Apparatus, systems and methods for event recognition based on a wireless signal

Patent metadata

Field	Value
Publication number	US-10374863-B2
Application number	US-201816203299-A
Country	US
Kind code	B2
Filing date	Nov 28, 2018
Priority date	Dec 5, 2012
Publication date	Aug 6, 2019
Grant date	Aug 6, 2019

Abstract

Apparatus, systems and methods for recognizing and classifying events in a venue based on a wireless signal are disclosed. In one example, a disclosed system comprises a first transmitter, a second transmitter, at least one first receiver, at least one second receiver, and an event recognition engine, in the venue. The first transmitter transmits a training wireless signal through a wireless multipath channel impacted by a known event in the venue in a training time period associated with the known event. Each first receiver receives asynchronously the training wireless signal, and obtains, asynchronously based on the training wireless signal, at least one time series of training channel information of the wireless multipath channel between the first receiver and the first transmitter. The second transmitter transmits a current wireless signal through the wireless multipath channel impacted by a current event in a current time period associated with the current event. Each second receiver receives asynchronously the current wireless signal, and obtains, asynchronously based on the current wireless signal, at least one time series of current channel information of the wireless multipath channel between the second receiver and the second transmitter. The event recognition engine trains a classifier based on the training channel information; and apples the classifier to: classify the current channel information and associate the current event with at least one of: a known event, an unknown event and another event.

First claim

We claim:

1. 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 recognizing an event, comprising:
for each of at least one known event happening in a venue in a respective training time period:
transmitting, by an antenna of a first transmitter, a respective training wireless signal to at least one first receiver through a wireless multipath channel impacted by the known event in the venue in the training time period associated with the known event,
obtaining, asynchronously by each of the at least one first receiver based on the training wireless signal, at least one time series of training channel information (training CI time series) of the wireless multipath channel between the first receiver and the first transmitter in the training time period associated with the known event, and
pre-processing the at least one training CI time series;

training a projection for CI using a dimension reduction method based on the training CI time series associated with the at least one known event,
training at least one classifier for the at least one known event based on the at least one training CI time series and the projection; and
for a current event happening in the venue in a current time period,
transmitting, by an antenna of a second transmitter, a current wireless signal to at least one second receiver through the wireless multipath channel impacted by the current event in the venue in the current time period associated with the current event,
obtaining, asynchronously by each of the at least one second receiver based on the current wireless signal, at least one time series of current channel information (current CI time series) of the wireless multipath channel between the second receiver and the second transmitter in the current time period associated with the current event,
pre-processing the at least one current CI time series, and
applying the at least one classifier to:
classify, based on the projection, at least one of: the at least one current CI time series, a portion of a particular current CI time series, and a combination of the portion of the particular current CI time series and a portion of an additional CI time series, and
associate the current event with at least one of: a known event, an unknown event and another event,

wherein a training CI time series associated with a first receiver and a current CI time series associated with a second receiver have at least one of:
different starting times,
different time durations,
different stopping times,
different counts of items in their respective time series,
different sampling frequencies,
different sampling periods between two consecutive items in their respective time series, and
channel information (CI) with different features.

2. The method of claim 1 , further comprising:
aligning a first section of a first time duration of a first CI time series and a second section of a second time duration of a second CI time series, and
determining a mapping between items of the first section and items of the second section.

3. The method of claim 2 , wherein:
the first CI time series is processed by a first operation;
the second CI time series is processed by a second operation; and
at least one of the first operation and the second operation comprises at least one of: subsampling, re-sampling, interpolation, filtering, transformation, feature extraction, and pre-processing.

4. The method of claim 2 , further comprising mapping a first item of the first section to a second item of the second section, wherein at least one constraint is applied on at least one function of at least one of:
the first item of the first section of the first CI time series;
another item of the first CI time series;
a time stamp of the first item;
a time difference of the first item;
a time differential of the first item;
a neighboring time stamp of the first item;
another time stamp associated with the first item;
the second item of the second section of the second CI time series; another item of the second CI time series;
a time stamp of the second item;
a time difference of the second item;
a time differential of the second item;
a neighboring time stamp of the second item; and
another time stamp associated with the second item.

5. The method of claim 4 , wherein one of the at least one constraint is that a difference between the time stamp of the first item and the time stamp of the second item is upper-bounded by an adaptive upper threshold and lower-bounded by an adaptive lower threshold.

6. The method of claim 1 , further comprising:
determining a section of a time duration of a CI time series adaptively; and
determining a starting time and an ending time of the section, wherein determining the section comprises:
computing a tentative section of the CI time series, and
determining the section by removing a beginning portion and an ending portion of the tentative section.

7. The method of claim 6 , wherein determining the section further comprises:
determining the beginning portion of the tentative section by:
considering items of the tentative section with increasing time stamp as a current item iteratively, one item at a time,
computing recursively an activity measure associated with at least one of:
the current item associated with a current time stamp,
past items of the tentative section with time stamps not larger than the current time stamp, and
future items of the tentative section with time stamps not smaller than the current time stamp,

adding the current item to the beginning portion of the tentative section when a first criterion associated with the activity measure is satisfied; and

determining the ending portion of the tentative section by:
considering items of the tentative section with decreasing time stamp as a current item iteratively, one item at a time,
iteratively computing and determining at least one activity measure associated with at least one of:
the current item associated with a current time stamp,
past items of the tentative section with time stamps not larger than the current time stamp, and
future items of the tentative section with time stamps not smaller than the current time stamp,

adding the current item to the ending portion of the tentative section when a second criterion associated with the at least one activity measure is satisfied.

8. The method of claim 7 , wherein:
at least one of the first criterion and the second criterion comprises at least one of:
the activity measure is smaller than an adaptive upper threshold,
the activity measure is larger than an adaptive lower threshold,
the activity measure is smaller than an adaptive upper threshold consecutively for at least a predetermined amount of consecutive time stamps,
the activity measure is larger than an adaptive lower threshold consecutively for at least an additional predetermined amount of consecutive time stamps,
the activity measure is smaller than an adaptive upper threshold consecutively for at least a predetermined percentage of the predetermined amount of consecutive time stamps,
the activity measure is larger than an adaptive lower threshold consecutively for at least another predetermined percentage of the additional predetermined amount of consecutive time stamps,
another activity measure associated with another time stamp associated with the current time stamp is smaller than another adaptive upper threshold and larger than another adaptive lower threshold,
at least one activity measure associated with at least one respective time stamp associated with the current time stamp is smaller than respective u

Assignees

• Origin Wireless Inc

Inventors

• Xu Qinyi
• Zhang Feng
• Chen Chen
• 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

Classifications

• H04W72/21

  in the uplink direction of a wireless link, i.e. towards the network · CPC title

• G08B13/187

  by interference of a radiation field · CPC title

• G08B29/20

  Calibration, including self-calibrating arrangements · CPC title

• G08B29/186

  Fuzzy logic; neural networks · CPC title

• H04L25/0202

  Channel estimation · CPC title