector model configured by the network and thus represents a "known anomaly pattern". Different tags may correspond to different anomalies.

The flagged anomaly patterns may also be associated with location information, meaning that the detector model not only detects anomalies, but implicitly finds the current location of the mobile device.

Examples of markers include: roadblocks, left-handed turns, right-handed turns, highway entrances, highway exits, speed bumps, and the like.

It will be appreciated that the data size of the indicia may be significantly smaller than the data size of the measurement comprising the at least one anomaly. This helps to reduce the required bandwidth.

For example, if at least one anomaly is identified as having a high significance, e.g. in relation to a given significance threshold, the at least one anomaly may be indicated by a short marker in the respective control data sent to the network node, rather than by a large part of the time sequence.

The term "significance" as used herein may refer to the certainty that at least one anomaly is identified by a detector model configured by the network. For example, an identification significance of 0% may indicate that a detector model configured by the network is unavailable or has been configured based on an anomaly other than the at least one anomaly. Conversely, a recognition significance of 100% may indicate that the at least one anomaly is encountered again by the network-configured detector model after the detector model has been configured based on the at least one anomaly. Due to the analog nature of the physical observables monitored, the identification significance can be less than 100%.

The term "detector model" as used herein may refer to a model built from sample data that enables anomaly detection in a time series of measurement values. For example, a simple statistical detector model may involve a multiple of the moving average of the time series as a threshold to determine outliers or anomalies in the time series. More complex detector models may, for example, involve machine learning, particularly machine learning based on artificial neural networks.

According to some embodiments, the control data is indicative of at least one of a portion of a time series of measurements including the at least one anomaly and location information of the respective mobile device at the time of occurrence of the at least one anomaly.

Such an implementation of control data may be helpful in situations where it is not possible to reliably detect anomalies at each individual mobile device. For example, the significance of a given mobile device detecting a given anomaly may be limited. Then, based on the measurement values obtained in the control data from the plurality of mobile devices, more reliable abnormality detection can be collectively performed, for example, by correlation between various measurement values.

Furthermore, such an implementation of control data may be helpful in cases where it is not easy to classify each anomaly into a given label, for example due to complexity. Uncertainty can then be avoided by providing measurements. Also, a priori knowledge about the type of anomaly may not be available.

Furthermore, such an implementation of control data may be helpful in cases where the detector model used to detect the anomaly has not been properly trained.

Advantageously, using the measurements to compare anomalies indicated by control data from the plurality of mobile devices facilitates assigning the plurality of mobile devices into a location sensing group when a large base of sensor data is not yet available and/or if no observed anomalies are yet available.

Based on the time series portion of the measurements, it may be possible to train a correlation model. This may help to identify whether certain anomalies are in principle suitable for use as decision criteria in the grouping of devices.

The term "training" as used herein may generally refer to the process of inferring functionality, such as decision-making functionality, from data collected in the past. In particular in the machine learning context, training may involve supervised learning based on a set of training examples consisting of input values or vectors and expected output values, or unsupervised learning based on training examples in which control data from a plurality of mobile devices is used as input and the result of a comparison of anomalies indicated by the control data from the plurality of mobile devices is used as an expected output value.

The term "machine learning" as used herein may refer to computational methods for data-driven learning and decision-making, and does not involve any data-specific programming.

The term "timestamp" as used herein may refer to timing information for a portion of a time series within the time series, and/or timing information relative to absolute time. For example, the timestamp may represent a start time and/or an end time of a portion of the time series that includes the at least one anomaly. A common time reference may be used for multiple devices.

The term "part of the time series" as used herein may refer to a part of the time series having no gaps or gaps, but in any case comprises those measurements indicative of at least one anomaly.

The term "location information" as used herein may specify information defining a particular geographic location. For example, the location information may include latitude and longitude information, optionally altitude information, and may be expressed, for example, as a decimal, as a degree-minute-second, or any other representation. The location information may represent the last known access point or cell of the wireless or cellular network, a segment of a cell, or the location of the mobile device itself.

According to some embodiments, the physical observable is selected from the group consisting of; acceleration, position, rotation, sound pressure, temperature, pressure, brightness.

According to some embodiments, the method further comprises: the anomalies of the plurality of mobile devices are compared based on the correlation model. At least one parameter of the correlation model is configured by a machine learning technique.

Advantageously, machine learning may allow for continuous adjustment and improvement of device groupings as more sensor data is captured in the field system. For example, as indicated above, the correlation model may be trained based on measurements received with the control data.

Advantageously, machine learning may allow data-driven learning and decision-making without involving any data-specific programming.

Advantageously, machine learning may allow for a reduction in reporting frequency of mobile devices and/or an increase in cluster granularity by inferring which anomalies are relevant or important for device grouping based on a comparison of anomalies.

The term "correlation model" as used herein may refer to any model that is capable of correlating anomalies, such as a signature or portion based on a corresponding time series of measurements. For example, a simple correlation model may involve a cross-correlation that is a measure of similarity of two parts of different time series based on their respective timestamps aligned with each other. More complex correlation models may for example relate to machine learning, in particular machine learning based on artificial neural networks.

According to some embodiments, the machine learning technique operates based on a time series of measurement values. A part of which may be indicated by control data.

According to some embodiments, the method further comprises: the determined assignment is validated based on baseline control data that does not originate from sensors of the plurality of mobile devices.

Advantageously, this enables appropriate action to be identified and taken if the position sensing set deviates from expectations.

The term "baseline control data" as used herein may refer to external data, such as a package list or an order database, that reflects one or more expected group assignments and against which a determined location sensing group may be compared.

According to some embodiments, the machine learning technique also operates based on baseline control data.

Advantageously, the desired output value is provided by the reference control data, which may facilitate machine learning based on a training example consisting of the input value or vector and the desired output value.

According to some embodiments, the method further comprises: receiving uplink training control data indicative of a time sequence of measurement values from at least one of the plurality of mobile devices; based on the uplink training control data: configuring at least one parameter of a respective detector model used by the at least one of the plurality of mobile devices to detect anomalies, and sending downlink control data including the at least one parameter of the respective detector model to the at least one of the plurality of mobile devices.

Additionally, the configuration may be based on machine learning techniques.

The term "uplink" as used herein may refer to the direction of communication from a terminal device, in particular a mobile device, to a network, in particular a wireless network.

Advantageously, based on the comparison of the respective up-link training control data and the anomalies indicated by the up-link training control data from the plurality of mobile devices, the respective detector models may be configured and may be further refined as more sensor data is captured in the field system. This may help to detect anomalies more reliably. In addition, new types of anomalies may be trained. A corresponding label may be assigned.

The term "downlink" as used herein may refer to the direction of communication from a network, in particular a wireless network, to a terminal device, in particular a mobile device.

According to some embodiments, configuring at least one parameter of a respective detector model comprises: training a respective detector model used by the at least one of the plurality of mobile devices to detect anomalies.

Advantageously, training the respective detector models may allow data-driven learning and decision-making without involving any data-specific programming.

According to a second aspect, there is provided a method of operating a mobile device, the method comprising the steps of: receiving downlink control data comprising at least one parameter of a detector model from a network node of a network, detecting at least one anomaly in a time sequence of physically observable measurements monitored by sensors of the mobile device based on the detector model configured according to the at least one parameter, and sending control data indicative of the at least one anomaly to the network node.

Advantageously, detecting at least one anomaly based on the detector model may reduce battery consumption of the respective mobile device by transmitting only the basic control data. The control signaling overhead is reduced. If the marked anomaly already has associated location information, battery consumption may be further reduced, as the mobile device is not required to run any positioning method to find the current location.

The term "network node" as used herein may refer to a cloud server infrastructure that provides services (e.g., packets of mobile devices) via available WAN connectivity. The cloud server infrastructure may be implemented by server hardware/software and/or distributed processing. The network node may be part of a wireless network or a data network, such as the internet.

According to some embodiments, the method further comprises implementing group sensor reporting according to at least one position sensing group established by the control data.

Advantageously, implementing group sensor reporting according to determined location sensing groups may reduce battery consumption of a plurality of mobile devices of the respective location sensing groups, as these mobile devices may be considered entities. A cluster head may be available. The grouped sensor reports may be shared among the grouped devices.

According to some embodiments, the method further comprises: selecting between periodic reporting and aperiodic reporting for sending the control data depending on the significance of identifying the at least one anomaly.

Advantageously, this may speed up the grouping of mobile devices or reduce battery consumption of the respective mobile devices in response to the availability of new sensor data.

For example, if the at least one anomaly is identified as having a high significance, e.g. in relation to a first given significance threshold, the corresponding control data may be immediately sent to the network node, i.e. non-periodically reported, e.g. to improve the positioning accuracy of the existing position sensing group.

Alternatively or additionally, if at least one anomaly is identified as having low significance, for example in relation to a second given significance threshold, aperiodic reporting may be appropriate. In this case, the at least one anomaly may not have been encountered by the network-configured detector model, and the corresponding control data may also facilitate grouping of mobile devices into a location detection group.

Aperiodic reports may rely on dedicated resources. Here, the uplink scheduling request and the downlink scheduling grant may be transmitted in response to a need for an aperiodic report to obtain dedicated resources.

In all other cases, or when reducing battery consumption is an important consideration, periodic reporting may be appropriate. The periodic reporting may utilize pre-scheduled resources. For example, semi-persistently scheduled resources that reoccur with a particular time pattern/periodic reporting schedule may be used for periodic reporting. Dedicated resources may not be required.

According to some embodiments, the method further comprises: aggregating a plurality of anomalies into a message of the control data according to a periodic reporting schedule.

Advantageously, this may preserve battery resources of the respective mobile devices by transmitting detected anomalies less frequently due to the transmission overhead per transmission.

According to a third aspect, there is provided a mobile device comprising a sensor; and a processor, the sensor being adapted to receive downlink control data from a network node of the network including at least one parameter of a detector model, detect at least one anomaly in a time series of measured values of physical observables monitored by the sensor of the mobile device based on the detector model configured according to the at least one parameter, send control data indicative of the at least one anomaly to the network node (50).

The term "wireless interface" as used herein may refer to a functional entity of a device for providing wireless connectivity to a corresponding wireless communication network.

The term "processor" as used herein may refer to a functional entity of an apparatus for performing the method steps provided in a memory of the apparatus.

According to some embodiments, the processor is further adapted to perform the methods of the various embodiments.

Advantageously, the technical effects and advantages described above in connection with the method according to the second aspect are equally applicable to a mobile device having corresponding features.

According to a fourth aspect, there is provided a network node comprising a processor adapted to receive control data from each of a plurality of mobile devices, the control data being indicative of at least one anomaly detected in a time series of physically observable measurements monitored by sensors of the respective mobile devices, determine an assignment of the plurality of mobile devices into at least one location sensing group based on a comparison of anomalies indicated by the control data from the plurality of mobile devices, and implement group sensor reporting in accordance with the at least one location sensing group.

The term "network interface" as used herein may refer to a functional entity of a device for providing network connectivity to a corresponding communication network.

According to some embodiments, the processor is further adapted to perform the methods of the various embodiments.

Advantageously, the technical effects and advantages described above in connection with the method according to the first aspect are equally applicable to a network node having corresponding features.

According to a fifth aspect, a system is provided. The system includes various embodiments of a mobile device and various embodiments of a network node.

Drawings

Embodiments of the present invention will be described with reference to the accompanying drawings, wherein like or similar reference numerals refer to like or similar elements.

Fig. 1 is a schematic diagram illustrating a method according to an embodiment.

Fig. 2 is a diagram illustrating uplink training control data transferred in a method according to an embodiment.

Fig. 3 is a schematic diagram illustrating a modification of the method according to the embodiment.

Fig. 4 is a schematic diagram illustrating a further variation of the method according to an embodiment.

Fig. 5 is a schematic diagram illustrating control data transmitted in a method according to an embodiment.

Fig. 6 is a schematic diagram for illustrating a mobile device according to an embodiment.

Fig. 7 is a schematic diagram illustrating a network node according to an embodiment.

Detailed Description

Exemplary embodiments of the present invention will now be described with reference to the accompanying drawings. Although some embodiments will be described in the context of a particular field of application, the embodiments are not limited to that field of application. Furthermore, features of the various embodiments may be combined with each other unless specifically stated to the contrary.

The figures are to be regarded as schematic representations and the elements shown in the figures are not necessarily shown to scale. Rather, the various elements are represented such that their function and general purpose will become apparent to those skilled in the art.

Fig. 1 is a schematic diagram illustrating methods 10A, 10B according to an embodiment.

These embodiments implement the grouping of mobile devices 40A-40C based on control data 30A indicating that an anomaly of 12 is detected without using a detector model configured by the network.

The method 10A shown on the left side of fig. 1 is for operating a mobile device 40A-40C of the plurality of mobile devices 40A-40C, while the method 10B shown on the right side of fig. 1 is for operating a network node 50.

According to the method 10A, each mobile device 40A-40C of the plurality of mobile devices 40A-40C includes a respective sensor 43, which may be a low cost sensor such as an accelerometer, microphone, or the like. Each sensor 43 monitors a respective physical observable captured in a respective time series of measurements. The corresponding physical observable may be acceleration, position, rotation, sound pressure, temperature, pressure, brightness, etc.

The different mobile devices 40A-40C may include corresponding sensors. In some cases, each mobile device 40A-40C includes more than a single sensor.

The respective mobile devices 40A-40C may individually detect 12 at least one anomaly in the respective time series of measurement values.

If so, in the example of fig. 1, the respective mobile device 40A-40C sends 15A control data 30A indicating at least one anomaly to the network node 50. As shown in fig. 1, the transmitting step 15A of method 10A performs the transmission of control data 30A by the respective mobile device 40A-40C, while the receiving step 15B of method 10B performs the corresponding reception of control data 30A by the network node 50.

Initially, a default detector model, e.g. a simple statistical detector model, may be assumed, so that the detection 12 may typically be performed without any assistance of the detector model configured by the network. Accordingly, the respective mobile device 40A-40C sends 15A respective uplink control data 30A to the network node 50 indicating at least one anomaly, see fig. 2. Here, the uplink control data 30A includes an associated part of a time series of the time stamp and the measurement value. Optionally, the uplink control data 30A includes the measured position.

This part of the time series of measurement values is included in the control data 30A, since typically an untrained detector model is rather unreliable.

Referring again to fig. 1, when transmitting 15A uplink control data, the respective mobile device 40A-40C may select 13 between periodic and aperiodic reporting of the control data 30A depending on the significance of identifying the at least one anomaly.

As previously mentioned, it is assumed that the detection 12 does not rely on a detector model configured according to at least one parameter received from the network node 50. Thus, the at least one anomaly is not identified as a "known anomaly pattern" and the aperiodic report is selected to provide the control data 30A to the network node 50 as soon as possible to take into account the at least one anomaly when creating the detector model of the network configuration.

According to the method 10B, the network node 50 receives 15B respective control data 30A from each mobile device 40A-40C of the plurality of mobile devices 40A-40C.

The network node 50 may then compare 16 the anomalies of the plurality of mobile devices 40A-40C based on the correlation model. At least one parameter of the correlation model is determined by a machine learning technique that can operate based on a time series of measurement values and can also operate based on reference control data.

The network node 50 then determines 17 an assignment of the plurality of mobile devices 40A-40C into at least one location sensing group based on a comparison 16 of the anomalies indicated by the respective control data 30A from the plurality of mobile devices 40A-40C.

The network node 50 may then verify 18 the determined group assignment based on baseline control data not originating from the sensors 43 of the plurality of mobile devices 40A-40C, which baseline control data reflects one or more expected group assignments, such as a list of packages or an order database.

The network node 50 then implements 20B group sensor reporting according to the at least one location sensing group. For example, this may involve assigning and communicating a respective reporting frequency to each mobile device 40A-40C according to a respective one of the at least one location sensing sets.

According to the method 10A, each of the plurality of mobile devices 40A-40C may also implement 20A set of sensor reports based on at least one location sensing set established according to the control data 30A. For example, this may include receiving and applying, by each mobile device 40A-40C, a respective reporting frequency from a respective one of the at least one location sensing sets.

Fig. 2 is a schematic diagram illustrating uplink control data 30A communicated in the methods 10A, 10B according to an embodiment.

The uplink control data 30A indicates at least one of a timestamp 31 of the at least one anomaly, a portion 32 of the time series that includes a measurement of the at least one anomaly, and location information 33 of the respective mobile device 40A-40C at the time of the occurrence of the at least one anomaly.

Assigning the plurality of mobile devices 40A-40C to a particular location sensing group may require that at least one mobile device 40A-40C of the plurality of mobile devices 40A-40C has provided its location information 33 in uplink control data 30A sent to and received by the network node 50.

Fig. 3 is a schematic diagram illustrating a variation of the methods 10A, 10B according to an embodiment.

These embodiments implement machine learning techniques for creating a respective detector model for use by at least one mobile device 40A-40C of the plurality of mobile devices 40A-40C to detect at least one anomaly.

According to the method 10B, the network node 50 receives 15B uplink training control data 99A from at least one mobile device 40A-40C of the plurality of mobile devices 40A-40C, the uplink training control data 99A indicating a time sequence of measurement values, i.e. a sequence of measurement values indexed in time order as described above. Here, the mobile devices 40A-40C are not generally required to have identified any anomalies in the time series of measurements. For example, there may not be a detector model available at the mobile devices 40A-40C.

The network node 50 then configures 19 at least one parameter of a respective detector model used by at least one mobile device 40A-40C of the plurality of mobile devices 40A-40C to detect the at least one anomaly based on the uplink training control data 99A. The configuration step 19 may additionally be based on machine learning techniques.

Configuring 19 at least one parameter of a respective detector model may include training the respective detector model for use by at least one mobile apparatus 40A-40C of the plurality of mobile apparatuses 40A-40C to detect anomalies. For example, the training may involve unsupervised learning based on a training example consisting of input values or vectors and expected output values, where uplink training control data from multiple mobile devices is used as input.

The network node 50 then transmits 11B downlink control data 99B comprising at least one parameter of the respective detector model to at least one mobile device 40A-40C of the plurality of mobile devices 40A-40C.

Fig. 4 is a schematic diagram illustrating further variations of the methods 10A, 10B according to embodiments.

The embodiments implement the grouping of mobile devices 40A-40C based on control data 30A indicating an anomaly detected 12 using a detector model of a network configuration.

The same reference numerals as in fig. 2 denote the same elements and need not be further mentioned.

According to the method 10A, the respective mobile device 40A-40C receives 11A downlink control data 99B in response to the mode 11B of the network node 50. The downlink control data comprises at least one parameter of the respective detector model. For example, the detector model may be trained using the received uplink training control data 99A. The detector model is typically composed of algorithms/methods and parameters. A very basic example would be that training finds that a linear regression y — B0+ B1 x can be used. The model then takes B0 and B1 as parameters. Y can then be predicted by providing x. Additionally, the algorithm/method of the detector model may be able to be updated.

The respective mobile device 40A-40C then detects 12 at least one anomaly in the time series of physically observable measurements monitored by the sensors 43 of the mobile device 40A-40C based on the detector model configured according to the at least one parameter.

The respective mobile device 40A-40C may then select 13 between periodic reporting and aperiodic reporting for transmission of said control data 30A, 30B depending on the significance of identifying said at least one anomaly.

According to the selected periodic reporting schedule, the respective mobile devices 40A-40C may aggregate 14 the plurality of anomalies into a message of the control data 30B.

The respective mobile device 40A-40C then sends 15A control data 30A, 30B indicating the at least one anomaly to the network node 50. For example, the same mobile device 40A-40C of the plurality of mobile devices 40A-40C may selectively transmit control data 30A or control data 30B indicative of the at least one anomaly, depending on a need for a respective significance of the identification of the potential at least one anomaly.

In particular, less battery resources may be required to transmit 15A control data 30B including a marker 34 for a "known abnormal pattern" than to transmit 15A uplink control data 30A including a portion 32 indicating a time sequence of measurement values and location information 33.

In the example of fig. 4, control data 30B is transmitted.

According to the method 10B, the network node 50 receives 15B respective control data 30B from each mobile device 40A-40C of the plurality of mobile devices 40A-40C. In general, some mobile devices 40A-40C may transmit control data 30A; while other mobile devices 40A-40C may transmit control data 30B.

From here, the same method sequence as in fig. 2 can be executed based on the control data 30A or the control data 30B.

In particular, comparing 16 anomalies of multiple mobile devices 40A-40C may be performed between tags 34 having the same or similar timestamps 31, and between portions of the time series 32 having the same or similar timestamps 31.

Fig. 5 is a schematic diagram illustrating control data 30B communicated in methods 10A, 10B according to an embodiment.

The control data 30B indicates at least one of a timestamp 31 of the at least one anomaly, and a marker 34 associated with the at least one anomaly. The markers 34 are identified according to respective detector models used by respective ones of the plurality of mobile devices 40A-40C to detect anomalies in the time series of measurement values.

As will be appreciated, control data 30B has a reduced size compared to control data 30A.

FIG. 6 is a schematic diagram illustrating a mobile device 40A-40C according to an embodiment.

The mobile devices 40A-40C include a processor 41, a wireless interface 42, and sensors 43.

The processor 41 and the wireless interface 42 are adapted to receive downlink control data 11A comprising at least one parameter of the detector model from a network node 50 of the network.

The processor 41 is adapted to detect 12 at least one anomaly in a time series of physically observable measurements monitored by the sensors 43 of the mobile devices 40A-40C based on a detector model configured according to at least one parameter.

Additionally, the sensors 43 may include position estimation capabilities that generate position information.

The processor 41 and the wireless interface 42 are further adapted to send 15A control data 30A, 30B indicating at least one anomaly to the network node 50.

The processor 41 is also adapted to perform a method 10A of operating the mobile device 40A-40C according to various embodiments.

Fig. 7 is a schematic diagram for illustrating a network node 50 according to an embodiment.

The network node 50 comprises a processor 51 and a network interface 52.

The processor 51 and the network interface 52 are adapted to receive 15B control data 30A, 30B from each mobile device 40A-40C of the plurality of mobile devices 40A-40C, the control data 30A, 30B being indicative of at least one anomaly detected in a time series of physically observable measurements monitored by the sensors 43 of the respective mobile devices 40A-40C.

The processor 51 is adapted to determine 17 an assignment of the plurality of mobile devices 40A-40C into the at least one location sensing group based on a comparison of anomalies indicated by the control data 30A, 30B from the plurality of mobile devices 40A-40C.

The processor 51 is further adapted to implement 20B group sensor reporting according to the at least one location sensing group and to perform the method 10B of operating the network node 50 according to various embodiments.