Proving resistance to conventional attacks, e.g., differential, linear, and integral attacks, is essential for designing a secure symmetric-key cipher. Recent advances in automatic search and deep learning-based methods have made this time-consuming task relatively easy, yet concerns persist over expertise requirements and potential oversights. To overcome these concerns, Kimura et al. proposed neural network-based output prediction (NN) attacks, offering simplicity, generality, and reduced coding mistakes. NN attacks could be helpful for designing secure symmetric-key ciphers, especially the S-box-based block ciphers. Inspired by their work, we first apply NN attacks to Simon , one of the AND-Rotation-XOR-based block ciphers, and identify structures susceptible to NN attacks and the vulnerabilities detected thereby. Next, we take a closer look at the vulnerable structures. The most vulnerable structure has the lowest diffusion property compared to others. This fact implies that NN attacks may detect such a property. We then focus on a biased event of the core function in vulnerable Simon -like ciphers and build effective linear approximations caused by such an event. Finally, we use these linear approximations to reveal that the vulnerable structures are more susceptible to a linear key recovery attack than the original one. We conclude that our analysis can be a solid step toward making NN attacks a helpful tool for designing a secure symmetric-key cipher.

Users are required to set passwords that are both complex and not included in leaked lists, with the recent increase in unauthorised access due to password list attacks. When password security is evaluated by the user, the burden on the user is large and impractical due to the increased size of the compromised password list. Therefore, there are methods to evaluate passwords on an external server, but there is a possibility that passwords may be compromised due to eavesdropping by the server administrator. In this paper, we propose, implement and evaluate an outsourcing password evaluation system using Intel SGX, which can prevent passwords from being eavesdropped.

There exist Adversarial Examples (AEs) that intentionally misclassified images of neural networks to a different class. We propose a method of attack by AEs that uses lasers to facilitate control of misclassification. A previous work has shown that road signs can be misclassified by irradiating them with lasers. We focus on the effective attack in the situations that the attacker can prepare the road signs. Namely, in our attack, the attacker prepares the road signs made by using AEs with missing perturbation, and makes misclassified by using laser irradiation in effective. In this paper, we present principle and method for misclassifying the road signs made by the AEs with missing perturbation by using laser irradiation. In addition, we demonstrate the effectiveness of the proposed method by experiments.

Aggregator oblivious encryption was proposed by Shi et al. (NDSS 2011). In this method, an aggregator can compute an aggregated sum of data and is unable to learn anything else (aggregator obliviousness). Since the aggregator does not learn individual data that may reveal users’ habits and behaviors, several applications including privacy-preserving smart metering have been considered. In this paper, we propose an aggregator oblivious encryption scheme with public verifiability where the aggregator is required to generate a proof of an aggregated sum, and anyone can verify whether the aggregated sum has been correctly computed by the aggregator. Although Leontiadis et al. (CANS 2015) considered verifiability, their scheme requires an interactive complexity assumption to provide the unforgeability of the proof. Our scheme is proven to be unforgeable under a static and simple assumption (a variant of the Computational Diffie–Hellman assumption). Moreover, our scheme inherits the tightness of the reduction of the Benhamouda et al. scheme (ACM TISSEC 2016) for proving aggregator obliviousness. This tight reduction allows us to employ elliptic curves of a smaller order and leads to efficient implementation. Specifically, for 112-bit security, we can employ Barreto–Naehrig (BN) curves with a 383-bit prime order, whereas we need to employ curves with a 1031-bit prime order to implement the Leontiadis et al. scheme. We give implementations of two schemes and evaluate their performances under those curves. We employ a Raspberry-Pi as a power-constrained device such as a smart meter. Consequently, we demonstrate that the running time of the data encryption, data aggregation and verification in our scheme are reduced by approximately 74%, 64% and 89%, respectively, compared to those of the Leontiadis et al. scheme.

In this paper, we propose an edge computing platform that can coordinate with each computing resources of device, edge node and cloud for diversifying data can be modularized and actively deployed in cooperation within the wide area network. This platform brings transparent processing environment for IoT devices according to a data characteristic or use purpose. We have been developing the platform based on open-source software for container and its automating deployment, scaling, and management tool.