Leixiao Li

At EUROCRYPT 2024, recovering key bits via superpoly (over 20 trillion terms for 851‐round Trivium) hit a bottleneck due to its complexity. While existing machine learning (ML)‐based cryptanalysis has shown promise on block ciphers, applying it to stream ciphers is difficult due to rapid diffusion. In this work, we propose a novel ML‐based analytic framework: weighted ensemble neural network for key prediction (WENNKP), aimed at investigating potential weak initialization vectors (IVs) and their induced statistical biases. Unlike standard ciphertext‐only attacks, we model the problem as a chosen‐IV analysis. Our preliminary observations suggest that for a specific, empirically selected IV candidate, the keystream of 1152‐round Trivium and Kreyvium appears to exhibit a distinguishable statistical dependence on the secret key. WENNKP integrates multiple neural networks to act as a distinguisher, indicating this potential nonrandomness with a prediction accuracy exceeding 50% (random guess). Experimental results show that WENNKP detects signs of these biases in over 70 out of 100 validation datasets. This work provides empirical evidence supporting the hypothesis of weak IVs in modern stream ciphers and explores the capability of ML as a sensitive tool for detecting deep statistical leaks.

The rapid increase in carbon emissions from the logistics transportation industry has underscored the urgent need for low-carbon logistics solutions. Electric logistics vehicles (ELVs) are increasingly being considered as replacements for traditional fuel-powered vehicles to reduce emissions in urban logistics. However, ELVs are typically limited by their battery capacity and load constraints. Additionally, effective scheduling of charging and the management of transportation duration are critical factors that must be addressed. This paper addresses low energy consumption scheduling (LECS) problem, which aims to minimize the total energy consumption of heterogeneous ELVs with varying load and battery capacities, considering the availability of multiple charging stations (CSs). Given that the complexity of LECS problem, this study proposes a heterogeneous attention model based on encoder-decoder architecture (HAMEDA) approach, which employs a heterogeneous graph attention network and introduces a novel decoding procedure to enhance solution quality and learning efficiency during the encoding and decoding phases. Trained via deep reinforcement learning (DRL) in an unsupervised manner, HAMEDA is adept at autonomously deriving optimal transportation routes for each ELV from specific cases presented. Comprehensive simulations have verified that HAMEDA can diminish overall energy utilization by no less than 1.64% compared with other traditional heuristic or learning-based algorithms. Additionally, HAMEDA excels in maintaining an advantageous equilibrium between execution speed and the quality of solutions, rendering it exceptionally apt for expansive tasks that necessitate swift decision-making processes.