Resource efficient data compression algorithms for demanding, WSN based biomedical applications.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Medical Informatics

Background:

• Wireless Sensor Networks (WSNs) are crucial for advanced medical research, including epilepsy monitoring.
• Limited memory and bandwidth in WSNs hinder the analysis of critical physiological data.
• Existing data compression methods lack comprehensive evaluation for specialized medical applications.

Purpose of the Study:

• To implement and experimentally evaluate prominent and novel data compression algorithms for WSNs.
• To assess algorithm performance using real-world Electroencephalography (EEG) and Electrocardiography (ECG) datasets for epilepsy monitoring.
• To identify compression techniques offering an optimal trade-off between compression rate and execution latency.

Main Methods:

• Implementation of various compression algorithms within a unified Matlab framework.
• Experimental evaluation using real-world EEG and ECG datasets, focusing on epilepsy-related data.
• Analysis of key performance metrics: compression rate and execution latency.

Main Results:

• Algorithm complexity significantly impacts execution latency versus compression rate.
• Proposed compression schemes demonstrate a considerable advantage in achieving an optimal compression rate-latency trade-off.
• A specific proposed algorithm achieved high compression with minimal latency, indicating real-time capability.

Conclusions:

• Data compression is essential for overcoming WSN limitations in demanding medical applications like epilepsy monitoring.
• The evaluated algorithms exhibit distinct performance characteristics crucial for WSN deployment.
• Novel proposed schemes offer significant improvements, particularly in achieving real-time processing for critical physiological data analysis.