On the Security of an Efficient and Robust Certificateless Signature Scheme for IIoT Environments

Wenjie Yang, Shangpeng Wang, Xinyi Huang, Yi MuIEEE Access 2019 · DOI

Introduction

In the Industrial Internet of Things (IIoT), the integrity and authenticity of data transmitted from edge sensors to cloud infrastructure are paramount. To secure these communications without the administrative burden of traditional Public Key Infrastructure (PKI) or the key escrow risks of Identity-Based Cryptography (IBC), researchers often turn to Certificateless Signature (CLS) schemes.

High-level IIoT data flow between owner, cloud, and consumer.

Recently, a scheme proposed by Zhang et al. was presented as an "efficient and robust" solution tailored for IIoT. However, this paper provides a critical cryptanalysis of that scheme, demonstrating that it fails to provide the basic security guarantees required for industrial applications.

The Core Vulnerability: Public Key Replacement

The fundamental flaw identified is the scheme's inability to withstand a public key replacement attack. In a CLS environment, a user's public key is not bound by a certificate. A "Type I" adversary—one who does not know the system's master secret key but can replace a user's public key—can exploit this.

The authors demonstrate that by selecting arbitrary values to construct a fraudulent public key, an attacker can mathematically manipulate the signature components ($t$ and $\delta$). This manipulation ensures that the verification equation balances perfectly, even if the attacker possesses no knowledge of the legitimate user's secret key. This allows for the forgery of valid signatures on any message for any identity.

System Architecture and Cryptanalysis

The CLS architecture typically involves a Key Generation Center (KGC) that provides partial private keys, while users generate their own secret keys and public keys to complete the process.

Detailed architecture of the Certificateless Signature (CLS) system for IIoT, showing the interaction between the KGC, data owner, and cloud server.

The authors' methodology involved two primary steps:

  1. Attack Construction: A concrete mathematical procedure showing how an adversary can bypass the "Verify" algorithm by exploiting the degrees of freedom in the public key and signature values.
  2. Proof Dissection: A rigorous review of the original security proof. The authors identified four major logical "drawbacks" in Zhang et al.’s proof. They found that the original simulation—intended to link the scheme's security to the $q$-Strong Diffie-Hellman problem—was flawed. Specifically, the simulation would "abort" in the exact scenarios where an adversary succeeded, rendering the security proof invalid.

Key Insights and Results

  • Verification Degrees of Freedom: The core flaw lies in the verification equation. Because the attacker can choose both the public key and the signature elements, they can "cancel out" unknown variables.
  • Soundness Gap: The paper highlights a critical lesson for cryptographic research: a security proof is only as strong as its simulation. If the simulation cannot handle successful adversarial queries, it fails to prove security.
  • IIoT Impact: In a real-world IIoT deployment, this vulnerability would allow an attacker to spoof sensor data or impersonate industrial controllers, potentially leading to catastrophic physical or financial consequences.

Conclusion

The analysis concludes that the Zhang et al. scheme is fundamentally insecure and should not be deployed in IIoT environments. The work serves as a vital warning to the community and a reminder of the necessity for independent verification of cryptographic proofs and the ongoing challenge of designing efficient yet robust authentication protocols for industrial scale.

Summary updated

BibTeX
@article{yang2019security,
  author = {Wenjie Yang and Shangpeng Wang and Xinyi Huang and Yi Mu},
  title = {{On the Security of an Efficient and Robust Certificateless Signature Scheme for IIoT Environments}},
  journal = {IEEE Access},
  year = {2019},
  doi = {10.1109/access.2019.2927597},
}
Figure