Secure and Privacy-Preserving Body Sensor Data Collection and Query Scheme

Hui Zhu, Lijuan Gao, Hui Li

TL;DR

This paper introduces SPCQ, a novel scheme for secure and private collection and querying of body sensor data. It leverages an improved homomorphic encryption technique and a specialized Weighted Euclidean Distance Contrast (WEDC) algorithm to ensure data confidentiality and query privacy while maintaining high accuracy, even when data is processed on an untrusted cloud server.

What Problem Does This Paper Address?

The proliferation of wearable sensors and smartphones has led to the collection of vast amounts of personal health data. While this data holds immense potential for public health insights and scientific discovery, its sensitive nature raises significant privacy concerns. Storing and processing this data on cloud servers, which is becoming increasingly common due to the computational and storage demands, introduces further risks. Existing solutions like differential privacy can compromise accuracy, while traditional homomorphic encryption and searchable encryption have limitations in handling multi-dimensional data queries efficiently and privately. This paper specifically addresses the challenges of securing body sensor data collection and enabling accurate, private data queries in an outsourced cloud computing environment, focusing on threats from telemetry interfaces.

Methods

The core of the SPCQ scheme lies in its innovative approach to data handling and querying:

  • Secure Data Collection: Sensor users (SUs) collect multi-dimensional personal data. Before uploading to the cloud server (CS), this data is first "perturbed" by adding a hash of a secret base ($B_1$) to each dimension. This increases the sample space, making exhaustive attacks more difficult. Subsequently, the perturbed data is transformed into an encrypted search index using a combination of secret bases ($B_1$, $B_2$), weights ($W$), and random numbers, ensuring that the CS cannot directly access the raw personal data. This encrypted index is then signed by the SU for authentication.
  • Privacy-Preserving Query Service: Data users (DUs) generate encrypted queries. Their query parameters are also perturbed by adding the hash of $B_1$. The query condition is then encrypted using secret bases and weights, and the entire query is further encrypted using the CS's public key. This ensures that the CS only sees encrypted data and queries.
  • Weighted Euclidean Distance Contrast (WEDC) Algorithm: This is the key algorithmic contribution. When the CS receives an encrypted query, it uses its private key to decrypt the query parameters. It then applies the WEDC algorithm to compare the encrypted data items with the encrypted query. This algorithm is designed to compute the weighted Euclidean distance between encrypted multi-dimensional vectors without decrypting them. The algorithm leverages properties of bilinear pairings and homomorphic encryption to perform these computations efficiently.
  • Outsourced Computing: The heavy lifting of data processing and query matching is offloaded to the cloud server, which is assumed to be "honest-but-curious" – it follows instructions but might try to infer information from the data it processes. The scheme ensures that even with this assumption, neither the SU's personal data nor the DU's query information is revealed.

Datasets

The paper mentions using a real medical database for evaluation:

  • Diabetes Database: A dataset containing 100,000 items with 55 attributes was used to simulate a real-world scenario. This database was used to populate the cloud server's storage.

Key Insights and Ideas

  • Perturbation for Enhanced Privacy: The simple yet effective technique of adding a hash of a secret base to data dimensions before encryption significantly enhances privacy by making direct analysis of raw data by the cloud server infeasible.
  • WEDC for Encrypted Multi-Dimensional Queries: The development of a WEDC algorithm that operates directly on encrypted multi-dimensional data is a crucial advancement. This allows for accurate distance-based queries without compromising the privacy of either the data owner or the data requester.
  • Balancing Security and Efficiency: The scheme demonstrates that it's possible to achieve strong security and privacy guarantees (confidentiality, privacy, authentication) without incurring prohibitive computational or communication costs, making it suitable for resource-constrained devices like smartphones.
  • Leveraging Composite Order Groups: The use of improved homomorphic encryption over composite order groups is key to enabling efficient computations on encrypted data, which is often a bottleneck in privacy-preserving systems.

Key Results

  • Security Guarantees: The SPCQ scheme is shown to provide confidentiality of SU's personal data, privacy of DU's query information, and source authentication for data, queries, and responses. It is designed to resist various security threats, including exhaustive attacks.
  • Accuracy: The WEDC algorithm ensures accurate query results, with the paper demonstrating that the scheme can correctly identify data items satisfying the query condition.
  • Efficiency:
    • Computational Costs: The paper presents a comparison with a privacy-preserving range query scheme (PPRQ). For SU, DU, and CS, SPCQ shows significantly lower computational costs. For instance, DU's computational cost is 2n * Ce, while PPRQ's is 4m * Ce + 2m * Ce. CS's cost for SPCQ is nN * Cp + 2nN * Cm, which is substantially less than PPRQ's 23mnN * Ce + 23mnN * Cm.
  • Communication Costs: SPCQ also demonstrates more efficient communication. DU's communication length is 164 * n bytes (vs. 512 * n bytes for PPRQ), and CS's is 256 bytes (vs. 1024 * N bytes for PPRQ).
  • Real-world Implementation: The scheme was implemented on an embedded device, smartphone, and laptop, demonstrating its feasibility and efficiency in a practical setting. The integrated performance in a real environment showed query response times of about five seconds, which is deemed acceptable.

Limitations and Future Work

  • Collusion Attack: The current security model does not consider collusion attacks where multiple parties might conspire to disclose a third party's privacy. The authors suggest this as a direction for future research.
  • Scalability with Large Datasets: While the paper demonstrates efficiency with 100,000 data items, the performance with significantly larger datasets might require further investigation.
  • Complexity of WEDC: While efficient, the WEDC algorithm's complexity is still dependent on the number of dimensions and data items, which could be a factor in extremely large-scale deployments.

The paper concludes by emphasizing that SPCQ offers a secure, privacy-preserving, and efficient solution for body sensor data collection and querying, suitable for outsourced computing environments.

Summary updated

Abstract

With the development of body sensor networks and the pervasiveness of smart phones, different types of personal data can be collected in real time by body sensors, and the potential value of massive personal data has attracted considerable interest recently. However, the privacy issues of sensitive personal data are still challenging today. Aiming at these challenges, in this paper, we focus on the threats from telemetry interface and present a secure and privacy-preserving body sensor data collection and query scheme, named SPCQ, for outsourced computing. In the proposed SPCQ scheme, users’ personal information is collected by body sensors in different types and converted into multi-dimension data, and each dimension is converted into the form of a number and uploaded to the cloud server, which provides a secure, efficient and accurate data query service, while the privacy of sensitive personal information and users’ query data is guaranteed. Specifically, based on an improved homomorphic encryption technology over composite order group, we propose a special weighted Euclidean distance contrast algorithm (WEDC) for multi-dimension vectors over encrypted data. With the SPCQ scheme, the confidentiality of sensitive personal data, the privacy of data users’ queries and accurate query service can be achieved in the cloud server. Detailed analysis shows that SPCQ can resist various security threats from telemetry interface. In addition, we also implement SPCQ on an embedded device, smart phone and laptop with a real medical database, and extensive simulation results demonstrate that our proposed SPCQ scheme is highly efficient in terms of computation and communication costs.
BibTeX
@inproceedings{zhu2016secure,
  author = {Hui Zhu and Lijuan Gao and Hui Li},
  title = {{Secure and Privacy-Preserving Body Sensor Data Collection and Query Scheme}},
  booktitle = {Sensors},
  year = {2016},
  url = {https://openreview.net/forum?id=2qfVzmXXl5},
}
Figure