GB2499274A

GB2499274A - Determining a motion status of a user of a mobile device

Info

Publication number: GB2499274A
Application number: GB1213599.2A
Authority: GB
Inventors: Alexandr Vasilievich Garmonov; Yury Nikolaevich Pribytkov; Sergey Nikolaevich Moiseev; Andrey Yurievich Savinkov; Artem Aleksandrovich Khramov; Mahmoud Hadef; Nicolas Papachristos
Original assignee: Samsung Electronics Co Ltd
Current assignee: Samsung Electronics Co Ltd
Priority date: 2012-02-10
Filing date: 2012-07-31
Publication date: 2013-08-14
Anticipated expiration: 2032-07-31
Also published as: RU2012104711A; GB201213599D0; GB2499274B

Abstract

A motion state of a user is detected using at least one sensor of a mobile electronic device 100, the sensor being an electronic compass 110 arranged to output signals based on its orientation. A plurality of signal samples 120 are received from the electronic compass 110, and statistical processing 130 is performed on the plurality of signal samples to determine one of a plurality of motion states. The motion states may comprise at least: a stationary state, a walking state, a running state, and a driving state. The device may further be arranged to determine whether the plurality of signal samples are representative of oscillatory movement. The mobile electronic device 100 may be adapted for use in a wireless communication system.

Description

1

MOTION STATE DETECTION

Technical Field

The present invention relates to motion state detection using mobile devices.

5 More particularly, the present invention relates to a method for detecting a motion state of a user using at least one compass sensor of a mobile electronic device. The mobile electronic device may be adapted for use in a wireless communication system.

Background

10 Mobile devices, such as so-called mobile or smart phones, are an essential part of modern wireless communication systems. Their primary function is wireless communication, for which the mobile devices are arranged to wirelessly communicate with base stations in a wireless communication system. The wireless communication system may support one or more of the following standards: Global System for 15 Mobile Communications (GSM); Code Division Multiple Access (CDMA) and CDMA derivatives; any standard within the Institute of Electrical and Electronic Engineers (IEEE) 802.16 and 802.11 families; and 3rd Generation Partnership Project's Long Term Evolution (3GPP LTE). However, modern mobile devices are also often general purpose computing devices and, as well as voice or data 20 communications over the wireless communications system, are able to perform multiple ancillary functions. These functions include: wireless or wired communication with local devices and equipment; communication with satellite navigation systems such as the Global Navigation Satellite System (GLONASS) or the Global Positioning System (GPS); entertainment functions, such as playing music, 25 video and games; and personal alert functions, such as calendars and reminders.

The ancillary functions of modern mobile devices enable advanced functionality for a user. For example, location services, based on determining user location using global navigation systems, have become popular. For example, in US Patent Application 2007/0250261 Al, entitled "Motion classification methods for 30 personal navigation", publication date 25 October 2007, international classification G01C 21/00, a personal navigation system is disclosed. This system is composed of multiple sensors, some of which are attached to a user, and a computation unit. The

2

multiple sensors consist of one or more GPS sensors, one or more magnetic sensors, one or more pressure sensors or altimeters, one or more accelerometers, and one or more gyroscopes. The personal navigation system aims to classify the motion of a user - for example walking, crawling, sidestepping, etc. - so as to measure more 5 accurately measured a distance covered by a user. This personal navigation system requires a large number of sensors, including GPS sensors, and complex processing. This consumes much power and so quickly depletes a battery of a user's mobile device.

Similarly, US Patent Application 2005/0033515 Al, entitled "Wireless 10 personal tracking and navigation system", publication date 10 February 2005, discloses a method for user location based on pedometer attached to a user's leg and connected to an electronic mobile device compass. The pedometer counts number of steps and the compass measures the user's direction of motion. The current position of the user is computed based on these data.

15 In another approach, US Patent Application, 2009/0227271 Al, "Apparatus and methods using radio signals", publication date 10 September 2009, international classification H04L12/26 describes a method of locating a user based on a statistical analysis of radio signals from base stations in a wireless communication system. A similar technique is used in a paper by Ian Anderson and Henk Muller of the 20 University of Bristol, entitled "Context Awareness via GSM Signal Strength Fluctuation", published at the 4th International Conference on Pervasive Computing, ISBN 3-85403-207-2, pp. 27-31, in May 2006. This also uses base station signal measurements to infer contextual information such as a mode of travel. In the paper a neural network that is trained using the received signal power from a number of base 25 stations. Both of these approaches are dependent on base station signal levels, which may be corrupted by multiple uncontrolled factors, and requires time for learning statistical patterns.

As another example, US Patent Application, 2006/0187847 Al, entitled "Techniques for determining communication state using accelerometer data", 30 publication date 24 August 2006 in the name of Cisco Technology Inc., international classification H04W4/02, uses data from an accelerometer to determine a type of network communication for communicating with the user. The accelerometer can be

3

used to determine whether the mobile device is moving or stationary. When using an accelerometer there can be a problem of accidental motions. A solution to this problem is described in US Patent 7,763,842 B2, "Motion determination apparatus and method thereof', publication date 27 July 2010, international classification G05B 5 11/06.

In the art there are thus solutions that make use of ancillary sensors within a mobile device. These solutions are typically complex, requiring multiple sensors and satellite navigation system capabilities. There is thus a need for simple, easy to implement systems that can make creative use of sensors that are typically supplied 10 with a mobile device.

Summary

In accordance with a first aspect of the present invention, there is provided a method for determining a status of a user using a mobile device, the mobile device being 15 coupled to an electronic compass, the electronic compass arranged to output a signal based on its orientation, the method comprising receiving a plurality of signal samples from the electronic compass and performing statistical processing of said plurality of signal samples to determine one of a plurality of motion states representative of said samples.

20 In accordance with a second aspect of the present invention, there is provided a mobile device comprising an electronic compass for outputting a signal based on its orientation, a memory for storing a plurality of signal samples from the electronic compass and a computational module for performing statistical processing of said plurality of signal samples to output one of a plurality of motion states representative 25 of said samples.

Further features and advantages of the invention will become apparent from the following description of preferred embodiments of the invention, given by way of example only, which is made with reference to the accompanying drawings.

30 Brief Description of the Drawings

Figure 1 is a simplified schematic diagram showing an exemplary mobile device;

4

Figure 2 is a flow diagram showing a processing algorithm according to a first embodiment;

Figure 3 is a flow diagram showing a number of exemplary steps that may form part of a second embodiment;

5 Figure 4 is a time series chart showing typical data signals measured by a compass for four mobile device motion states, wherein Figures 4A to 4D separately show each of the data signals shown superimposed in Figure 4;

Figure 5 is a time series chart showing resultant values of a first exemplary decision function used to distinguish two mobile device motion states; 10 Figure 6 is a time series chart showing resultant values of a second exemplary decision function used to determine one out of four mobile device motion states;

Figure 7 is a time series chart showing resultant values of a third exemplary decision function used to separate four mobile device motion states into two groups;

Table 1 is a table showing exemplary algorithm parameter values for the first 15 exemplary decision function;

Table 2 is a table showing exemplary algorithm parameter values for the second exemplary decision function; and

Table 3 is a table showing exemplary algorithm parameter values for the third exemplary decision function.

20

Detailed Description

Certain embodiments of the present invention provide a method and apparatus to distinguish motion states of a user using a mobile device that forms part of a wireless communication system. In particular, certain embodiments of the present 25 invention analyse and process data measured by a compass embedded in a mobile device to determine a motion state.

Recognition of motion states of a user using a mobile device plays an important practical role in extending the functionality of such a mobile device. For example, according to certain embodiments described herein, possible motion states 30 that may be determined by a mobile device include: a stationary state, a walking state, a running state, and a driving state. Other states are envisaged, such as travelling on different modes of transport (e.g. bicycle, train, aeroplane, bus, etc.), carrying or

5

pushing an object (e.g. a small child, pushchair or wheelchair) and/or different movement types (e.g. going up stairs, in a lift etc).

Knowledge of these motion states provides a number of ways to extend mobile device functionality. For example, mobile device battery power can be saved by 5 switching off communication with satellite navigation systems such as GPS when the user is in a stationary state; user alerts in a user alert system can alert the user based on the motion state of the user, e.g. to inform the user that certain actions should be taken in a particular motion state; a volume adaptation system can be used to adapt the volume of the mobile device depending on whether the user is walking, driving or is 10 stationary; and a call forwarding system can be automatically activated to direct a voice call from the mobile phone to a hands-free device while in a driving motion state.

Figure 1 shows an exemplary mobile device 100 that may be used to implement a first embodiment of the present invention. Mobile device 100 comprises 15 a compass 110, a memory 120, a processing component 130, an operating system 140 and a display 150. Compass 110 may be an electronic magnetometer arranged to output an analogue or digital signal proportional to the direction of the mobile device 100 in relation to a frame of reference that is stationary relative to the surface of the earth, i.e. the orientation of the compass 110 and by extension the mobile device 100. 20 It is assumed that the compass 110 is statically arranged within the mobile device 100, for example it may comprise solid state electronics embedded within a chip on circuitry of the mobile device 100. In other embodiments, the compass may be an external device electrically coupled to the mobile device 100, for example using a universal serial bus (USB) or proprietary coupling. Analogue signals produced by the 25 compass 110 may be digitised and it will be assumed for the sake of explanation that at least digital signal representative of the orientation of the compass is available. This digital signal may comprise an 8, 16, 32 or 64-bit value representative of an orientation angle between 0 and 360 degrees.

In Figure 1, a digital signal output by the compass 110 is sampled at regular 30 intervals, i.e. periodically, and placed in memory 120. Memory 120 may comprise a buffer of a set size, for example a First In First Out (FIFO) or cyclical buffer, arranged to store a set number of samples, i.e. data from the compass 110. For example, the

6

buffer may have capacity for 50 samples at a 1Hz sampling rate; thus every second a sample of the digital signal produced by the compass 110 is stored in the buffer and a sample from 50 seconds ago is discarded. The buffer may store the data in the form of a numerical array.

5 In Figure 1, the memory 120 is accessible to a processing component 130. The processing component 130 is arranged to read the data from the memory 120 and determine a motion state of a user of the mobile device. For example, the processing component 130 may perform statistical processing on the data so as to determine whether the motion state is one of stationary, walking, running or driving. The 10 processing component 130 is arranged to use one or more signal processing algorithms that are capable in discriminating the movement type. The algorithms may be designed to filter the type of movement, taking into account the compass data variations. The processing component 130 may output a motion status notification to an operating system 140 or other control processing of the mobile device. The 15 notification may be used by one or more mobile applications being processed by the mobile device, which may change their state accordingly. A user may also be able to view a current motion state or a history of motion states by viewing the display 150.

Figure 2 shows an example of processing 200 that is performed by the processing component 130 according to a first embodiment. At step 210 compass data 20 is acquired. The data may comprise magnetic field strength vector angles. In certain embodiments, memory 120 holds a set number of samples and the contents of the memory are read at predetermined intervals so as to operate on a numeric data array. At step 220, there is a step of recognising a motion state of a user's mobile device using the compass data. Particular algorithms are described in more detail below with 25 reference to Figures 3 to 7. In these algorithms, statistical properties of the compass data are calculated. These statistical properties are then used as parameters in one or more discrimination functions that are applied to the compass data to determine a motion state. In certain embodiments a series of filters are applied to determine a specific motion state. For example, a first set of one or more filters may discriminate 30 between two or more groups of motion states, each group having one or more motion states; a second set of one or more filters may then perform a specific discrimination within the group. In the example of Figure 2, step 220 results in a discrimination

7

function value or weight for each of four states: a stationary state 230, a walking state 240, a running state 250, and a driving state 260. The discrimination function value may be a binary value for each state indicating whether the compass data is representative of that state, e.g. a value of 1 for the stationary state 230 may indicate 5 that the compass data is representative of a stationary motion state. Alternatively, the discrimination function value may comprise a probabilistic value, e.g. a value between 0 and 1 indicating a likelihood of a particular motion state. At step 270, the processing component makes a final decision on a particular motion state based on the results for each of motion states 230 to 260. For example, this may comprise taking 10 the highest likelihood value using maximum likelihood methods or may comprise performing additional filtering of compass data if a particular motion state has a particular value, e.g. additional discriminate processing if the motion state is not stationary. The output of the processing 200 is a motion state that may be notified to other applications within the mobile device.

15 A processing algorithm applied by a second embodiment of the present invention will now be described with regard to Figure 3. The steps of Figure 3 should be taken as one possible implementation and certain steps may be removed, moved or added depending on the implementation circumstances. Any of the steps of the second embodiment may also be used in the first embodiment and vice versa. 20 At step 302 data from the compass is periodically read by the mobile device.

At step 304, during a preset period, the data read from the compass is recorded to a file by software means, either directly or via memory 120. In certain embodiments, the data is saved in the form of a numeric data array. The preset period may comprise a particular sampling time period. The data recorded to the file will be referred to as 25 source data, which may be a numeric array of magnetic field strength vector angle values.

After a sampling period within which source data from the compass are obtained, a mean or median value for the source data may be obtained at step 306. A mean or median value may be obtained by filtering the source data; for example, 30 using a smoothing median filter of preset length, where in certain examples the preset length is equal to the sampling period and/or the number of samples in the buffer memory 120. The output of step 306 is referred to as auxiliary data. At step 308, the

8

auxiliary data are shifted by a preset value and then are subtracted from the source data. This may be equivalent to subtracting a mean or median value for a first set time period from each of a set of source data values for a second set time period. The two set time periods may be the same time period (i.e. shifted by a preset value of 0) or, 5 for certain cases, the two time period may be of different length and/or shifted in time. At step 310, the obtained data are squared and then smoothed by filtering by a filter with a preset length, for example, by means of a median filter. Steps 306 to 310 effectively calculate the variance of the source data with respect to a mean or median value, the mean or median value possibly relating to a subset of previous source data 10 samples. At step 312, a preset threshold value is subtracted from the obtained data. The preset threshold value is a discrimination threshold in the form of a constant value. The value may be optimized based on observed test data. In certain implementations, the obtained data are optionally non-linearly transformed at step 314 depending on the form of the data.

15 The output of steps 308 to 314 is a first decision value or statistic. This is the result of a first discrimination function that effectively compares a variance of the compass data with a discrimination threshold. At step 316 a test is performed on the first decision value. The first decision value is compared with zero and a preliminary decision about recognition of a mobile device state in one of two state groups is made. 20 If the first decision value is less than zero then at step 320 it is determined that compass data is representative of a first group and the user of a mobile device is in one of two motion states: a stationary state or a driving state. If the first decision value is greater or equal to zero, it is determined at step 318 that the compass data is representative of a second group and the user of a mobile device is in one of two 25 states: a walking state or a running state. As will be clear with reference to Figures 4A to 4C, this discrimination is made based on the knowledge that walking and running states generate compass data with a much greater variance than stationary or driving states; hence, if the variance is above a threshold value the former states are more likely.

30 At step 322, following a first or preliminary discrimination stage as represented by steps 316 to 320, a second or clarifying discrimination function is applied to the first decision value. The second discrimination function is selected

9

based on whether the source data was deemed representative of the first group or the second group. The second discrimination function outputs a second decision value that provides further in-group discrimination; i.e. in this example discriminates between a stationary and driving state in the first group and a walking and running 5 state in the second group. For the first group, the second discrimination function may be a multiplication, i.e. an amplification function. For the second group, the second discrimination function may be a non-linear transformation. The second decision value in each case may be compared with zero to make a binary distinction between the two states in each group, i.e. a value greater than or equal to zero may represent a 10 first motion state and a value less than zero may represent a second motion state. At step 324 a final decision about a motion state detected by the mobile device is made by using the output of the second discrimination function. The detected motion state can then be used to extend the functional capabilities of the mobile device.

As many mobile devices are now equipped with built in magnetometer 15 sensors, it is possible to measure angles of the magnetic field strength vector without changing mobile device hardware. Computation module software may be adapted to record measured compass data to one or more files. Data are then read from these one or more files and processed. Statistical processing is then performed on the data to determine a motion state of a user of the mobile device. For example, in the 20 embodiments above, data realizations are obtained for four mobile device states: stationary, driving, walking and running. The sampling rate and time interval between two adjacent samples are configurable and may be selected for particular implementations and compass specifications.

Figure 4 illustrates typical measured realisations of the compass data for all 25 four states for the sampling rate of 50 Hz. In this example, the compass data is a digital angle value between 0 and 360 degrees. Figure 4A shows representative compass data for a stationary state; Figure 4B shows representative compass data for a driving state; Figure 4C shows representative data for a walking state; and Figure 4D shows representative data for a running state. It is evident from the realisations that 30 the statistical characteristics of the compass data differ for all four states. The running state is characterized by high signal fluctuations, for the walking state the signal fluctuation is less, for the driving state the signal fluctuation is quite small and the

10

stationary state is characterised by the compass signal data matching background noise fluctuations. The motion state estimation algorithms are based on these typical differences.

A number of exemplary algorithms for calculating the decision values will

5 now be described. A first discrimination function may be based on the calculation:

a0=d* -h d*

A variance estimate value at time 1 can be calculated in the manner below:

< = me{(4_(J, _1)/2 -nt Y,...,(A_(TI _m_Ti +1"^_r2+lf}, = me{4 +1},

10

where 1 - the current time moment;

A

' - an angle measured by the compass at the time moment1;

T T

1 and 2 - two averaging intervals; and me{x, _ sampie estimate of the sampling median x' 'xt-i—'xt-T+i

15 The value A is a preset threshold value. Hence, the first decision function may use a d*

value based on the difference between the variance estimate value at time 1 and the preset threshold value h. The function me{} may be replaced in some embodiments with a function to calculate a sample estimate of the sampling mean.

In one case, the difference, ao, between the variance estimate value and the 20 preset threshold value may be used directly, wherein the value ao is used to perform a first discrimination into stationary and moving states: if ao is less than zero a stationary state is indicated and if ao is greater than or equal to zero a moving state is indicated.

In a second case, which is a preferred embodiment of the invention, the first 25 discrimination function may further comprise a non-linear transformation of the difference value a0 in certain cases, as described with regard to step 314 in Figure 3. For example, a first discrimination function to distinguish between two groups as described with regard to steps 316 to 320 above, i.e. to recognize the motion state as

11

one of the stationary or driving state and the walking or running state, may further comprise:

\ajh, a0< 0,

j(ao) = i

[ln(l + a0)/5, a0> 0,

i.e. if the difference value is less than or equal to zero, the function divides the 5 difference value ao by the constant h and if the difference value ao is greater than zero the function adds one to the difference value ao, takes the natural logarithm and divides the result by a constant, in this case five. Table 3 provides algorithm parameters for this case. If the first decision value, i.e. the result of the first decision function set out above, is greater than zero, the second group is selected, i.e. the group 10 comprising walking or running states. If the first decision value is less than or equal to zero, the first group is selected, i.e. the group comprising stationary or driving states. A robust second discrimination function to generate a second decision value a a — f* (CI ")

for state recognition has the following form: ' °, where ao is the difference value. The form of the second discrimination function and the algorithm

15 parameters (e.g. h, Ti, and T2) differ for different states.

For the first group described with regard to step 320 of Figure 3, to distinguish between a stationary state and a driving state, the second discrimination function may have a form as follows: j c ^ difference value ao is multiplied by a constant, in this case 2. Table 1 provides estimation algorithm parameters for this 20 case. The second decision value a produced by such a function is shown in Figure 5. If the second decision value is greater than zero, the motion state is determined to be a driving state. If the second decision value is less than or equal to zero, the motion state is determined to be a stationary state.

For the second group described with regard to step 318 of Figure 3, to 25 distinguish between a walking state and running state, the second discrimination function has the following form

\ajh, a0 < 0,

j(ao) = i

I ln(l + a0)/5, a0> 0,

12

i.e. the same form as the preferred version of the first discrimination function described above. However, for this case different algorithm parameters are used which are given in Table 2. The second decision value a produced by such a function is shown in Figure 6. If the second decision value is greater than zero, the motion state 5 is determined to be a running state. If the second decision statistics value is less than or equal to zero, the motion state is determined to be a walking state.

Figure 5 shows illustrative second decision values to distinguish between a stationary state and driving state. It is evident that the motion state recognition algorithm recognizes these states without errors as a second decision value for the 10 driving state is above zero, whereas decision values for the stationary state are below zero.

Figure 6 shows illustrative second decision values to distinguish between a walking state and running state. It is evident that the motion state recognition algorithm recognizes these states without errors as a second decision value for the 15 running state is above zero, whereas decision values for the walking state are below zero.

Figure 7 shows illustrative second decision values to distinguish between the two state groups comprising, in a first group, a stationary or driving state and, in a second group, the walking or running state. It is evident that the recognition algorithm 20 recognizes these states without errors as second decision values for the walking or running states are above zero, whereas second decision values of the stationary or driving state are below zero.

These examples demonstrate that compass data processing can provide additional information concerning motion states, which may not be available from 25 other sensors. No other sensors are needed to accurately discriminate the motion states. Embodiments of the present invention have the advantage of requiring a small amount of memory as it can successfully discriminate between states with a low number of samples. The filtering described herein to effect the statistical processing can be simply implemented without overloading an operating system, for example the 30 methods can be implemented in software and require few processor resources. In this manner, the described embodiment also results in a low additional power consumption that does not greatly drain device batteries.

13

Embodiments of the present invention provide an advantage over available solutions. Certain embodiments ensure higher recognition accuracy of at least four mobile device states in a wireless communication system, which therefore increases the efficiency of using mobile devices in wireless communication systems. This 5 advantage is achieved by the proposed analysis and processing procedure of data (angles of the magnetic field strength vector) measured by a mobile device with a built-in compass. The use of the mobile device with a built-in compass allows obtaining auxiliary data that cannot be obtained by other sensors and can be used to recognise a motion state. Mobile device applications can apply the recognised motion 10 state to provide additional benefits, namely: the obtained accurate motion state estimates in a wireless communication system allows performing various actions to extend the functionality of a mobile device. These include: saving mobile device battery power by turning off communication with global positioning systems when the user is in stationary still state; using motion state estimates to inform the user that 15 certain actions should be taken depending on the motion state; using motion state estimates to adapt the phone volume depending on whether the user is walking, driving or is stationary; and using motion state estimates for call forwarding from the mobile phone to a hands-free automatic audio playback device while driving, as well as extending other functional capabilities of the mobile device. 20 In certain cases a computer program product comprising a non-transitory computer-readable storage medium having computer readable instructions stored thereon may be used to implement at least a portion of the described examples. In this case, the computer readable instructions are executable by a computerized device to cause the computerized device to perform a method for determining a motion state of 25 a user of a mobile device, the method comprising receiving a plurality of signal samples from an electronic compass, the electronic compass being arranged to output a signal based on its orientation and performing statistical processing of said plurality of signal samples to determine one of a plurality of motion states representative of said samples.

30 The above embodiments are to be understood as illustrative examples of the invention. Further embodiments of the invention are envisaged. For example, the claimed invention can also be implemented on a user portable computer according to

14

the above algorithm. In this case, the portable computer may be coupled to an electronic compass device and may comprise a computation module or processor-module, which is available in most devices. It will be evident to that the specific functions described above may be varied depending on the implementation, for 5 example the algorithm parameters, function calculations and constants may be varied and/or optimized to improve performance. For example, functions other than natural logarithms may be used and the multiplication constant 2 and division constant 5 in the functions may be varied. When using threshold values it will be evident that variations can be made while maintaining functionality, i.e. a decision point other than 10 zero may be selected resulting in a different value for the preset value h. Likewise, the comparison "greater than or equal to" may be replaced with a "greater than" comparison depending on implementation and similar logic applies to "less than" comparisons. In certain embodiments, a handling filter may be used to discount intermittent movement of the mobile device generated by user handling. This filter 15 may use the change in direction or orientation of the compass between samples, as opposed to the absolute angle value, also compensating for angle changes around 0 and 360 degrees. Handling movements generally do not have oscillatory patterns, but are predominantly large swings in one direction. If the compass data does not demonstrate oscillatory patterns it may be dismissed as representative of a handling 20 movement. A handling filter may also test the data to see if a change in direction is balanced, i.e. that the changes in movement happen in both directions. If it does it is likely to be representative of a motion state rather than a handling motion. In certain implementations the determination of a motion state may be performed by one or more machine learning techniques such as hidden Markov model estimation or 25 support vector machines. Feature analysis may also be used to extract salient features from compass data samples. It is to be understood that any feature described in relation to any one embodiment may be used alone, or in combination with other features described, and may also be used in combination with one or more features of any other of the embodiments, or any combination of any other of the embodiments. 30 Furthermore, equivalents and modifications not described above may also be employed without departing from the scope of the invention, which is defined in the accompanying claims.

15

Claims

Claims

1. A method for determining a status of a user using a mobile device, the mobile device being coupled to an electronic compass, the electronic compass arranged to output a signal based on its orientation, the method comprising:

receiving a plurality of signal samples from the electronic compass; and performing statistical processing of said plurality of signal samples to determine one of a plurality of motion states representative of said samples.

2. The method of claim 1, wherein the statistical processing comprises: determining a variance of the plurality of signal samples;

applying a discrimination function to the variance;

using an output of the discrimination function to determine one of a plurality of motion states representative of said samples.

3. The method of claim 1, wherein the statistical processing comprises: determining a variance estimate for the plurality of signal samples; and comparing a discrimination value based on the variance to a threshold value, wherein if the discrimination value is less than the threshold value a motion state of stationary is selected and if the discrimination value is at least greater than the threshold value further statistical processing of the plurality of signal samples is performed.

4. The method of any one of claims 1 to 3, comprising:

determining whether the plurality of signal samples are representative of oscillatory movement; and performing statistical processing if it is determined that the plurality of signal samples are representative of oscillatory movement.

5. The method of claim 4, wherein the step of determining is performed based on changes between consecutive signal samples.

16

6. The method of claim 1, wherein the statistical processing comprises: determining a variance estimate for at least a subset of the plurality of signal samples;

applying a first decision function to the variance estimate, including 5 selectively applying a non-linear transformation to the variance estimate if the variance estimate is below a first threshold value; and making a motion state decision based the output of said first decision function.

7. The method of claim 6, wherein the step of determining a variance 10 estimate comprises:

using a first median filter to generate a first median estimate for at least a first subset of the plurality of signal samples;

subtracting the first median estimate from at least a second subset of the plurality of signal samples and squaring the result to generate a range of variance 15 values; and applying a second median filter to the range of variance values to generate a discrimination value.

8. The method of claim 6, wherein the step of making a motion state 20 decision comprises:

comparing the output of the first decision function to a second threshold value; applying a first version of a second decision function if the output of the first decision function is at least below the second threshold value, the first version of the second decision function comprising an amplification of the output of the first 25 decision function;

applying a second version of the second decision function if the output of the first decision function is at least above the second threshold value, the second version of the second decision function comprising a non-linear transformation of the output of the first decision function; and 30 comparing the output of the applied second decision function to a third threshold value to determine a motion state.

17

9. The method of any one of claims claim 1 to 8, wherein the plurality of motion states comprise at least: a stationary state, a walking state, a running state and a driving state.

10. A mobile device comprising:

an electronic compass for outputting a signal based on its orientation;

a memory for storing a plurality of signal samples from the electronic compass; and a computational module for performing statistical processing of said plurality of signal samples to output one of a plurality of motion states representative of said samples.

11. The mobile device of claim 10, comprising:

a wireless communications module for wirelessly communicating with a wireless communications network.

12. The mobile device of claim 10 or claim 11, comprising:

a discrimination function for application to a variance of the plurality of signal samples, wherein an output of the discrimination function is used to determine one of a plurality of motion states representative of said samples.

13. The mobile device of any one of claims 10 to 12, comprising:

a comparison function for application to a variance of the plurality of signal samples, wherein if the output of the comparison function is less than a threshold value a motion state of stationary is selected and if the output of the comparison function is at least greater than the threshold further statistical processing of the plurality of signal samples is performed by the computational module.

14. The mobile device of any one of claims 10 to 13, comprising:

a handling filter for determining whether the plurality of signal samples are representative of oscillatory movement, wherein statistical processing is performed by

18

the computational module if it is determined that the plurality of signal samples are representative of oscillatory movement.

15. The mobile device of claim 14, wherein the handling filter is arranged 5 to use changes between consecutive signal samples.

16. The mobile device of any one of claims 10 to 15, comprising:

a variance estimator for determining a variance estimate for at least a subset of the plurality of signal samples; and 10 a first decision function for application to the variance estimate, the first decision function including a non-linear transformation for selective application to the variance estimate if the variance estimate is below a first threshold value,

wherein a motion state decision is made by the computational module based the output of said first decision function.

15

17. The mobile device of claim 16, comprising:

a first median filter for generating a first median estimate for at least a first subset of the plurality of signal samples; and a second median filter for generating a discrimination value based on an 20 output of the variance estimator;

wherein the variance estimator is arranged to subtract the first median estimate from at least a second subset of the plurality of signal samples and square the result to generate the output of the variance estimator.

25 18. The mobile device of claim 16, comprising:

a first comparison function for comparing the output of the first decision function to a second threshold value;

a first version of a second decision function for application to the output of the first decision function if the output of the first decision function is at least below the 30 second threshold value, the first version of the second decision function comprising an amplification of the output of the first decision function;

19

a second version of the second decision function for application to the output of the first decision function if the output of the first decision function is at least above the second threshold value, the second version of the second decision function comprising a non-linear transformation of the output of the first decision function; and 5 a second comparison function for comparing the output of the applied second decision function to a third threshold value to determine a motion state.

•.'????.• INTELLECTUAL

*.*. .V PROPERTY OFFICE

20

Application No: GB 1213599.2 Examiner: Richard Kerslake

Claims searched: 1-18 Date of search: 15 November 2012

Patents Act 1977: Search Report under Section 17

Documents considered to be relevant:

Category

Relevant to claims

Identity of document and passage or figure of particular relevance

X

1,2,10 & 12

US2010/0063768 Al (MESMIC) See paragraphs 10 & 27-32

X

1,2,10 & 12

US2009/0248352 Al

(ALPS ELECTRIC) See paragraphs 13,34-38 & 49

X

1 & 10

US2010/0177037 Al

(SAMSUNG ELECTRONICS) See whole document, especially paragraphs 30-34

X

1 & 10

CN202013196 U

(SHENYANG SECURITY) See WPI abstract

Categories:

X

Document indicating lack of novelty or inventive

A

Document indicating technological background and/or state

step

of the art.

Y

Document indicating lack of inventive step if

P

Document published on or after the declared priority date but

combined with one or more other documents of

before the filing date of this invention.

same category.

&

Member of the same patent family

E

Patent document published on or after, but with priority date

earlier than, the filing date of this application.

Field of Search:

x

Search of GB, EP, WO & US patent documents classified in the following areas of the UKC :

Worldwide search of patent documents classified in the following areas of the IPC

G01C; H04W

The following online and other databases have been used in the preparation of this search report

EPODOC,WPI,TXTE

International Classification:

Subclass

Subgroup

Valid From

G01C

0017/28

01/01/2006

H04W

0004/02

01/01/2009

Intellectual Property Office is an operating name of the Patent Office www.ipo.gov.uk