Encoding Distortion Modeling For DWT-Based Wireless EEG Monitoring System

TL;DR

This paper presents an analytical distortion model for DWT-based EEG encoding in wireless systems, enabling adaptive reconfiguration to balance power consumption and reconstruction accuracy amid channel variations.

Contribution

It introduces a novel distortion model for EEG encoding that allows dynamic adjustment of encoder parameters based on application constraints and channel conditions.

Findings

Main distortion factors are compression ratio and filter length.

Wireless channel variations significantly impact distortion levels.

The model helps optimize encoder settings for better performance.

Abstract

Recent advances in wireless body area sensor net- works leverage wireless and mobile communication technologies to facilitate development of innovative medical applications that can significantly enhance healthcare services and improve quality of life. Specifically, Electroencephalography (EEG)-based applications lie at the heart of these promising technologies. However, the design and operation of such applications is challenging. Power consumption requirements of the sensor nodes may turn some of these applications impractical. Hence, implementing efficient encoding schemes are essential to reduce power consumption in such applications. In this paper, we propose an analytical distortion model for the EEG-based encoding systems. Using this model, the encoder can effectively reconfigure its complexity by adjusting its control parameters to satisfy application constraints while maintaining reconstruction accuracy at the receiver side. The simulation results illustrate that the main parameters that affect the distortion are compression ratio and filter length of the considered DWT-based encoder. Furthermore, it is found that the wireless channel variations have a significant influence on the estimated distortion at the receiver side.