# US 12,604,258 B2 · Method and system for identifying Wi-Fi devices

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

- Number: US 12,604,258 B2
- Document ID: US12604258B2
- Status: Granted
- Kind: B2
- Application number: 18/687,559
- Filed: 2022-07-18
- Published: 2024-11-14
- Granted: 2026-04-14
- Earliest claimed: 2022-06-17
- Assignees: Galgus Global S.R.L.
- Applicants: Galgus Global S.R.L.
- Inventors: Abraham Pérez Hernández; Pablo Aguilera Bonet; Jesús Fernández Manzano; José Ayub González Garrido; José Antonio Delgado Alonso
- Web page: https://patentgrep.com/patent/US12604258B2

## CPC

- H04W 12/062: Pre-authentication
- H04W 48/16: Discovering, processing access restriction or access information
- H04W 4/80: Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
- H04W 12/71: Hardware identity
- H04W 12/79: Radio fingerprint
- H04W 24/08: Testing, {supervising or monitoring} using real traffic
- H04W 84/12: WLAN [Wireless Local Area Networks]
- H04W: WIRELESS COMMUNICATION NETWORKS (broadcast communication H04H; communication systems using wireless links for non-selective communication, e.g. wireless extensions H04M1/72)

## Abstract

A method and system for identifying Wi-Fi devices, the system (200) having a plurality N of access points (AP1, AP2, . . . , APN) arranged at an installation (204) and a server (210). The access points are configured to receive probe request frames (104) sent by Wi-Fi devices (202); generate a fingerprint (108) associated to each probe request frame (104); and send the generated fingerprint (108), a timestamp (502) and an RSSI measurement (508) associated to each probe request frame (104) to a server (210) that has a processing unit (220) to identify bursts (114) of probe request frames using their associated timestamps (502); for each burst i (114), determine an averaged RSSI measurement (118) received by each access point; for each fingerprint (108), perform an N-dimensional cluster analysis on the averaged RSSI measurements (118), obtaining at least one cluster (122) per fingerprint (108); and identify a Wi-Fi device for each different cluster (122).

## Claims

### Claim 1

A method of identifying Wi-Fi devices, the method comprising:
receiving, by a plurality N of access points (AP1, AP2, . . . , APN) arranged at an installation, probe request frames sent by Wi-Fi devices;
generating, by the plurality N of access points (AP1, AP2, . . . , APN), a fingerprint associated to each probe request frame;
sending, from the plurality N of access points (AP1, AP2, . . . , APN) to a server, the generated fingerprint, a timestamp and a Received Signal Strength Indicator (RSSI) measurement associated to each probe request frame;
identifying, by the server, bursts of probe request frames using their associated timestamps;
for each burst i of probe request frames, determining an averaged RSSI measurement (Xi1, Xi2, . . . , XiN) received by each access point (AP1, AP2, . . . , APN);
for each fingerprint, performing an N-dimensional cluster analysis on the averaged RSSI measurements (Xi1, Xi2, XiN) received by each access point (AP1, AP2, . . . , APN), obtaining at least one cluster per fingerprint; and
identifying a Wi-Fi device for each different cluster.

### Claim 2 (depends on 1)

The method of claim 1, wherein the cluster analysis includes:
computing a maximum distance in the N-dimensional RSSI space of two points of any cluster to obtain a spread scoring; and
if the spread scoring is higher that a clustering threshold, incrementing the number of clusters (Nc) by one.

### Claim 3 (depends on 2)

The method of claim 2, wherein the clustering threshold is a value trained using training probe requests frames received by the plurality N of access points (AP1, AP2, . . . APN), wherein the training probe request frames contain real Medium Access Control (MAC) addresses of Wi-Fi devices.

### Claim 4 (depends on 1)

The method of claim 1, wherein the cluster analysis is performed on the averaged RSSI measurements received by each access point (AP1, AP2, . . . , APN) within a determined time window.

### Claim 5 (depends on 4)

The method of claim 4, wherein a temporal span of the time window is trained using training probe requests frames received by the plurality N of access points (AP1, AP2, . . . , APN), wherein the training probe requests frames contain real MAC addresses of Wi-Fi devices.

### Claim 6

A system for identifying Wi-Fi devices, the system comprising:
a plurality N of access points (AP1, AP2, . . . , APN) arranged at an installation and configured to:
receive probe request frames sent by Wi-Fi devices;
generate a fingerprint associated to each probe request frame; and
send the generated fingerprint, a timestamp and a Received Signal Strength
Indicator (RSSI) measurement associated to each probe request frame to a server; and
the server, comprising a memory and a processing unit configured to:
identify bursts of probe request frames using their associated timestamps;
for each burst i of probe request frames, determine an averaged RSSI measurement (Xi1, Xi2, XiN) received by each access point (AP1, AP2, . . . , APN);
for each fingerprint, perform an N-dimensional cluster analysis on the averaged RSSI measurements (Xi1, Xi2, . . . , XiN) received by each access point (AP1, AP2, . . . , APN), obtaining at least one cluster per fingerprint; and
identify a Wi-Fi device for each different cluster.

### Claim 7 (depends on 6)

The system of claim 6, wherein for the cluster analysis the processing unit is configured to:
compute a maximum distance in the N-dimensional RSSI space of two points of any cluster to obtain a spread scoring; and
if the spread scoring is higher that a clustering threshold, increment the number of clusters (Nc) by one.

### Claim 8 (depends on 7)

The system of claim 7, wherein the clustering threshold is a value trained using training probe requests frames received by the plurality N of access points (AP1, AP2, . . . , APN), wherein the training probe requests frames contain real MAC addresses of Wi-Fi devices.

### Claim 9 (depends on 6)

The system of claim 6, wherein the processing unit is configured to perform the cluster analysis on the averaged RSSI measurements received by each access point (AP1, AP2, . . . , APN) within a determined time window.

### Claim 10 (depends on 9)

The system of claim 9, wherein a temporal span of the time window is trained using training probe requests frames received by the plurality N of access points (AP1, AP2, . . . , APN), wherein the training probe requests frames contain real MAC addresses of Wi-Fi devices.