[Paper Review] Decentralized Identifier-based Privacy-preserving Authenticated Key Exchange Protocol for Electric Vehicle Charging in Smart Grid

Rohini Poolat Parameswarath, Prosanta Gope+1 more|arXiv (Cornell University)|Jun 27, 2022

Advanced Authentication Protocols Security4 citations

TL;DR

This paper proposes a decentralized identifier (DID)-based, privacy-preserving authenticated key exchange protocol for electric vehicle (EV) charging in smart grids, leveraging Verifiable Credentials (VCs) and zero-knowledge proofs (ZKPs) to enable user-empowered identity control and mutual authentication without exposing real identities. The protocol ensures strong security properties, including unlinkability, resistance to replay and location forgery attacks, and a secure key recovery mechanism, while maintaining low computational overhead suitable for real-world deployment.

ABSTRACT

The popularity of Electric Vehicles (EVs) has been rising across the globe in recent years. Smart grids will be the backbone for EV charging and enable efficient consumption of electricity by the EVs. With the demand for EVs, associated cyber threats are also increasing. Users expose their personal information while charging their EVs, leading to privacy threats. This paper proposes a user-empowered, privacy-aware authenticated key exchange protocol for EV charging in smart grid. The proposed protocol is based on the concept of Decentralized Identifier (DID) and Verifiable Credentials (VCs). The use of DIDs empowers users by helping them to have complete control over their identities. The charging station and the user verify that the other party is legitimate before proceeding with the charging services using VC. Key recovery is another issue we address in this paper. A method to recover lost keys is incorporated into the proposed protocol. We present formal security proof and informal analysis to show that protocol's robustness against several attacks. We also provide a detailed performance analysis to show that the proposed protocol is efficient.

Motivation & Objective

• To address growing privacy and security threats in EV charging within smart grids, especially tracking and identity exposure.
• To empower users with full control over their digital identities through Decentralized Identifiers (DIDs), eliminating reliance on centralized authorities.
• To ensure mutual authentication and privacy during EV charging using Verifiable Credentials (VCs) and zero-knowledge proofs (ZKPs).
• To provide a secure and efficient key recovery mechanism in case of private key loss.
• To achieve comprehensive security properties—such as unlinkability, resistance to replay and location forgery attacks—while maintaining practical performance.

Proposed method

• The protocol uses Decentralized Identifiers (DIDs) to allow users to self-manage their identities without central authority assignment.
• Verifiable Credentials (VCs) issued by trusted parties are used to cryptographically bind user attributes and enable third-party verification of legitimacy.
• Zero-knowledge proofs (ZKPs) are employed to prove possession of a valid VC without revealing any underlying identity or credential details.
• Mutual authentication is achieved through a three-message key exchange protocol where both the EV and charging station (CS) verify each other’s credentials using VCs and ZKPs.
• A key recovery mechanism is integrated to allow users to regain access if their private key is lost, ensuring resilience against accidental key loss.
• Security is formally proven using the ProVerif tool, and performance is evaluated through computation cost analysis on standard hardware.

Figure 3: DID and Verifiable Credential.

Experimental results

Research questions

• RQ1 How can EV users be empowered to control their digital identities in smart grid charging systems without relying on centralized authorities?
• RQ2 To what extent can a DID- and VC-based protocol ensure user anonymity and unlinkability in EV charging transactions?
• RQ3 Can a ZKP-based authentication mechanism prevent the charging station from learning sensitive user information while still enabling mutual authentication?
• RQ4 How does the proposed protocol resist common attacks such as replay, location forgery, and impersonation?
• RQ5 What is the computational overhead of the proposed protocol, and how does it compare to existing schemes in terms of efficiency and scalability?

Key findings

• The proposed protocol achieves strong privacy guarantees, including unlinkability and anonymity, by using ZKPs to hide user identities during authentication.
• The protocol is formally proven secure against common threats such as replay attacks, impersonation, and key compromise impersonation.
• The protocol resists location forgery attacks, a critical vulnerability in many existing schemes, by binding location data to verifiable credentials.
• Performance evaluation shows that the total computation cost per authentication is within acceptable limits, with the USP’s cost slightly higher due to VC generation, but still feasible given its computational resources.
• The protocol supports non-repudiation and includes a secure key recovery mechanism, enhancing resilience to private key loss.
• Compared to existing schemes, the proposed protocol uniquely combines user empowerment via DIDs with comprehensive security and privacy features at a reasonable computational cost.

Figure 4: System model.