CSI-F: A Human Motion Recognition Method Based on Channel-State-Information Signal Feature Fusion

Juan Niu,Xiuqing He,Bei Fang,Guangxin Han,Xu Wang,Juhou He

DOI: https://doi.org/10.3390/s24030862

IF: 3.9

2024-01-30

Sensors

Abstract:The recognition of human activity is crucial as the Internet of Things (IoT) progresses toward future smart homes. Wi-Fi-based motion-recognition stands out due to its non-contact nature and widespread applicability. However, the channel state information (CSI) related to human movement in indoor environments changes with the direction of movement, which poses challenges for existing Wi-Fi movement-recognition methods. These challenges include limited directions of movement that can be detected, short detection distances, and inaccurate feature extraction, all of which significantly constrain the wide-scale application of Wi-Fi action-recognition. To address this issue, we propose a direction-independent CSI fusion and sharing model named CSI-F, one which combines Convolutional Neural Networks (CNN) and Gated Recurrent Units (GRU). Specifically, we have introduced a series of signal-processing techniques that utilize antenna diversity to eliminate random phase shifts, thereby removing noise influences unrelated to motion information. Later, by amplifying the Doppler frequency shift effect through cyclic actions and generating a spectrogram, we further enhance the impact of actions on CSI. To demonstrate the effectiveness of this method, we conducted experiments on datasets collected in natural environments. We confirmed that the superposition of periodic actions on CSI can improve the accuracy of the process. CSI-F can achieve higher recognition accuracy compared with other methods and a monitoring coverage of up to 6 m.

engineering, electrical & electronic,chemistry, analytical,instruments & instrumentation

What problem does this paper attempt to address?

The problems that this paper attempts to solve are several key challenges in using Channel State Information (CSI) in Wi - Fi signals for human activity recognition. Specifically, these challenges include:

1. **Direction - Dependence**: Existing Wi - Fi motion recognition methods can usually only detect motions in specific directions. When the user faces different directions, the robustness of recognition will decrease significantly.
2. **Limited Detection Distance**: The detection distance of most existing methods is relatively short, which limits their application scope in practical scenarios.
3. **Inaccurate Feature Extraction**: Due to the complexity of the indoor environment, such as multipath effects and noise interference, existing methods have inaccuracies in feature extraction, which affects the recognition accuracy.

To address these challenges, the author proposes a direction - independent model CSI - F based on CSI signal feature fusion. This model combines Convolutional Neural Network (CNN) and Gated Recurrent Unit (GRU), and uses a series of signal processing techniques to eliminate random phase offsets, amplify Doppler frequency offset effects, and generate spectrograms, thereby enhancing the influence of actions on CSI. Experimental results show that CSI - F can achieve higher recognition accuracy and a wider monitoring coverage range (up to 6 meters) in natural environments.