Hua Guo

The fog computing paradigm enables mobile users to seamlessly interact with crowds of nearby IoT devices for low-latency services. However, existing authentication and key agreement (AKA) protocols suffer from critical limitations in this mobile context. While existing AKA schemes are optimized for mobile edge environments, they incur substantial overhead in multi-device scenarios: they require repeated authentication for each device and generate distinct session keys per user-device pair, leading to high key management complexity at scale. Furthermore, traditional one-to-many protocols are unsuitable for distributed fog environments due to their reliance on trusted central nodes (e.g., gateways or servers) and incompatible communication models. To address these issues, this paper proposes an efficient and secure authentication protocol for fog computing that integrates elliptic curve cryptography with secret sharing. Our solution enables a user to authenticate an entire group of devices managed by a semi-trusted fog node (honest-but-curious) through a single protocol execution, effectively eliminating the authentication bottleneck caused by scaling devices. Formal security proof under the Real-Or-Random (ROR) model is provided, along with informal analysis, demonstrating the protocol ensures forward secrecy, user anonymity, and resistance to a comprehensive set of attacks, including ephemeral secret leakage, key compromise impersonation, and replay attacks. Performance analysis confirms the scheme maintains low communication and computational overhead while achieving comprehensive security.

In this letter, we identify critical vulnerabilities in both protocols proposed by Zhao et al. (published in IEEE TIFS, doi: 10.1109/TIFS.2025.3574993), showing that they are susceptible to collusion attacks. In the first protocol, the leader P₁ can determine whether its items are held by any honest party by colluding with other participants. In the second protocol, P₁ can infer whether all honest parties hold a specific item without any collusion, violating the fundamental privacy guarantees of multi-party private set intersection.

The increasing complexity of the smart grid (SG) raises significant concerns regarding the security of SG communication. As a countermeasure, the authentication and key agreement (AKA) protocol ensures the secure transmission of sensitive information between legitimate entities by achieving mutual authentication and establishing session keys. One of the most urgent and critical security threats in the AKA protocol concerns the ephemeral secret leakage (ESL) attack, due to its threat to session key secrecy. However, there remains a lack of systematic understanding of how to resist ESL attacks in the SG environment. Therefore, we categorize the ESL attack into three different types, then conduct an in-depth analysis of their root causes and propose corresponding recommendations to mitigate it. To further illustrate the effectiveness of the recommendations, we design a secure and efficient AKA protocol based on elliptic curve cryptography accordingly. The proposed protocol is proven secure through rigorous security proof under the random oracle model and formally verified by the AVISPA tool. Performance comparisons indicate that the proposed protocol outperforms other related protocols due to its lightweight nature and adherence to all fundamental security attributes, making it well-suited for deployment in a resource-constrained SG environment.

Transformer has been successfully used in practical applications due to its powerful advantages. However, users' input is leaked to the model provider during the service. With people's attention to privacy, privacy-preserving Transformer inference is on the demand of such services. Secure protocols for non-linear functions are crucial in privacy-preserving Transformer inference, which are not well studied. Thus, designing practical secure protocols for non-linear functions is hard but significant to model performance. In this work, we propose a framework East to enable efficient and accurate secure Transformer inference. Firstly, we propose a new oblivious piecewise polynomial evaluation algorithm and apply it to the activation functions, which reduces the runtime and communication of GELU by over 1.5× and 2.5×, compared to prior arts. Secondly, the secure protocols for softmax and layer normalization are carefully designed to faithfully maintain the desired functionality. Thirdly, several optimizations are conducted in detail to enhance the overall efficiency. We applied East to BERT and the results show that the inference accuracy remains consistent with the plaintext inference without fine-tuning. Compared to Iron, we achieve about 1.8× lower communication within 1.2× lower runtime.