Research on Human Activity Recognition Methods Based on Wi-Fi Channel State Information

DOI

10.2991/978-94-6463-823-3_7How to use a DOI?

Keywords

Wi-Fi Channel State Information (CSI); Human Activity Recognition; Traditional Machine Learning; Self-Supervised Learning

Abstract

Human activity recognition based on Wi-Fi Channel State Information (CSI) has gained attention in smart sensing due to its non-invasive and widespread applicability. This paper reviews four main approaches. First, traditional machine learning and signal processing use hand-crafted time-frequency features (like DTW or PCA) with classifiers such as SVM or KNN. While interpretable, they rely heavily on manual design and lack generalization. Second, deep learning methods, including CNN and LSTM, process CSI spectrograms or time-series data end-to-end, achieving high accuracy in complex scenarios like healthcare monitoring. However, they face challenges with model complexity and costly data labeling. Third, emerging techniques like self-supervised learning tackle data scarcity, using pre-training and fine-tuning to improve adaptability across environments. Fourth, multi-modal fusion and hardware optimization integrate diverse data (e.g., IMU) or tools like Nexmon, showing promise in multi-user settings through hardware-algorithm synergy. This paper explores key aspects. By highlighting four future directions—merging lightweight models with unsupervised learning, creating algorithms for dynamic settings, promoting multi - modal data fusion, and enhancing privacy and device compatibility—it offers valuable insights. It examines method evolution, technical hurdles, and real - world issues, providing theoretical and practical guidance for the field’s growth. These findings are expected to spark further research, driving the field forward.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Xufeng Zhang
PY  - 2025
DA  - 2025/08/31
TI  - Research on Human Activity Recognition Methods Based on Wi-Fi Channel State Information
BT  - Proceedings of the 2025 3rd International Conference on Image, Algorithms, and Artificial Intelligence (ICIAAI 2025)
PB  - Atlantis Press
SP  - 67
EP  - 83
SN  - 2352-538X
UR  - https://doi.org/10.2991/978-94-6463-823-3_7
DO  - 10.2991/978-94-6463-823-3_7
ID  - Zhang2025
ER  -