Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Time Series Anomaly Detection with CNN for Environmental Sensors in Healthcare-IoT

Mirza Akhi Khatun, Mangolika Bhattacharya, Ciarán Eising, Lubna Luxmi Dhirani

TL;DR

This work addresses the need for reliable anomaly detection in healthcare-IoT environmental sensor time series, where cyber threats can compromise patient safety. It proposes a time-series CNN approach tailored to healthcare-IoT, validated in a Cooja/Contiki simulation with the WSN_DDoS_Attack_H-IoT2023 dataset. The key contributions include a dataset for environmental sensor anomalies, a 1D CNN architecture that achieves 92% accuracy and faster training compared with SVM and ensembles, and a methodology that preserves temporal structure during data splitting. The findings support the practical impact of CNN-based anomaly detection for improving data integrity and patient safety in healthcare facilities.

Abstract

This research develops a new method to detect anomalies in time series data using Convolutional Neural Networks (CNNs) in healthcare-IoT. The proposed method creates a Distributed Denial of Service (DDoS) attack using an IoT network simulator, Cooja, which emulates environmental sensors such as temperature and humidity. CNNs detect anomalies in time series data, resulting in a 92\% accuracy in identifying possible attacks.

Time Series Anomaly Detection with CNN for Environmental Sensors in Healthcare-IoT

TL;DR

This work addresses the need for reliable anomaly detection in healthcare-IoT environmental sensor time series, where cyber threats can compromise patient safety. It proposes a time-series CNN approach tailored to healthcare-IoT, validated in a Cooja/Contiki simulation with the WSN_DDoS_Attack_H-IoT2023 dataset. The key contributions include a dataset for environmental sensor anomalies, a 1D CNN architecture that achieves 92% accuracy and faster training compared with SVM and ensembles, and a methodology that preserves temporal structure during data splitting. The findings support the practical impact of CNN-based anomaly detection for improving data integrity and patient safety in healthcare facilities.

Abstract

This research develops a new method to detect anomalies in time series data using Convolutional Neural Networks (CNNs) in healthcare-IoT. The proposed method creates a Distributed Denial of Service (DDoS) attack using an IoT network simulator, Cooja, which emulates environmental sensors such as temperature and humidity. CNNs detect anomalies in time series data, resulting in a 92\% accuracy in identifying possible attacks.
Paper Structure (9 sections, 3 equations, 1 figure, 1 table)

This paper contains 9 sections, 3 equations, 1 figure, 1 table.

Table of Contents

  1. Introduction
  2. Methodology
  3. Data Acquisition
  4. Data Preprocessing
  5. CNN Model Architecture
  6. Experimental Setup
  7. Metrics for Performance Evaluation
  8. Convolutional Neural Network Classifier
  9. Conclusion

Figures (1)

  • Figure 1: Comparative Performance: Accuracy vs. Error Rate