Wireless Sensing CSI Filtering for Human Activity Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wireless sensing technologies face challenges in accurately detecting human activity due to the complexity and variability of real-world environments, which introduce confounding factors such as multipath propagation and interference from stationary and intermittently moving objects.
Innovation Solution
The implementation of a preprocessing pipeline that includes a motion target indicator (MTI) filter and a self-supervised learning framework to train a Bayesian Convolutional Neural Network (CNN) using contrastive data augmentation, which enhances the accuracy of human activity predictions by filtering out stationary object channel responses and amplifying human movement-induced variations in CSI data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wireless sensing is used to detect human activity in real-world environments, then activity detection capability is provided, but accuracy deteriorates due to confounding factors such as multipath propagation and interference from stationary objects
Solution Approach 1:
The patent extracts and removes channel responses corresponding to stationary objects from the CSI data using the MTI filter. This separation isolates the stationary object interference from the moving object signals, allowing the system to focus on detecting human activity without contamination from static environmental factors.
Solution Approach 2:
The system performs preliminary filtering of stationary object channel responses before activity detection. By pre-processing the CSI data to remove stationary object interference, the system prepares cleaner input data for subsequent activity recognition algorithms, improving overall detection accuracy.
2Reliability
If traditional wireless sensing methods are used, then basic presence detection is achieved, but reliability deteriorates in complex environments with multiple objects
Solution Approach 1:
The patent segments the channel responses in CSI data into different categories: stationary object responses and moving object responses. This segmentation allows the system to process and analyze each type separately, improving reliability by focusing on relevant moving object signals while filtering out stationary background interference.
Solution Approach 2:
The MTI filter acts as an intermediary component between the raw CSI data and the activity detection algorithm. It mediates the processing by selectively removing stationary object channel responses, thereby protecting the downstream detection system from environmental complexity and improving prediction reliability.
3Productivity
If all channel responses are processed for activity detection, then comprehensive data analysis is performed, but processing efficiency deteriorates due to redundant stationary object data
Solution Approach 1:
The system extracts and removes redundant stationary object channel responses from the CSI data before processing. This extraction eliminates unnecessary computational workload on static environmental data, significantly improving processing efficiency while preserving all relevant information about moving objects and human activity.
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
This approach significantly improves the accuracy of human activity detection by isolating and amplifying human movement-induced variations in CSI data, while reducing the impact of confounding factors, thereby enabling more reliable predictions of human presence and activity.
Implementation Method 1
When the wireless signal encounters an object, the wireless signal may be reflected, refracted, scattered, or absorbed
Implementation Method 2
When the wireless signal encounters an object, the wireless signal may be reflected, refracted, scattered, or absorbed
Implementation Method 3
When the wireless signal encounters an object, the wireless signal may be reflected, refracted, scattered, or absorbed
Implementation Method 4
When the wireless signal encounters an object, the wireless signal may be reflected, refracted, scattered, or absorbed
Implementation Method 5
moving objects in the area may also produce shifts in the wireless signal's frequency
Data Source
AI summary
The present disclosure provides an approach that captures one or more wireless signals in a geographic area. Each one of the one or more wireless signals includes channel state information (CSI) data. The present disclosure produces a channel state information (CSI) representation based on the CSI data that indicates multiple channel responses corresponding to the one or more wireless signals. The present disclosure filters the CSI representation to remove at least one of the channel responses that correspond to a stationary object within the geographic area to produce a filtered CSI representation. The present disclosure predicts a presence of a moving object within the geographic area based on the filtered CSI representation.