Lightweight and Privacy-Preserving Medical Services Access for Healthcare Cloud

Jingwei Liu, Huifang Tang, Rong Sun, Xiaojiang Du, Mohsen GuizaniIEEE Access 2019 · DOI

Abstract

Cloud computing is revolutionizing healthcare by enabling remote diagnosis and resource sharing, but it introduces significant privacy risks. Traditional Attribute-Based Signature (ABS) schemes provide anonymity but are computationally intensive for mobile devices. This paper presents LPP-MSA, a lightweight and privacy-preserving medical services access protocol. By utilizing multi-authority attribute management, online/offline signing, and server-aided verification, LPP-MSA reduces the computational burden on mobile devices by over 95% compared to existing methods, achieving constant-time performance regardless of attribute complexity.

Introduction

The integration of cloud computing into healthcare allows for seamless data sharing among hospitals, community centers, and health departments. However, ensuring that only authorized users access sensitive medical services without compromising their identity remains a major challenge. While Attribute-Based Signature (ABS) is the standard cryptographic solution, its high computational cost often drains mobile batteries and causes significant latency.

Problem Statement

Traditional ABS requires complex mathematical operations, such as bilinear pairings and exponentiations, which are difficult for resource-constrained devices like smartphones to handle efficiently. Furthermore, in practical healthcare ecosystems, attributes (e.g., "Doctor," "Hospital A," "Researcher") are often issued by different, independent authorities. Most existing schemes struggle to handle these multi-authority scenarios without becoming prohibitively slow.

General architecture of a healthcare cloud system involving various stakeholders and devices.

Methods

The researchers introduced LPP-MSA (Lightweight and Privacy-Preserving Medical Services Access), which employs three core strategies to achieve high security with low overhead:

  1. Multi-Authority ABS: Unlike single-authority systems, this scheme allows different organizations to manage different attributes independently, making it more practical for complex healthcare ecosystems.
  2. Online/Offline Signing: The "heavy lifting" of the cryptographic signature is performed "offline" while the device is idle. When the user requests a service, the "online" phase requires only a few simple operations, making the process nearly instantaneous.
  3. Server-Aided Verification: Verification is typically the most expensive part of the process for the Medical Service Provider (MSP). LPP-MSA outsources the bulk of this math to a cloud server. A "transformation" step ensures the cloud server can assist with the computation without ever seeing sensitive data or being able to forge a signature.
System model of the LPP-MSA protocol illustrating the interaction between the requester, cloud, and authorities.

Evaluation and Key Results

The protocol was evaluated using a simulation platform (Linux Ubuntu 16.04, Intel Core i5-4200U CPU) and the pbc-0.5.14 library for pairing-based cryptography. The results demonstrate significant performance gains over existing schemes like EABS and DMA-ABS:

  • Constant-Time Performance: A major breakthrough of LPP-MSA is that computational costs for both signing and verification remain fixed. In older schemes, the time required increased linearly with the number of attributes.
  • Signing Speed: While competing schemes can take up to 442.8 ms for 10 attributes, LPP-MSA completes online signing in a constant 5.136 ms.
  • Verification Speed: For the Medical Service Provider, verification time is reduced to a constant 3.885 ms, compared to over 250 ms in alternative methods.
  • Security: The paper proves the scheme is resistant to collusion, even if an attacker collaborates with the helper cloud server.

Conclusion and Future Work

LPP-MSA effectively shifts the burden of proof from mobile devices to the cloud, making it an ideal solution for the Internet of Medical Things (IoMT). While the scheme is highly efficient, the authors note that the signature size is slightly larger than some less secure alternatives. Future research will focus on optimizing storage overhead and developing specialized protocols for attribute revocation in multi-authority settings.

Summary updated

BibTeX
@article{liu2019lightweight,
  author = {Jingwei Liu and Huifang Tang and Rong Sun and Xiaojiang Du and Mohsen Guizani},
  title = {{Lightweight and Privacy-Preserving Medical Services Access for Healthcare Cloud}},
  journal = {IEEE Access},
  year = {2019},
  doi = {10.1109/access.2019.2931917},
}
Figure