# US 2024/0187822 A1 · Handling concept drift in wi-fi-based localization

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2024/0187822 A1
- Document ID: US20240187822A1
- Status: Published application
- Kind: A1
- Application number: 18/438,864
- Filed: 2024-02-12
- Published: 2024-06-06
- Earliest claimed: 2019-02-22
- Assignees: Aerial Technologies Inc.
- Applicants: Aerial Technologies Inc.
- Inventors: QIANYU LIU; NEGAR GHOURCHIAN; MICHEL ALLEGUE MARTINEZ; RAIHAN SERAJ
- Web page: https://patentgrep.com/patent/US20240187822A1

## CPC

- G01S 7/417: {involving the use of neural networks}
- H04W 4/029: Location-based management or tracking services
- G01S 7/415: {Identification of targets based on measurements of movement associated with the target}
- G01S 13/003: {Bistatic radar systems; Multistatic radar systems}
- G01S 13/56: for presence detection {(presence detection using near field arrangements G01V3/00, e.g. G01V3/08, G01V3/12; burglar, theft or intruder alarms with electrical actuation G08B13/22 - G08B13/26)}
- G01S 13/72: for two-dimensional [2D] tracking, e.g. combination of angle and range tracking, track-while-scan radar
- H04B 17/19: Self-testing arrangements
- H04B 17/26: using historical data, averaging values or statistics
- H04W 84/12: WLAN [Wireless Local Area Networks]
- G01S: RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)
- H04B: TRANSMISSION

## Abstract

Systems and methods are provided for human activity analysis and localization using channel state information (CSI). A streamlined data processing and feature extraction approach that addresses concept drifts for time series data is provided. Data obtained from CSI or any other mechanism is used to estimate the wireless channel between two different wireless nodes (e.g., an access point (AP) and an associated station (STA)), and can be used to train a robust system capable of performing room level localization. A phase and magnitude augmented feature space along with a standardization technique that is little affected by drifts is also used.

## Claims

### Claim 1

A method for handling drift in wireless signals, the method comprising:
acquiring a series of channel state information (CSI) data sets, each CSI data set within the series of CSI data sets comprising channel state information (CSI) data from a wireless access point for a wireless link incorporating the wireless access point associated with an area comprising areas of interest at a point in time;
processing the acquired series of CSI data sets with a pre-processing module forming part of a system comprising at least a microprocessor to generate a processed series of CSI data sets;
generating with a feature acquisition module forming part of the system a series of sets of statistical properties, each set of statistical properties generated in dependence upon a CSI data set of the processed series of CSI data sets and representative of the area of interest;
processing the series of sets of statistical properties with a post-processing module forming part of the system where the post-processing module generates a stream of predictions of a location of a user within the area and filters the stream of predictions; wherein
each prediction of location of the user is an area of interest within the area.

### Claim 2 (depends on 1)

The method according to claim 1, wherein
the filtering applied by the post-processing module leverages current predictions and previous predictions within the stream of predictions to eliminate outliers within the stream of predictions; wherein
the current predictions and previous predictions are within a short period of time.

### Claim 3 (depends on 1)

The method according to claim 1, wherein
the filtering applied by the post-processing module leverages series of sets of statistical properties to eliminate outliers within the stream of predictions.

### Claim 4 (depends on 1)

The method according to claim 1, wherein
the post-processing module comprises a clustering unit; and
the clustering unit comprises a supervised learner comprising one or more machine learning algorithms and an unsupervised learner comprising one or more other machine learning algorithms.

### Claim 5 (depends on 1)

The method according to claim 1, wherein
the post-processing module comprises a clustering unit comprising a supervised learner comprising one or more machine learning algorithms and an unsupervised learner comprising one or more other machine learning algorithms; and
the unsupervised learner accounts for some areas of interest being similar in their CSI profiles which are employed by the post-processing module to generate a predictions.

### Claim 6 (depends on 1)

The method according to claim 1, wherein
the post-processing module comprises a clustering unit comprising a supervised learner comprising one or more machine learning algorithms and an unsupervised learner comprising one or more other machine learning algorithms;
the unsupervised learner accounts for some areas of interest being similar in their CSI profiles which are employed by the post-processing module to generate a predictions; and
the areas of interest within the area are defined by another user prior to the system localizing the user via the stream of predictions.

### Claim 7 (depends on 1)

The method according to claim 1, wherein
the feature acquisition module comprises a stream and subcarrier selection unit such that the series of sets of statistical properties are generated from sub-bands of the wireless link; and
the sub-bands are a fixed number of sub-bands and are uniformly spaced.

### Claim 8 (depends on 1)

The method according to claim 1, wherein
the feature acquisition module comprises a stream and subcarrier selection unit such that the series of sets of statistical properties are generated from sub-bands of the wireless link; and
the sub-bands are a fixed number of sub-bands and are uniformly spaced.

### Claim 9 (depends on 1)

The method according to claim 1, wherein
the feature acquisition module comprises a stream and subcarrier selection unit such that the series of sets of statistical properties are generated from a number of sub-carriers of the wireless link which comprises a plurality of sub-carriers;
the sub-bands are established by a processing comprising:
initially selecting a fixed number of sub-carriers which are uniformly spaced;
for each sub-carrier of the number of sub-carriers establishing a magnitude of the sub-carrier and its neighboring sub-carriers within the plurality of sub-carriers;
determining in dependence upon the magnitude of the sub-carrier and its neighboring sub-carriers within the plurality of sub-carriers whether to change the initially selected sub-carrier to another sub-carrier of the plurality of sub-carriers; and
upon a determination to change the initially selected sub-carrier to another sub-carrier of the plurality of sub-carriers changing the initially selected sub-carrier to another sub-carrier of the plurality of sub-carriers.

### Claim 10 (depends on 1)

The method according to claim 1, wherein
each CSI data set within the series of CSI data sets comprises a frame and a packet index;
the pre-processing module processes each CSI data set within the series of CSI data sets to generate a first derivative of the frame with respect to packet index and a second derivative of the frame with respect to packet index.

### Claim 11 (depends on 1)

The method according to claim 1, wherein
each CSI data set within the series of CSI data sets comprises a frame and a packet index;
the pre-processing module processes each CSI data set within the series of CSI data sets by:
applying a data preparation process with respect to subcarrier index to compensate for temporal instabilities by applying a fixed-score scaling normalization and digital filtering to the frame; and
applying another data preparation step along a frequency axis by applying a normalization and digital filtering to the frame;
generating a first derivative of the frame with respect to packet index; and
generating a second derivative of the frame with respect to packet index.

### Claim 12 (depends on 1)

The method according to claim 1, wherein
generating with the feature acquisition module forming part of the system the series of sets of statistical properties comprises applying one or more feature extraction strategies to each CSI data set of the processed series of CSI data sets;
each feature extraction strategy of the one or more feature extraction strategies assumes that motion of the user within the area creates one of identifiable features over time and identifiable changes over multiple signal paths and frequency bands.

### Claim 13 (depends on 1)

The method according to claim 1, wherein
each CSI data set within the series of CSI data sets comprises a frame and a packet index;
generating with the feature acquisition module forming part of the system the series of sets of statistical properties comprises applying one or more feature extraction strategies to each CSI data set of the processed series of CSI data sets; and
each feature extraction strategy of the one or more feature extraction strategies assumes that motion of the user within the area creates one of identifiable features over time and identifiable changes over multiple signal paths and frequency bands;
each feature extraction strategy calculates a statistical feature along a first axis of the frame an averages the statistical feature along a second axis of the frame; and
an axis of the frame is one of a packet axis, stream axis and a sub-carrier axis.

### Claim 14 (depends on 1)

The method according to claim 1, wherein
the pre-processing module applies a filter to the received series of CSI data sets; and
the filter executes a process of analysing successive packets of received data received with the wireless access point, establishing for each packet a received signal strength indicator (RSSI) value, identifying a sudden peak in RSSI value; and discarding those packets associated with the sudden peak in RSSI value.
