ENERGY-EFFICIENT ECG DATA SECURITY USING SIDECHAIN TECHNOLOGY

Abstract

Electrocardiogram (ECG) signals are classified into heart disease categories through the design of high-efficiency signal processing models. These models must be able to perform Pre-processing, Segmentation of signal, Extraction of Feature, Classification, and Post-Processing steps with high efficiency in terms of accuracy and power metrics. To attain this, researchers have proposed the use of reinforcement learning models, that assist in offloading computationally complex tasks on the cloud. Due to this offloading, there are communication vulnerabilities between the ECG sensing device and the cloud data center. To mitigate the different vulnerabilities, researchers have proposed the use of various data security mechanisms, which adds to the computational complexity of the underlying ECG model. This increases the delay needed for communication, thereby reducing the real-time performance capabilities. While maintaining high security, side blockchain-based systems were introduced, which possess characteristics like immutability, transparency, traceability, and distributed computing. But the delay needed for the main blockchain is directly proportional to the chain length, thus they are not suited for big data applications like ECG signal communication.  The proposed design of a novel side chain high security and QoS model for energy-aware ECG classification deployments to overcome these issues. While improving security performance making the model highly applicable for extensive variation of health real-time IoT data in which Delay, Energy, Throughput, and PDR are evaluated.

Keywords: ECG, Low Power, WOA, Optimization, Delay, Splitting, Merging, Security, Attacks