Differential Privacy Preserving in Big Data Analytics for Connected Health

Area of Science:

• Biomedical Engineering
• Data Science
• Cybersecurity

Background:

• Body Area Networks (BANs) generate large volumes of sensitive data from wearable sensors.
• Existing privacy protection methods for BANs are insufficient, leading to potential data exposure.
• The need for robust privacy solutions is critical for the secure utilization of health data.

Purpose of the Study:

• To develop a differential privacy protection scheme tailored for big data in Body Area Networks.
• To enhance the availability and reliability of privacy protection for sensitive sensor data.
• To address the limitations of previous methods in safeguarding user privacy.

Main Methods:

• Implementation of a differential privacy scheme specifically designed for Body Area Network big data.
• Introduction of dynamic noise thresholds to adapt privacy protection to varying data characteristics.
• Evaluation of the scheme's effectiveness against potential attackers with extensive background knowledge.

Main Results:

• The proposed scheme offers improved privacy protection compared to existing methods.
• Dynamic noise thresholds enhance the scheme's suitability for processing large datasets.
• Experimental results confirm the scheme's ability to preserve privacy even under sophisticated attacks.

Conclusions:

• The developed differential privacy scheme provides a reliable and available solution for protecting sensitive big data in Body Area Networks.
• The use of dynamic noise thresholds is a key innovation for effective big data privacy in BSNs.
• This research contributes to the secure advancement of wearable health technology and data utilization.