A Secure and Robust ECG Signal Transmission System for Cardiac Arrhythmia Identification

This paper proposes a dynamic encryption method for cardiac arrhythmia identification based on electrocardiogram (ECG) signal classification, which is collected from a Wireless Body Area Network (WBAN). We extract five features from collected ECG signals to modify the accuracy of the K Nearest Neighbor (KNN) classifier. KNN gives better performance with the features compared with other classifiers. From our proposed dynamic encryption method in this paper, time-varying encryption keys are provided for the security of consecutive ECG data frames, and Alamouti-based cooperative relays are applied for transmission robustness under noisy transmission channels.