A low cost quadratic level ECG compression algorithm and its hardware optimization for body sensor network system.

Area of Science:

• Biomedical Engineering
• Signal Processing
• Computer Engineering

Background:

• Body sensor network systems require efficient data compression to manage power consumption and transmission bandwidth.
• Existing compression algorithms often face trade-offs between compression ratio, signal quality, and hardware complexity.
• Low-cost, high-performance compression is crucial for widespread adoption of wearable health monitoring.

Purpose of the Study:

• To propose a novel, low-cost quadratic level compression algorithm for body sensor network systems.
• To reduce encoding delay and hardware costs associated with data compression.
• To maintain high reconstructed signal quality while achieving significant data compression.

Main Methods:

• Development of a quadratic compression algorithm utilizing mean deviation for level determination.
• Implementation of a 16-bit sensor node processor supporting the proposed algorithm.
• Evaluation of compression ratio (CR), peak residual distortion (PRD), and encoding rate.

Main Results:

• Achieved an overall compression ratio (CR) of 8.4:1.
• Maintained a peak residual distortion (PRD) of 0.897%, indicating high signal fidelity.
• Demonstrated an encoding rate of 6.4 Mbps.
• Designed a 16-bit sensor node processor with low power consumption (0.56 nJ/bit at 1 MHz).

Conclusions:

• The proposed quadratic compression algorithm offers an effective solution for body sensor networks.
• The algorithm successfully balances compression efficiency, signal quality, and reduced hardware requirements.
• The integrated sensor node processor enables low-power, high-performance data compression for wearable devices.