Frequency-domain channel extrapolation is effective in reducing pilot overhead for massive multiple-input multiple-output (MIMO) systems. Recently, deep learning (DL) based channel extrapolators have become promising candidates for modeling complex frequency-domain dependency. Nevertheless, current DL extrapolators fail to operate in unseen environments under distribution shift, which poses challenges for large-scale deployment. In this paper, environment generalizable learning for channel extrapolation is achieved by realizing distribution alignment from a physics perspective. Firstly, the distribution shift of wireless channels is rigorously analyzed, which comprises the distribution shift of multipath structure and single-path response. Secondly, a physics-based progressive distribution alignment strategy is proposed to address the distribution shift, which includes successive path-oriented design and path alignment. Path-oriented DL extrapolator decomposes multipath channel extrapolation into parallel extrapolations of the extracted paths, which can mitigate the distribution shift of multipath structure. Path alignment is proposed to address the distribution shift of single-path response in path-oriented DL extrapolators, which eventually enables generalizable learning for channel extrapolation. In the simulation, distinct wireless environments are generated using the precise ray-tracing tool. Based on extensive evaluations, the proposed path-oriented DL extrapolator with path alignment can reduce extrapolation error by more than 6 dB in unseen environments compared to the state-of-the-arts.

Large Vision-Language Models (LVLMs) signify a groundbreaking paradigm shift within the Artificial Intelligence (AI) community, extending beyond the capabilities of Large Language Models (LLMs) by assimilating additional modalities (e.g., images). Despite this advancement, the safety of LVLMs remains adequately underexplored, with a potential overreliance on the safety assurances purported by their underlying LLMs. In this paper, we propose FigStep, a straightforward yet effective black-box jailbreak algorithm against LVLMs. Instead of feeding textual harmful instructions directly, FigStep converts the prohibited content into images through typography to bypass the safety alignment. The experimental results indicate that FigStep can achieve an average attack success rate of 82.50% on six promising open-source LVLMs. Not merely to demonstrate the efficacy of FigStep, we conduct comprehensive ablation studies and analyze the distribution of the semantic embeddings to uncover that the reason behind the success of FigStep is the deficiency of safety alignment for visual embeddings. Moreover, we compare FigStep with five text-only jailbreaks and four image-based jailbreaks to demonstrate the superiority of FigStep, i.e., negligible attack costs and better attack performance. Above all, our work reveals that current LVLMs are vulnerable to jailbreak attacks, which highlights the necessity of novel cross-modality safety alignment techniques.

Private set union (PSU) is a cryptographic protocol that allows two parties to compute the union of their sets without revealing anything else. Despite some efficient PSU protocols that have been proposed, they mainly focus on the balanced setting, where the sets held by the parties are of similar size. Recently, Tu et al. (CCS 2023) proposed the first unbalanced PSU protocol which achieves sublinear communication complexity in the size of the larger set.

State-of-the-art Byzantine fault-tolerant (BFT) protocols assuming partial synchrony such as SBFT and HotStuff use regular certificates obtained from 2f + 1 (partial) signatures. We show that one can use weak certificates obtained from only f + 1 signatures to assist in designing more robust and more efficient BFT protocols. We design and implement two BFT systems: Dashing (a family of two HotStuff-style BFT protocols) and Star (a parallel BFT framework).

Facebook public pages are a popular form of online social network (OSN) communities. The "like" connections between public pages create a graph of pages on Facebook. Geographic location is a crucial piece of metadata for pages, but it is often omitted by page managers. We propose a classification algorithm to restore the missing subdivision location of Facebook public pages. We propose neighborhood state distribution vectors as features for graph neural networks to classify the state of the pages. Then, we define intrastate and interstate Facebook public pages based on the high-probability state label outputted by the classification model. Finally, we profile states with different influences over the online communities by analyzing the classification confusion matrix, interstate page percentages, and interstate pages across state borders. Our method achieves better accuracy (87.52%) and F1 score (0.8756) than previous studies (66.2% and 73.08%).

Graph Neural Networks (GNNs) have gained the interest of industry with Relational Graph Convolutional Networks (R-GCNs) showing promise for fraud detection. Taking existing workflows that leverage graph features to train a gradient boosted decision tree (GBDT) and replacing the graph features with GNN produced embedding achieves an increase in accuracy. However, recent work has shown that the combination of graph attributes with GNN embeddings provides the biggest lift in accuracy.

This study revisits the side‐channel security of the elliptic curve cryptography (ECC) scalar multiplication implemented with Montgomery ladder. Focusing on a specific implementation that does not use the y ‐coordinate for point addition (ECADD) and point doubling (ECDBL), the authors show that Montgomery ladder on Weierstrass curves is vulnerable to a chosen base‐point attack. Unlike the normal implementation with y ‐coordinate, in the scenario of this study, the chosen base‐point strategy will not lead to operations with two same inputs during the ECADD and/or ECDBL. Instead, by choosing a suitable base‐point, one will find that there are operations that share a common operand; while it is not the case if the base‐point is not chosen correctly. This results in the recovery of the secret (fixed) scalar. They also experiment the methods of shared operand detection on a real‐world SoC, where a secp256k1 dedicated Montgomery ladder scalar multiplication with x ‐only coordinate is implemented, to show the efficiency of the scalar recovery attack. Naturally, the attack can be generalised to other Weierstrass curves when they contain special points.

This paper implements an OpenCL-based FPGA accelerator for YOLOv2 on Arria-10 GX1150 FPGA board. The hardware architecture adopts a scalable pipeline design to support multi-resolution input image, and improves resource utilization by full 8-bit fixed-point computation and CONV+BN+Leaky-ReLU layer fusion technology. The proposed design achieves a peak throughput of 566 GOPs under 190 MHz working frequency. The accelerator could run YOLOv2 inference with 288×288 input resolution and tiny YOLOv2 with 416×416 input resolution at the speed of 35 and 71 FPS, respectively.

The integral attack, exploits the balanced property of the output in the distinguisher. Usually, adversaries append some rounds after the distinguisher, guess the corresponding key bits and check whether the target bits are balanced. Few works add rounds before the distinguisher to make the key recovery attack. In the first full‐round attack on MISTY1, Todo adds one FL layer (key‐dependent linear function) before the distinguisher. In this study, the authors extend his method and give a general method, which they can use to extend some rounds (non‐linear) before the distinguisher to attack more rounds with data complexity smaller than the whole space and little extra time consumption. The basic idea is that for different subkeys guessed in the forward rounds, they set different constant values for the input of the distinguisher. Finally, the selected data space is not full. For substitution permutation network (SPN) (Feistel with SPN round function) structures with 4 bit S‐box and bit permutation, they estimate the data complexity when adding one round before the distinguishers for all 4 bit S‐boxes. Using the method, they improve the integral attacks on PRESENT, RECTANGLE, TWINE and LBlock, and their results could cover one more round.

Dynamic key-guessing techniques, which exploit the property of AND operation, could improve the differential and linear cryptanalytic results by reducing the number of guessed subkey bits and lead to good cryptanalytic results for SIMON. They have only been applied in differential and linear attacks as far as we know. In this paper, dynamic key-guessing techniques are first introduced in integral cryptanalysis. According to the features of integral cryptanalysis, we extend dynamic key-guessing techniques and get better integral cryptanalysis results than before. As a result, we present integral attacks on 24-round SIMON32, 24-round SIMON48/72, and 25-round SIMON48/96. In terms of the number of attacked rounds, our attack on SIMON32 is better than any previously known attacks, and our attacks on SIMON48 are the same as the best attacks.

Low Entropy Masking Schemes (LEMS) are countermeasure techniques to mitigate the high performance overhead of masked hardware and software implementations of symmetric block ciphers by reducing the entropy of the mask sets. The security of LEMS depends on the choice of the mask sets. Previous research mainly focused on searching balanced mask sets for hardware implementations. In this paper, we find that those balanced mask sets may have vulnerabilities in terms of absolute difference when applied in software implemented LEMS. The experiments verify that such vulnerabilities certainly make the software LEMS implementations insecure. To fix the vulnerabilities, we present a selection criterion to choose the mask sets. When some feasible mask sets are already picked out by certain searching algorithms, our selection criterion could be a reference factor to help decide on a more secure one for software LEMS.

Online social network (OSN) discussion groups are exerting significant effects on political dialogue. In the absence of access control mechanisms, any user can contribute to any OSN thread. Individuals can exploit this characteristic to execute targeted attacks, which increases the potential for subsequent malicious behaviors such as phishing and malware distribution. These kinds of actions will also disrupt bridges among the media, politicians, and their constituencies. For the concern of Security Management, blending malicious cyberattacks with online social interactions has introduced a brand new challenge. In this paper we describe our proposal for a novel approach to studying and understanding the strategies that attackers use to spread malicious URLs across Facebook discussion groups. We define and analyze problems tied to predicting the potential for attacks focused on threads created by news media organizations. We use a mix of macro static features and the micro dynamic evolution of posts and threads to identify likely targets with greater than 90% accuracy. One of our secondary goals is to make such predictions within a short (10 minute) time frame. It is our hope that the data and analyses presented in this paper will support a better understanding of attacker strategies and footprints, thereby developing new system management methodologies in handing cyber attacks on social networks.

In this paper, we propose a genetic algorithm for solving the shortest vector problem (SVP) based on sparse representation of short lattice vectors, which, we prove, can guarantee finding the shortest lattice vector under a Markov analysis. With some heuristic improvements (local search and heuristic pruning), the SVP genetic algorithm, by experimental results, outperforms other SVP algorithms, like the famous Kannan-Helfrich algorithm under SVP challenge benchmarks. In summary, we, for the first time, adopt the genetic algorithm in solving the shortest vector problem, based on which lattice-based cryptosystem is as a promising candidate for post-quantum cryptography.

In this study, the authors study the security of hash functions SM3 and BLAKE‐256 against boomerang attack. SM3 is designed by Wang et al . and published by Chinese Commercial Cryptography Administration Office for the use of electronic certification service system in China. BLAKE is one of the five finalists of the NIST SHA‐3 competition submitted by Aumasson et al . For SM3, they present boomerang distinguishers for the compression function reduced to 34/35/36/37 steps out of 64 steps, with time complexities 2 31.4 , 2 33.6 , 2 73.4 and 2 192 , respectively. Then, they show some incompatible problems existed in the previous boomerang attacks on SM3. Meanwhile, they launch boomerang attacks on up to 7‐ and 8‐round keyed permutation of BLAKE‐256, which are the first valid 7‐round and 8‐round boomerangs for BLAKE‐256. Especially, since the author's distinguishers on 34/35‐steps compression function of SM3 and 7‐round keyed permutation of BLAKE‐256 are practical, they are able to obtain boomerang quartets of these attacks. As far as they know, these are the best results against round‐reduced SM3 and BLAKE‐256.

Background: Clinical studies showed that abdominal acupuncture was an effective and safe treatment for depression. However, the underlying neural mechanism was still largely unknown. The present study investigated the brain spontaneous activity changes following with the abdominal acupuncture treatment in female depressive patients with different levels, to explore the neural basis underpinning abdominal acupuncture treatment efficacy. Methods: Fifteen female depression patients (nine mild and six moderate-serious participants) were recruited for the present study, these patients treated by abdominal acupuncture for 4 weeks. The resting-state functional magnetic resonance imaging (fMRI) data was collected at before and after treatment stages. The changes of resting state regional homogeneity (ReHo) were examined between these two stages, to suggest the acupuncture treatment effect on spontaneous neural activity. Further analysis was performed to explore the correlation between ReHo changes and clinical measurements (Self-rating depression scale SDS and Montgomery-Asberg Depression Rating Scale MADRS scores). Results: The scores of SDS and MADRS had gradually declined after-treatment stage in mild and moderate-serious depression, compared with those of before-treatment stage. There were statistically differences of SDS and MADRS scores after treatment at 1st, 2nd, 3rd and 4th week's (P<;0.05). Compared with the before-treatment state in mild depression, brain regions with increased ReHo were found in Frontal_Sup_L, Frontal_Inf_Tri_L, Frontal_Inf_Tri_R, and Parietal_Inf_R. Meanwhile, the brain regions including Temporal_Pole_Sup_R, Occipital_Inf_R, Occipital_Inf_R, Temporal_Sup_R, Cuneus_R and Temporal_Mid_L were observed with decreased ReHo. In moderate-serious depression, brain regions with increased ReHo were found in Frontal_Med_Orb_Rand, and Cingulum_Mid_R; meanwhile, the brain regions including Lingual_L and Postcentral_R were observed with decreased ReHo. There were a large of quantity regional homogeneity changes which correlated with scores of SDS and MADRS. Conclusion: Abdominal acupuncture treatment could induce spontaneous brain activity change in depression patients. However, there were different regional homogeneity changes in different depression levels. ReHo changes in mild depression were mainly distributed in prefrontal and temporal lobe, but the ReHo change in prefrontal, Parietal, Occipital lobe and limbic system in moderate-serious depression. Prefrontal region took important roles in the acupuncture treatment effect on depression patients. Together, the findings of the present study provided new evidence for that abdominal acupuncture was an effective treatment for depression, and the ReHo was efficacy evaluation objective indicators of acupuncture treatment for depression.

Objective: This study aims to investigate the effect of Chinese herbs treatment of Kidney-strengthening therapy, Spleen-invigorating therapy and Liver-regulating therapy on monoamine neurotransmitters in ovariectomized female rats. Methods: 54 female rats were divided into six groups by random: control group, sham operation group, model group, Spleen-invigorating group, Kidney-strengthening group and Liver-regulating group. Each group contained 9 rats. The organ index of hypothalamic pituitary and uterus and serum serotonin (5-HT) and dopamine (DA) test was compared after treating with 8 weeks. Results: The organ index of hypothalamic pituitary and uterus in model group dcreased with significant difference (p0.05) comparing with model group. On the effect of serum serotonin and dopamine, model group showed significantly decline in the content of serum serotonin comparing to control group (P0.05) in all groups treated with Chinese herb formulae. Conclusion: Different Chinese herbs treatment of Spleen-invigorating therapy, Kidney-strengthening therapy and Liver-regulating therapy may improve monoamine neurotransmitters in ovariectomized female rats by different targets. Spleen-invigorating group was more effected in improving monoamine neurotransmitters than other two groups.

At ISIT 2008, Aguilar Melchor, Castagnos and Gaborit presented a lattice-based homomorphic encryption scheme (abbreviated as MCG). Its security is based on the Computational Knapsack Vector Problem. In this paper, we explore a secret linear relationship between the public keys and the secret keys, which can be used to construct a reduced-dimension lattice, and then we obtain a group of equivalent private keys by solving the Closest Vector Problem of the lattice. Moreover, our attack is practical on all the three settings of recommended parameters, and the running time to recover the equivalent private keys is only several hours on a single PC.

In this paper, we present an improvement of the Nguyen-Vidick heuristic sieve algorithm for shortest vector problem in general lattices, which time complexity is 20.3836n polynomial computations, and space complexity is 20.2557n. In the new algorithm, we introduce a new sieve technique with two-level instead of the previous one-level sieve, and complete the complexity estimation by calculating the irregular spherical cap covering.