International Conference on Computer and Knowledge Engineering

Home / 11th International Conference on Computer and Knowledge Engineering

CSI-Based Human Activity Recognition using Convolutional Neural Networks

Authors :

Parisa Fard Moshiri¹ Mohammad Nabati² Reza Shahbazian³ Seyed Ali Ghorashi⁴

1- Cognitive Telecommunication Research Group, Department of Electrical Engineering, Shahid Beheshti University G. C., Tehran, Iran 2- Cognitive Telecommunication Research Group, Department of Electrical Engineering, Shahid Beheshti University G. C., Tehran, Iran 3- Department of Electrical Engineering, Faculty of Technology and Engineering, Standard Research Institute, Alborz 31745-139, Iran 4- Department of Computer Science & Digital Technologies, School of Architecture, Computing, and Engineering, University of East London, London, UK

Keywords :

Activity Recognition, Channel State Information, Convolutional Neural Network, Deep Learning, WiFi

Abstract :

Human activity recognition (HAR) as an emerging technology can have undeniable impacts on several applications such as health monitoring, context-aware systems, transportation, robotics, and smart cities. Among the main research methods in HAR (sensor, image, and WiFi-based), the WiFi-based method has attracted more attention due to the ubiquity of WiFi devices. WiFi devices can be utilized to recognize daily human activities such as running, walking, and sleeping. These activities affect WiFi signal propagation and can be further used to identify human activities. This paper proposes a Deep Learning (DL) method for activity recognition tasks using WiFi channel state information (CSI). A new model is developed in which CSI data are converted to grayscale images. These images are then fed into a Convolutional Neural Network (CNN) with 2-dimensional convolutional layers for activity recognition. We take advantage of CNN's high accuracy on image classification along with WiFi-based preponderance. The experimental results demonstrate that our proposed approach achieves acceptable performance in HAR tasks.