Deep-Learning Wireless Sensing With Autocorrelation for Motion Differentiation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless sensing systems struggle to accurately differentiate between human and non-human movements, particularly in indoor environments, leading to high false alarm rates due to the presence of pets, robotic vacuum cleaners, and electrical appliances, which limits their practical adoption.
Innovation Solution
A wireless sensing system based on deep learning that utilizes a transmitter and receiver to process a time series of channel information through autocorrelation functions, extracting motion features from wireless multipath channels to distinguish between human and non-human motions using a Hidden Markov Model (HMM)-based state machine, suitable for edge devices with limited resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wireless sensing systems use traditional signal processing methods, then device complexity is reduced, but measurement precision deteriorates leading to high false alarm rates
Solution Approach 1:
The patent segments the motion detection task into multiple components: channel impulse response estimation, autocorrelation function computation, and deep learning classification. This segmentation allows each component to be optimized independently, improving overall precision while managing complexity through modular processing stages.
Solution Approach 2:
The patent introduces an intermediary processing layer using autocorrelation functions and feature extraction modules between the raw wireless signals and the final classification. This intermediary layer transforms complex raw data into meaningful motion features, enabling accurate differentiation between human and non-human motions without requiring overly complex end-to-end systems.
2Productivity
If wireless sensing systems operate in real-time on edge devices, then productivity is improved, but measurement precision deteriorates due to resource constraints
Solution Approach 1:
The patent performs preliminary signal processing actions including channel impulse response estimation and autocorrelation function computation before the final classification stage. These preliminary actions prepare the data in an optimized format that reduces computational burden during real-time classification, enabling accurate motion recognition on edge devices with limited resources.
Solution Approach 2:
The patent dynamically adjusts processing parameters such as the number of autocorrelation functions computed, feature extraction depth, and classification model complexity based on device capabilities and environmental conditions. This adaptive parameter adjustment maintains measurement precision while optimizing productivity for real-time operation on constrained edge hardware.
3Reliability
If wireless sensing systems differentiate between human and non-human motions, then reliability is improved, but device complexity increases
Solution Approach 1:
The patent employs self-service mechanisms where the system automatically learns and adapts to different motion patterns through deep learning models trained on diverse datasets. The autocorrelation function computation automatically extracts distinctive temporal patterns without requiring manual feature engineering, enabling reliable human-nonhuman differentiation while keeping the implementation relatively simple.
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
The system achieves high accuracy and low false alarm rates in identifying human and non-human subjects, including pets and appliances, in various environments without requiring additional training or environmental restrictions, making it suitable for real-time operation on edge devices.
Implementation Method 1
a wireless signal transmitted from a first device to a second device through a wireless multipath channel of a venue
Data Source
AI summary
Methods, apparatus and systems for wireless sensing based on deep learning are described. For example, a described method comprises: transmitting a wireless signal through a wireless multipath channel of a venue, wherein the wireless multipath channel is impacted by a motion of an object in the venue; receiving the wireless signal through the wireless multipath channel of the venue, wherein the received wireless signal differs from the transmitted wireless signal due to the wireless multipath channel and the motion of the object; obtaining a time series of channel information (TSCI) of the wireless multipath channel based on the received wireless signal; computing a plurality of autocorrelation functions based on the TSCI, each autocorrelation function (ACF) computed based on CI of the TSCI in a respective sliding time window; constructing at least one ACF vector, wherein each respective ACF vector is a vector associated with a respective ACF comprising multiple vector elements each associated with a respective time lag, each vector element being a value of the respective ACF evaluated at the respective time lag; rearranging the at least one ACF vector into rearranged ACF data, wherein each ACF vector is a one-dimensional (1D) ACF-block; and performing a wireless sensing task based on a task engine to do a processing using the rearranged ACF data as an input.