Apparatus, systems and methods for fall-down detection based on a wireless signal

• US 10,397,039 B2
• Filed: 11/28/2018
• Issued: 08/27/2019
• Est. Priority Date: 12/05/2012
• Status: Active Grant

• Current Assignee: Origin Wireless, Inc.
• Original Assignee: Origin Wireless, Inc.
• Inventors: Zhang, Feng, Chen, Chen, Xu, Qinyi, Wang, Beibei, Wu, Chenshu, Zhang, Hangfang, Wong, Chau-Wai, Claffey, David N., Chen, Chun-I, Lai, Hung-Quoc Duc, Wu, Zhung-Han, Wu, Min, Han, Yi, Au, Oscar Chi-Lim, Liu, K. J. Ray
• Primary Examiner(s): Lee, Siu M
• Application Number:
  US16/203,317
• Publication Number:
  US 20190158340A1
• Time in Patent Office: 272 Days
• CPC Class Codes: H04B 1/38 Transceivers, i.e. devices in which transmitter and receiver form a structural unit and in which at least one part is used for functions of transmitting and receiving
• CPC Class Codes: H04B 1/385 Transceivers carried on the body, e.g. in helmets
• CPC Class Codes: H04B 17/309 Measuring or estimating channel quality parameters
• CPC Class Codes: H04B 17/318 Received signal strength
• CPC Class Codes: H04B 17/364 Delay profiles
• CPC Class Codes: H04L 27/362 Modulation using more than one carrier, e.g. with quadrature carriers, separately amplitude modulated H04L27/366 takes precedence
• CPC Class Codes: H04W 72/21 in the uplink direction of a wireless link, i.e. towards the network

Abstract

Apparatus, systems and methods for detecting and monitoring fall-down actions and other transient motions are disclosed. In one example, a system for monitoring a transient motion in a venue is disclosed. The system comprises a transmitter, a receiver, and a transient motion monitor. The transmitter is located at a first position in the venue for transmitting a wireless signal through a wireless multipath channel impacted by the transient motion of an object in the venue. The receiver is located at a second position in the venue for: receiving the wireless signal through the wireless multipath channel, and obtaining a time series of channel information (CI) of the wireless multipath channel based on the wireless signal. The transient motion monitor is configured for: monitoring the transient motion based on the time series of CI, and triggering a response action based on the monitored transient motion.

Claims

1. A system having a processor, a memory communicatively coupled with the processor and a set of instructions stored in the memory for monitoring a transient motion in a venue, comprising:

a transmitter located at a first position in the venue and configured for transmitting a wireless signal through a wireless multipath channel impacted by the transient motion of an object in the venue;

a receiver located at a second position in the venue and configured for;

receiving the wireless signal through the wireless multipath channel impacted by the transient motion of the object in the venue, and

obtaining a time series of channel information (CI) of the wireless multipath channel based on the wireless signal; and

a transient motion monitor configured for;

monitoring the transient motion of the object based on the time series of CI, and

triggering a response action based on the monitored transient motion of the object.

2. The system of claim 1, wherein the transient motion monitor is further configured for:

determining a sliding time window having a duration comparable to the transient motion of the object; and

computing an intermediate quantity (IQ) based on CI of the time series in the sliding time window, wherein the transient motion of the object is monitored based on the IQ.

3. The system of claim 2, wherein computing the IQ based on the CI comprises:

computing an auto function (AF) based on the CI of the time series in the sliding time window;

computing an IQ-related feature of the AF; and

computing the IQ based on the IQ-related feature of the AF, wherein the AF comprises at least one of;

auto-correlation function, auto-covariance function, cross-correlation function, cross-covariance function, differential function, differential auto-correlation, differential auto-covariance, differential cross-correlation, differential cross-covariance, a frequency transform, frequency spectrum, and a power spectral density.

4. The system of claim 3, wherein computing the AF comprises:

transforming each CI of the time series in the sliding time window; and

computing the AF of each transformed CI in the sliding time window, wherein transforming each CI comprises performing at least one of the following on the CI;

a scalar operation, a magnitude, a phase, a nonlinear map, a linear map, a piecewise linear map, a function, a linear function, a nonlinear function, a piecewise linear function, a real function, a complex function, a vector-valued function, an inverse function, a derivative of a function, an integration of a function, a function of another function, a one-to-one function, a one-to-many function, a many-to-one function, a many-to-many function, a logarithmic function, an exponential function, a trigonometric function, a circular function, an algebraic function, a transcendental function, a power function, a root function, a polynomial, a square, a cube, a sine, a cosine, a tangent, a cotangent, a secant, a cosecant, an elliptical function, a parabolic function, a hyperbolic function, a gamma function, a zeta function, an absolute value, a thresholding, a quantization, a piecewise constant function, a piecewise linear function, a table look-up,

a set operation, a histogram, a combination, a permutation, a subset, an inclusion, a logical operation, an OR, an ADD, an XOR, a union, an intersection, a set difference, a symmetric difference a Cartesian product, a power set, a cardinality, an average, a weighted average, a mean, a variance, a range, a trimmed mean, a statistics, a percentile,

a vector operation, a norm, a metric, a vector addition, a vector subtraction, a vector multiplication, a vector division, a dot product, an inner product, an outer product, an arithmetic operation, a mathematical operation, a subspace, a projection, a decomposition, a vector transformation, a change of basis, a rotation, a translation, a vector function, a projection, a principal component, a kernel principal component, an independent component, a neighborhood component, a connected component,

a linear transform, a nonlinear transform, Fourier transform, Laplace transform, Hilbert transform, Sine transform, Cosine transform, Wavelet transform, Hadamard transform, Slant transform, Fast Fourier transform (FFT), discrete cosine transform (DCT), frequency transform, frequency decomposition, inverse transform, time transform, time decomposition, a frequency, a spectrum, a period, a time,

a matrix operation, a rank, a transpose, a determinant, a matrix addition, a matrix subtraction, a matrix multiplication, a matrix division, a matrix inverse, a basis vector, a basis, an eigenvalue, and an eigenvector.

5. The system of claim 3, wherein:

each CI has at least one component; and

computing the AF comprises;

transforming each of the at least one component of each CI of the time series in the sliding time window,

computing, for each component, a component auto function (CAF) based on the corresponding transformed component of each CI of the time series in the sliding time window, and

computing the AF by aggregating the at least one CAF associated with the at least one component based on weighted averaging.

6. The system of claim 3, wherein computing the IQ-related feature of the AF comprises:

determining a presence of a class of features comprising the IQ-related feature in the AF in a window of argument of the AF; and

computing the IQ-related feature of the AF in the window of argument of the AF, wherein determining the presence of the class of features in the window of argument of the AF comprises;

computing a first regression of the AF in the window of argument of the AF and a first regression error corresponding to the first regression,

computing a second regression of the AF in the window of argument of the AF and a second regression error corresponding to the second regression,

computing a function of the first regression error and the second regression error, and

determining the presence of the class of features in the AF by comparing the function with a threshold, wherein the IQ-related feature in the AF comprises at least one of the following of the AF;

a maximum, a minimum, a constrained maximum, a constrained minimum, a local maximum, a local minimum, a first local maximum, a first local minimum, a k-th local maximum, a k-th local minimum, an average, a weighted average, a percentile, a mean, a median, a mode, a trimmed mean, a conditional mean, and an argument associated with another quantity.

7. The system of claim 3, wherein computing the IQ based on the IQ-related feature of the AF comprises:

determining an IQ-related argument of the AF in a window of argument of the AF corresponding to the IQ-related feature; and

computing the IQ based on at least one of;

the IQ-related feature, the IQ-related argument, and the window of argument of the AF.

8. The system of claim 2, wherein the transient motion monitor is further configured for:

computing at least one feature of the IQ, wherein the transient motion of the object is monitored based on the at least one feature of the IQ, and wherein the at least one feature comprises at least one of;

an IQ within an analysis time window, a maximum, a minimum, a local maximum, a local minimum, an IQ within a target range, a thresholding, an incremental change, a rate of change, a first derivative, a second derivative, a higher order derivative, a mean, a median, a mode, a variance, a total variation, a filtered IQ, a lowpass filtered IQ, a highpass filtered IQ, a nonlinear filtered IQ, a transformed IQ, and a function of IQ.

9. The system of claim 2, wherein the transient motion monitor is further configured for:

comparing a feature of the IQ with a threshold; and

detecting the transient motion of the object when the feature of IQ is larger than the threshold, wherein the response action is triggered based on the detected transient motion of the object.

10. The system of claim 2, wherein the transient motion monitor is further configured for:

determining that a first significant fraction of IQ in an observation time window are within a first target range, and a second significant fraction of changing rate of IQ in the observation time window are within a second target range, wherein;

the transient motion is part of a series of repeated gait cycles of the object,

each of the observation time window, the first target range and the second target range is related to the repeated gait cycles of the object,

the transient motion of the object is monitored by analyzing the IQ in the observation time window and estimating an information of the repeated gait cycles of the object based on the IQ in the observation time window, and

the information comprises at least one of;

a cycle duration, a cycle period, a cycle frequency, a gait intensity, a stage of a gait cycle, a gait start time, a gait stop time, a time trend of the gait cycles, a classification of the gait, a behavior associated with the gait, an activity associated with the gait, a meaning associated with the gait, a state of the object associated with the gait, a change in gait, a normal gait, a deviation from normal gait, and another gait with a characteristics different from the repeated gait cycles.

11. The system of claim 10, wherein analyzing the IQ in the observation time window comprises:

identifying at least one of the following characteristic IQ features;

a significant feature of IQ, a feature of IQ in a target range, an IQ with a magnitude in a target IQ range, a local mean of IQ, a local variance of IQ, a local behavior of IQ, a local statistics of IQ, a global statistics of IQ, a global behavior of IQ, a local maximum IQ, a local minimum IQ, a zero-crossing of IQ, a zero-crossing of mean-subtracted IQ, a pair of adjacent local maximum IQ and local minimum IQ, time stamp associated with an identified characteristic IQ feature, a window of IQ that match a training pattern; and

computing the information based on the identified at least one characteristic IQ feature.

12. The system of claim 2, wherein the transient motion monitor is further configured for:

determining a first analysis time window of a first window length comprising at least one first time stamp, wherein each of the at least one first time stamp is associated with a respective sliding time window;

computing at least one first intermediate quantity (IQ) each associated with a respective first time stamp in the first analysis time window based on CI of the time series in the respective sliding time window;

comparing the at least one first IQ in the first analysis time window with at least one second IQ in a second analysis time window of a second window length, wherein each of the at least one second IQ is associated with a respective second time stamp in the second analysis time window;

computing a first nonlinear mapping between the at least one first time stamp in the first analysis time window and the at least one second time stamp in the second analysis time window based on the at least one first IQ in the first analysis time window and the at least one second IQ in the second analysis time window;

computing a first matching score between the at least one first IQ in the first analysis time window and the at least one second IQ in the second analysis time window based on the first nonlinear mapping;

monitoring the transient motion of the object based on the first matching score; and

detecting the transient motion of the object when the first matching score is less than or equal to a threshold.

13. The system of claim 12, wherein the transient motion monitor is further configured for:

computing a second nonlinear mapping between the at least one first time stamp in the first analysis time window and at least one third time stamp in a third analysis time window based on the at least one first IQ in the first analysis time window and at least one third IQ in the third analysis time window, wherein each of the at least one third IQ is associated with one of the at least one third time stamp in the third analysis time window;

computing a second matching score between the at least one first IQ in the first analysis time window and the at least one third IQ in the third analysis time window based on the second nonlinear mapping;

comparing the first matching score and the second matching score to generate a comparison result; and

identifying, based on the comparison result, the transient motion of the object as one of;

a first reference motion and a second reference motion, wherein the first reference motion is associated with the at least one second IQ in the second analysis time window, and wherein the second reference motion is associated with the at least one third IQ in the third analysis time window.

14. The system of claim 12, further comprising at least one heterogeneous transmitter and at least one heterogeneous receiver for obtaining at least one time series of training channel information (CI) in a training stage, wherein:

each of the at least one heterogeneous transmitter is located in a respective training venue and configured for transmitting a respective wireless signal through the wireless multipath channel impacted by a respective training transient motion of a respective training object in the respective training venue in the training stage;

each of the at least one heterogeneous receiver is located in the respective training venue and configured for;

receiving the respective wireless signal through the wireless multipath channel, and

obtaining a respective time series of training CI of the wireless multipath channel based on the respective wireless signal, wherein a respective time window of the respective time series of training CI is associated with the respective training transient motion; and

the transient motion monitor is further configured for;

computing at least one training IQ each associated with a respective time stamp in the respective time window, based on training CI of the respective time series in a sliding time window associated with the respective time stamp, and

computing the at least one second IQ in the second analysis time window of the second window length, based on the at least one training IQ associated with each of the at least one time series of training CI.

15. The system of claim 12, wherein the transient motion monitor is further configured for:

computing a third nonlinear mapping between time stamps of a first time window and time stamps of a second time window, based on at least one first training IQ in the first time window associated with a first time series of training CI and at least one second training IQ in the second time window associated with a second time series of training CI;

computing, based on the third nonlinear mapping, a third matching score between the at least one first training IQ in the first time window and the at least one second training IQ in the second time window; and

computing the at least one second IQ in the second analysis time window of the second window length, based on at least one of;

the third matching score, the third nonlinear mapping, the at least one first training IQ and the at least one second training IQ.

16. The system of claim 2, wherein the transient motion monitor is further configured for:

determining a first analysis time window of a first window length comprising at least one first time stamp, wherein each of the at least one first time stamp is associated with a respective sliding time window;

for each first time stamp in the first analysis time window, computing a set of N first intermediate quantity (IQ) comprising IQ_1, IQ_2, . . . , and IQ_N associated with the first time stamp based on CI of the time series in the respective sliding time window associated with the time stamp, wherein N is greater than 1;

comparing the set of N first IQ associated with each first time stamp in the first analysis time window with a set of N second IQ associated with one of at least one second time stamp in a second analysis time window of a second window length;

computing a nonlinear mapping between the at least one first time stamp in the first analysis time window and the at least one second time stamp in the second analysis time window based on all first IQ in the first analysis time window and all second IQ in the second analysis time window;

computing, based on the nonlinear mapping, a matching score between the first IQ in the first analysis time window and the second IQ in the second analysis time window;

monitoring the transient motion of the object based on the matching score; and

detecting the transient motion of the object when the matching score is less than or equal to a threshold.

17. The system of claim 1, wherein the response action comprises at least one of:

presenting a data related to the transient motion of the object, displaying the data visually, playing an audio related to the data, animating the data,

presenting an alarm, displaying an alarm, sounding an alarm, animating an alarm, sending a message to a designated user, informing emergency service, requesting emergency service, taking corrective measure, taking preventive measure, initiating emergency response procedure, activating counter measure,

controlling the object, controlling the transient motion of the object, controlling a next motion of the object, controlling a speed of the object, controlling a direction of the object, controlling a gait of the object, controlling a state of the object,

controlling a second device in response to the transient motion of the object, turning on the second device, turning off the second device, turning off the second device and turning on a third device,

sending a data to a fourth device, triggering the fourth device to respond collaboratively, presenting the data on the fourth device, displaying the data visually on the fourth device, playing an audio to present the data on the fourth device, and sounding an alarm on the fourth device.

18. The system of claim 1, further comprising:

an additional receiver configured for;

receiving the wireless signal through the wireless multipath channel impacted by the transient motion of the object in the venue, and

obtaining an additional time series of CI of the wireless multipath channel based on the wireless signal,

wherein the transient motion monitor is further configured for monitoring the transient motion of the object jointly based on the time series of CI and the additional time series of CI.

19. The system of claim 1, further comprising:

an additional transmitter configured for transmitting an additional wireless signal through the wireless multipath channel impacted by the transient motion of the object in the venue,

wherein the receiver is further configured for;

receiving the additional wireless signal through the wireless multipath channel, and

obtaining an additional time series of CI of the wireless multipath channel based on the additional wireless signal,

wherein the transient motion monitor is further configured for monitoring the transient motion of the object jointly based on the time series of CI and the additional time series of CI.

20. The system of claim 1, wherein the transient motion monitor is further configured for computing a location of the transient motion of the object based on at least one of:

a location of the transmitter;

a location of the receiver;

a location of a device that is at least one of;

an additional receiver in the venue for receiving the wireless signal from the transmitter through the wireless multipath channel, and

an additional transmitter in the venue for transmitting an additional wireless signal to the receiver through the wireless multipath channel; and

a joint monitoring of the transient motion of the object based on the time series of CI and at least one of;

an additional time series of CI extracted from the wireless signal received by the additional receiver, and

an additional time series of CI extracted from the additional wireless signal transmitted by the additional transmitter.

21. The system of claim 1, wherein:

the object is a person;

the transient motion represents a fall-down action of the person; and

the transient motion monitor is coupled to at least one of;

the transmitter, the receiver, an additional transmitter, an additional receiver, a cloud server, a fog server, a local server, and an edge server.

22. A method implemented on a machine having a processor, a memory communicatively coupled with the processor and a set of instructions stored in the memory for monitoring a transient motion in a venue, comprising:

receiving a wireless signal through a wireless multipath channel impacted by the transient motion of an object in the venue;

obtaining a time series of channel information (CI) of the wireless multipath channel based on the wireless signal;

monitoring the transient motion of the object based on the time series of CI; and

triggering a response action based on the monitored transient motion of the object.

23. The method of claim 22, further comprising:

determining a sliding time window having a duration comparable to the transient motion of the object; and

computing an intermediate quantity (IQ) based on CI of the time series in the sliding time window, wherein the transient motion of the object is monitored based on the IQ.

24. The method of claim 23, wherein computing the IQ based on the CI comprises:

computing an auto function (AF) based on the CI of the time series in the sliding time window;

computing an IQ-related feature of the AF; and

computing the IQ based on the IQ-related feature of the AF, wherein the AF comprises at least one of;

auto-correlation function, auto-covariance function, cross-correlation function, cross-covariance function, differential function, differential auto-correlation, differential auto-covariance, differential cross-correlation, differential cross-covariance, a frequency transform, frequency spectrum, and a power spectral density.

25. The method of claim 24, wherein computing the AF comprises:

transforming each CI of the time series in the sliding time window; and

computing the AF of each transformed CI in the sliding time window, wherein transforming each CI comprises performing at least one of the following on the CI;

a scalar operation, a magnitude, a phase, a nonlinear map, a linear map, a piecewise linear map, a function, a linear function, a nonlinear function, a piecewise linear function, a real function, a complex function, a vector-valued function, an inverse function, a derivative of a function, an integration of a function, a function of another function, a one-to-one function, a one-to-many function, a many-to-one function, a many-to-many function, a logarithmic function, an exponential function, a trigonometric function, a circular function, an algebraic function, a transcendental function, a power function, a root function, a polynomial, a square, a cube, a sine, a cosine, a tangent, a cotangent, a secant, a cosecant, an elliptical function, a parabolic function, a hyperbolic function, a gamma function, a zeta function, an absolute value, a thresholding, a quantization, a piecewise constant function, a piecewise linear function, a table look-up,

a set operation, a histogram, a combination, a permutation, a subset, an inclusion, a logical operation, an OR, an ADD, an XOR, a union, an intersection, a set difference, a symmetric difference a Cartesian product, a power set, a cardinality, an average, a weighted average, a mean, a variance, a range, a trimmed mean, a statistics, a percentile,

a vector operation, a norm, a metric, a vector addition, a vector subtraction, a vector multiplication, a vector division, a dot product, an inner product, an outer product, an arithmetic operation, a mathematical operation, a subspace, a projection, a decomposition, a vector transformation, a change of basis, a rotation, a translation, a vector function, a projection, a principal component, a kernel principal component, an independent component, a neighborhood component, a connected component,

a linear transform, a nonlinear transform, Fourier transform, Laplace transform, Hilbert transform, Sine transform, Cosine transform, Wavelet transform, Hadamard transform, Slant transform, Fast Fourier transform (FFT), discrete cosine transform (DCT), frequency transform, frequency decomposition, inverse transform, time transform, time decomposition, a frequency, a spectrum, a period, a time,

a matrix operation, a rank, a transpose, a determinant, a matrix addition, a matrix subtraction, a matrix multiplication, a matrix division, a matrix inverse, a basis vector, a basis, an eigenvalue, and an eigenvector.

26. The method of claim 24, wherein:

each CI has at least one component; and

computing the AF comprises;

transforming each of the at least one component of each CI of the time series in the sliding time window,

computing, for each component, a component auto function (CAF) based on the corresponding transformed component of each CI of the time series in the sliding window, and

computing the AF by aggregating the at least one CAF associated with the at least one component based on weighted averaging.

27. The method of claim 24, wherein computing the IQ-related feature of the AF comprises:

determining a presence of a class of features comprising the IQ-related feature in the AF in a window of argument of the AF; and

computing the IQ-related feature of the AF in the window of argument of the AF, wherein determining the presence of the class of features in the window of argument of the AF comprises;

computing a first regression of the AF in the window of argument of the AF and a first regression error corresponding to the first regression,

computing a second regression of the AF in the window of argument of the AF and a second regression error corresponding to the second regression,

computing a function of the first regression error and the second regression error, and

determining the presence of the class of features in the AF by comparing the function with a threshold, wherein the IQ-related feature in the AF comprises at least one of the following of the AF;

a maximum, a minimum, a constrained maximum, a constrained minimum, a local maximum, a local minimum, a first local maximum, a first local minimum, a k-th local maximum, a k-th local minimum, an average, a weighted average, a percentile, a mean, a median, a mode, a trimmed mean, a conditional mean, and an argument associated with another quantity.

28. The method of claim 24, wherein computing the IQ based on the IQ-related feature of the AF comprises:

determining an IQ-related argument of the AF in the sliding time window corresponding to the IQ-related feature; and

computing the IQ based on at least one of;

the IQ-related feature, the IQ-related argument, and the sliding time window.

29. A monitor of a motion monitoring system, comprising:

a processor;

a memory communicatively coupled with the processor; and

a set of instructions stored in the memory which, when executed, causes the processor to perform;

obtaining a time series of channel information (CI) of a wireless multipath channel from a receiver of the motion monitoring system, wherein the receiver extracts the time series of CI from a wireless signal received from a transmitter of the motion monitoring system through the wireless multipath channel impacted by a transient motion of an object in a venue,

monitoring the transient motion of the object based on the time series of CI, and

triggering a response action based on the monitored transient motion of the object.

30. A receiver of a motion monitoring system, comprising:

a wireless circuitry configured to receive a wireless signal through a wireless multipath channel impacted by a transient motion of an object in a venue, wherein the wireless signal is transmitted by a transmitter of the motion monitoring system;

a processor communicatively coupled with the wireless circuitry;

a memory communicatively coupled with the processor; and

a set of instructions stored in the memory which, when executed, causes the processor to obtain a time series of channel information (CI) of the wireless multipath channel based on the wireless signal, wherein;

the time series of CI is to be used by a transient motion monitor of the motion monitoring system to monitor the transient motion of the object and trigger a response action based on the monitored transient motion.