Privacy-Enhanced Data Fusion for COVID-19 Applications in Intelligent Internet of Medical Things

Hui Lin, Sahil Garg, Jia Hu, Xiaoding Wang, Md. Jalil Piran, M. Shamim Hossain

Abstract

With the worldwide large-scale outbreak of COVID-19, the Internet of Medical Things (IoMT), as a new type of Internet of Things (IoT)-based intelligent medical system, is being used for COVID-19 prevention and detection. However, since the widespread use of IoMT will generate a large amount of sensitive information related to patients, it is becoming more and more important yet challenging to ensure data security and privacy of COVID-19 applications in IoMT. The leakage of private information during IoMT data fusion process will cause serious problems and affect people's willingness to contribute data in IoMT. To address these challenges, this article proposes a new privacy-enhanced data fusion strategy (PDFS). The proposed PDFS consists of four important components, i.e., sensitive task classification, task completion assessment, incentive mechanism-based task contract design, and homomorphic encryption-based data fusion. The extensive simulation experiments demonstrate that PDFS can achieve high task classification accuracy, task completion rate, task data reliability and task participation rate, and low average error rate, while improving the privacy protection for data fusion under COVID-19 application environments based on IoMT.

Introduction: The IoMT Privacy Dilemma

The COVID-19 pandemic highlighted a critical bottleneck in modern healthcare: the need for massive, real-time data collection versus the absolute necessity of patient privacy. The Internet of Medical Things (IoMT)—a network of wearable sensors, smart thermometers, and medical monitors—can provide the data needed for disease tracking and vaccine development.

The Internet of Medical Things (IoMT) relies on a variety of body sensors and wearable devices to collect health data.

However, "data fusion" (the process of integrating data from multiple sources) is a privacy minefield. If patients fear their identity or health status will be leaked, they won't participate. Furthermore, traditional cloud-based IoMT architectures often suffer from high latency. This paper tackles these issues by proposing a Multi-access Edge Computing (MEC) architecture that processes data closer to the source while using advanced cryptography and AI to keep that data anonymous and trustworthy.

Methodology: The PDFS Framework

The Privacy-enhanced Data Fusion Strategy (PDFS) is built on four distinct pillars designed to balance utility and security:

  1. Sensitive Task Classification: The authors developed a privacy-preserving version of the K-means algorithm. By using "fake centers" and data perturbation, the system can categorize tasks by security level without the Data Fusion Center (DFC) ever seeing the raw, sensitive task details.
  2. Deep Reinforcement Learning (DDPG) for Incentives: Setting the right "price" for data is complex. The authors use the Deep Deterministic Policy Gradient (DDPG) algorithm to dynamically adjust payments to participants. Unlike older methods that work in discrete steps, DDPG operates in a continuous action space, allowing for much finer control over reward mechanisms.
  3. Task Completion Assessment: DDPG is also used to determine data reliability. It sets an optimal threshold to distinguish between honest participants and malicious actors providing "fake" or low-quality data, minimizing both false alarms and missed detections.
  4. Homomorphic Encryption: To protect identities, the system employs the Paillier cryptosystem. This allows the fusion center to perform mathematical operations on encrypted data so that the final recipient can verify the data's source without ever knowing the participant's true identity.

Evaluation and Results

The authors evaluated their strategy using Haberman’s Survival Data Set as a high-quality proxy for sensitive medical data. The simulation results indicate that PDFS significantly outperforms baseline strategies:

  • Reliability: PDFS achieved a data reliability rate of 88%, compared to roughly 58-60% for baseline methods.
  • Participation: Because the incentive mechanism is dynamic and the privacy protections are robust, the participation rate reached 93%.
  • Error Rates: The False Alarm Rate (FAR) for data validation was only 6% on average, significantly lower than DQN (28%) or standard Q-learning (34%).
Secure data fusion is vital for sensitive medical tasks like vaccine administration and pandemic tracking.

Conclusion

The PDFS framework demonstrates that privacy does not have to come at the cost of accuracy. By moving the fusion process to MEC servers at the "Edge," the authors address the latency issues of the cloud, which is vital for real-time pandemic monitoring. Future work will focus on addressing sophisticated multi-party collusion attacks and optimizing the computational overhead of homomorphic encryption for low-power wearable devices.

Summary updated

Abstract

With the worldwide large-scale outbreak of COVID-19, the Internet of Medical Things (IoMT), as a new type of Internet of Things (IoT)-based intelligent medical system, is being used for COVID-19 prevention and detection. However, since the widespread use of IoMT will generate a large amount of sensitive information related to patients, it is becoming more and more important yet challenging to ensure data security and privacy of COVID-19 applications in IoMT. The leakage of private information during IoMT data fusion process will cause serious problems and affect people's willingness to contribute data in IoMT. To address these challenges, this article proposes a new privacy-enhanced data fusion strategy (PDFS). The proposed PDFS consists of four important components, i.e., sensitive task classification, task completion assessment, incentive mechanism-based task contract design, and homomorphic encryption-based data fusion. The extensive simulation experiments demonstrate that PDFS can achieve high task classification accuracy, task completion rate, task data reliability and task participation rate, and low average error rate, while improving the privacy protection for data fusion under COVID-19 application environments based on IoMT.
BibTeX
@article{lin2020privacy,
  author = {Hui Lin and Sahil Garg and Jia Hu and Xiaoding Wang and Md. Jalil Piran and M. Shamim Hossain},
  title = {{Privacy-Enhanced Data Fusion for COVID-19 Applications in Intelligent Internet of Medical Things}},
  journal = {IEEE Internet of Things Journal},
  year = {2020},
  doi = {10.1109/jiot.2020.3033129},
}
Figure