Wireless Occupancy Detection for Static Humans in Noisy Venues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional occupancy detection methods, such as cameras, PIR sensors, and RFID sensors, are costly, privacy-invasive, and suffer from low accuracy, while existing WiFi sensing applications struggle to identify static humans and are ineffective in complex environments due to noise and interference.
Innovation Solution
A method utilizing a deep learning model to process autocorrelation function segments of channel information from wireless signals to detect user presence, independent of transmitter quantity and location, by generating feature maps and computing a probability of user presence based on aggregate representations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional methods such as cameras, PIR sensors, and RFID sensors are used for occupancy detection, then occupancy detection capability is provided, but cost increases and privacy concerns arise
Solution Approach 1:
The patent replaces physical sensing devices (cameras, PIR sensors, RFID sensors) with a wireless signal-based detection system. The system uses wireless signals (e.g., WiFi) that already exist in the environment to detect occupancy by analyzing changes in channel state information, eliminating the need for additional specialized sensors and reducing both cost and privacy intrusion.
Solution Approach 2:
The system uses existing wireless communication infrastructure and signals rather than creating new sensing systems. By repurposing wireless communication channels for sensing purposes, the invention avoids the costs and complexities associated with deploying dedicated sensing hardware while maintaining detection capability.
2Reliability
If existing WiFi sensing applications are used, then wireless occupancy detection is provided, but static humans cannot be identified and performance degrades in complex environments
Solution Approach 1:
The patent segments the channel state information into multiple components and processes them through separate processing streams. By dividing the complex sensing task into manageable segments and analyzing different aspects of the wireless signal separately, the system can extract more detailed information about static objects and better handle complex environmental conditions.
Solution Approach 2:
The system transitions from traditional motion-based detection to analyzing multiple dimensions of channel state information including amplitude, phase, and temporal characteristics. This multi-dimensional analysis enables the detection of static humans by capturing subtle changes in the wireless channel that are invisible to conventional motion-only systems.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances occupancy detection accuracy by identifying static humans and improving performance in complex environments, offering cost-effective and privacy-preserving solutions for energy management and security.
Implementation Method 1
transmitting, by each of a plurality of transmitters, a respective wireless signal through a wireless channel of a venue; receiving, by a receiver, the respective wireless signal from each of the plurality of transmitters through the wireless channel of the venue, wherein the received wireless signal differs from the transmitted wireless signal due to the wireless channel and a motion of a user
Data Source
AI summary
Examples for wireless based occupancy detection are described. In one example, a described method comprises: transmitting, by each of a plurality of transmitters, a respective wireless signal through a wireless channel of a venue; receiving, by a receiver, the respective wireless signal, the received wireless signal different from the transmitted wireless signal due to the wireless channel and a motion of a user when the user is present in the venue; obtaining a plurality of time series of channel information (TSCI) of the wireless channel; computing a plurality of autocorrelation function (ACF) segments based on the plurality of TSCI; generating a plurality of feature maps using a first deep learning model that is shared by the plurality of ACF segments, wherein each ACF segment is input into the first deep learning model individually to generate a respective feature map; inputting the plurality of feature maps together into a second deep learning model to generate an aggregate representation that is independent of a quantity and locations of the transmitters; computing a probability of user presence based on the aggregate representation; and detecting a presence of the user in the venue based on a threshold and the probability of user presence.