Secret Sharing Research Articles

Influence-Based Fair Selection for Sample-Discriminative Backdoor Attack

Backdoor attacks have posed a serious threat in machine learning models, wherein adversaries can poison training samples with maliciously crafted triggers to compromise the victim model. Advanced backdoor attack methods have focused on selectively poisoning more vulnerable training samples, achieving a higher attack success rate (ASR). However, we found that when the manipulation strength of the trigger is constrained to a very small value for imperceptible attacks, they suffer from extremely uneven class-wise ASR due to the unequal selection of instances per class. To solve this issue, we propose a novel backdoor attack method based on Influence-based Fair Selection (IFS), including two objectives: 1) selecting samples that significantly contribute to ASR and 2) ensuring class balance during the selection process. Specifically, we adapt Influence Functions, a classic technique in robust statistics, to evaluate the influence of trigger-embedded training samples on ASR. In this case, training samples contributing to reducing the backdoored test risk could possess higher influence scores. Further, a group-based pruning strategy is designed to avoid calculating the influence on ASR for all training samples, thereby significantly reducing the computational cost. Then, based on the influence score, we design an adaptive thresholding scheme to dynamically select samples with higher influence while maintaining class balance. Extensive experiments on four datasets verify the effectiveness of IFS compared with advanced methods.

Smart Cities Cybersecurity: The Synergy of Ai, E-Governance, and Stakeholder Involvement

This project presents a secure and intelligent web portal for smart city management, integrating AI, e-Governance, and stakeholder participation to enhance governance, security, and efficiency. It addresses cybersecurity risks using modern cryptographic techniques such as Shamir's Secret Sharing, OTP email validation, and AWS S3 encryption. Developed with Python and Flask, the system employs Role-Based Access Control (RBAC) to regulate access for Admin, Manager, and Citizen roles. By combining AI-driven functionalities with robust security measures, this solution ensures data privacy, public trust, and efficient urban resource management.

Tactical decomposable regular group divisible designs and their threshold schemes

Tactical decomposable regular group divisible designs and their threshold schemes

EBTM: An Enumeration-Based Thresholding Method for Degraded Circular Dot Segmentation.

Circular dots are widely used in various measurement applications due to their inherent symmetry, ease of detection, and scalability. However, when degraded by factors such as specular highlights, low contrast, strong noise, or friction damage, accurately extracting them from the background becomes a significant challenge. To address this issue, an enumeration-based thresholding method (EBTM) is proposed for degraded circular dot segmentation. Firstly, a series of candidate outputs are generated using an enumeration-based thresholding scheme. Next, an assessment criterion is developed to evaluate these candidate outputs. Finally, the optimal segments are selected from each candidate output and combined to produce a reasonable thresholding result. Unlike traditional methods, the novel approach does not focus on selecting the optimal threshold values, but instead aims to choose the best segments to produce the desired output. Owing to the enumeration-based thresholding mechanism, the novel approach demonstrates greater robustness in handling the challenges in degraded circular dot images. Extensive comparative studies demonstrate the superiority of the novel approach.

Secure and intelligent 5G-enabled remote patient monitoring using ANN and Choquet integral fuzzy VIKOR

Rapid advancements in healthcare technologies necessitate efficient and secure remote patient monitoring systems. This research develops an intelligent system that combines ANN technology and 5G infrastructure with MCDM methods based on Choquet Integral Fuzzy VIKOR to improve medical data acquisition processes. Physical Layer Security (PLS) is a main emphasis point since it protects transmitted healthcare data from eavesdroppers and cyber intruders. The proposed model implements Reinforcement Learning with Hyper-parameter tuning and Lasso regression to obtain a 97.25% accuracy level, which exceeds Physical-Layer Authentication with Superimposed Independent authentication Tags PLA-SIT (97%), Flexible Physical Layer Authentication FPLA (96.8%) and Privacy-Embedded Lightweight and Efficient Automated PLA (95.3%). The proposed model surpasses both CNN-based mechanisms by 94.7%, Shamir’s Secret Sharing Algorithm by 90.7%, and the Blowfish Algorithm by 82.3%. The enhanced quality of service alongside reliability produces the model as a dependable solution for MIoT applications that will exist in the next generation.

Enhanced brain image security using a hybrid of lifting wavelet transform and support vector machine

Thanks to technological improvements, digital picture watermarking has emerged as a useful method for preventing unlawful use and manipulation of digital photographs. Providing robustness against geometrical assault while maintaining an adequate level of security and imperceptibility is a basic challenge in digital picture watermarking. With the use of support vector machine (SVM) and lifting wavelet transform (LWT), this study offers an effective authentication approach for digital image watermarking on medical images. To distinguish between the region of interest (ROI) and the non-region of interest (NROI) in the medical image, SVM is first employed in this article. After that, LWT is used to incorporate watermark data into the medical image’s NROI section (cover image). Additionally, a shared secret key has been used to increase the suggested scheme’s resilience. A vast image database is used to test the method’s performance in various scenarios. To determine whether the current plan was acceptable, the study examined several experimental investigations. The experimental results give a PSNR value of 67.81 dB and a structural similarity index measure value of 0.9999, Where the PSNR improvement percentage is 13.9462 dB, showing durability and imperceptibility for the proposed watermarking model.

Lossless Recovery Scheme for Grayscale Visual Cryptography Based on Random Grids

With the advancement of digitalization, the demand for grayscale image encryption has been increasing. Previous visual cryptography schemes for grayscale images have encountered issues such as information loss during recovery, poor recovery quality, and significant pixel expansion. This study introduces novel grayscale visual cryptography schemes based on grayscale threshold segmentation, including the (n,n) and (k,n) schemes, utilizing random grids. These schemes enable lossless recovery of the secret image when all shares are involved in the decryption process. Furthermore, by incorporating the particle swarm optimization algorithm, a scheme for generating share images containing grayscale information was proposed. This method relies on the grayscale-overlay principle, allowing secret image recovery through the overlay of shares without requiring computational devices, demonstrating the applicability of the extended scheme. Theoretical analysis demonstrates the security of the proposed schemes, and experimental results further verify their effectiveness.

Securing Confidentiality in Distributed Ledger Systems with Secure Multi-party Computation for Financial Data Protection

This study addresses confidentiality challenges in financial Distributed Ledger Systems (DLS) using Secure Multi-Party Computation (SMPC). By analyzing real-world datasets, it evaluates privacy risks, protocol efficiency, and system resilience. Findings highlight SMPC’s role in enhancing security while balancing computational efficiency. Using the Elliptic AML Bitcoin Transactions dataset, anomaly detection (Isolation Forest) identifies financial confidentiality vulnerabilities, revealing that anomalous transactions exhibit a 336.1% increase in volume and a 15.5% rise in frequency, suggesting heightened risks. A comparative analysis of SMPC protocols utilizing the MP-SPDZ benchmark dataset and one-way ANOVA confirms that Yao’s Garbled Circuits is the most computationally efficient (180.50 ms execution time), whereas Shamir’s Secret Sharing offers superior security (0.73 high-probability security). Kaplan-Meier survival analysis of Verizon DBIR 2024 establishes that SMPC extends financial system longevity (36.11 months vs. 21.91 months for traditional encryption). Recommendations include integrating scalable SMPC models, standardizing regulatory frameworks, optimizing algorithmic efficiency, and enhancing anomaly detection in financial DLS.

All-You-Can-Compute: Packed Secret Sharing for Combined Resilience

Unprotected cryptographic implementations are vulnerable to implementation attacks, such as passive side-channel attacks and active fault injection attacks. Recently, countermeasures like polynomial masking and duplicated masking have been introduced to protect implementations against combined attacks that exploit leakage and faults simultaneously. While duplicated masking requires O (t · e) shares to resist an adversary capable of probing t values and faulting e values, polynomial masking requires only O (t · e) shares, which is particularly beneficial for affine computation. At CHES’24, Arnold et al. showed how to further improve the efficiency of polynomial masking in the presence of combined attacks by embedding two secrets into one polynomial sharing. This essentially reduces the complexity of previous constructions by half. The authors also observed that using techniques from packed secret sharing (Grosso et al., CHES’13) cannot easily achieve combined resilience to encode an arbitrary number of secrets in one polynomial encoding. In this work, we resolve these challenges and show that it is possible to embed an arbitrary number of secrets in one encoding and propose gadgets that are secure against combined attacks. We present two constructions that are generic and significantly improve the computational and randomness complexity of existing compilers, such as the laOla compiler presented by Berndt et al. at CRYPTO’23 and its improvement by Arnold et al. For example, for an AES evaluation that protects against t probes and e faults, we improve the randomness complexity of the state-of-the-art construction when t + e > 3, leading to an improvement of up to a factor of 2.41.

Communication Efficient Secure Three-Party Computation Using Lookup Tables for RNN Inference

Many leading technology companies currently offer Machine Learning as a Service Platform, enabling developers and organizations to access the inference capabilities of pre-trained models via API calls. However, due to concerns over user data privacy, inter-enterprise competition, and legal and regulatory constraints, directly utilizing pre-trained models in the cloud for inference faces security challenges. In this paper, we propose communication-efficient secure three-party protocols for recurrent neural network (RNN) inference. First, we design novel three-party secret-sharing protocols for digit decomposition, B2A conversion, enabling efficient transformation of secret shares between Boolean and arithmetic rings. Then, we propose the lookup table-based secure three-party protocol. Unlike the intuitive way of directly looking up tables to obtain results, we compute the results by utilizing the inherent mathematical properties of binary lookup tables, and the communication complexity of the lookup table protocol is only related to the output bit width. We also design secure three-party protocols for key functions in the RNN, including matrix multiplication, sigmoid function, and Tanh function. Our protocol divides the computation into online and offline phase, and places most of the computations locally. The theoretical analysis shows that the communication round of our work was reduced from four rounds to one round. The experiment results show that compared with the current SOTA-SIRNN, the online communication overhead of sigmoid and tanh functions decreased by 80.39% and 79.94%, respectively.

A Two-Phase Embedding Approach for Secure Distributed Steganography.

Steganography serves a crucial role in secure communications by concealing information within non-suspicious media, yet traditional methods often lack resilience and efficiency. Distributed steganography, which involves fragmenting messages across multiple containers using secret sharing schemes, offers improved security but increases complexity. This paper introduces a novel two-phase embedding algorithm that mitigates these issues, enhancing both security and practicality. Initially, the secret message is divided into shares using Shamir's Secret Sharing and embedded into distinct media containers via pseudo-random LSB paths determined by a unique internal stego key. Subsequently, this internal key is further divided and embedded using a shared stego key known only to the sender and receiver, adding an additional security layer. The algorithm effectively reduces key management complexity while enhancing resilience against sophisticated steganalytic attacks. Evaluation metrics, including Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM), demonstrate that stego images maintain high quality even when embedding up to 0.95 bits per pixel (bpp). Additionally, robustness tests with StegoExpose and Aletheia confirm the algorithm's stealthiness, as no detections are made by these advanced steganalysis tools. This research offers a secure and efficient advancement in distributed steganography, facilitating resilient information concealment in sophisticated communication environments.

Enhancing Crime Detection in Video Surveillance via a Lightweight Blockchain and Homomorphic Encryption-based Computer Vision System

Blockchain technology consists of distributed ledgers or database systems, regarded as immutable, secure, and innovative, characterized by unsupervised internal maintenance with a special security protocol used to prevent inference from malicious or third parties. The widespread use of this technology has led to deep research into the problems posed by this technology, which can be summarized in terms of computational cost and latency time. The crime detection process in video surveillance has made great progress with the use of technologies such as the Internet of Things and blockchain technologies. However, to reach high levels of security in the physical crime detection process in which data are sent to servers via a computer network, there must be a high degree of security for Internet of Things systems related to the crime detection process. There has been a significant increase in the number of problems associated with crime detection in video surveillance systems, including the modification of surveillance data during transfer to and from servers. For this reason, establishing a reliable and secure system for transferring video surveillance data to servers has become a high priority. This paper presents a lightweight security system to protect data generated in the crime detection process, both from video surveillance cameras and the servers that store these data. The challenges related to IoT-based video surveillance cameras and monitoring and control centers have been considered, turning the system primarily into a decentralized system. In this paper, a lightweight blockchain system based on a proof of secret share consensus algorithm technology is proposed, along with the encryption of surveillance data via modified Okamoto–Uchiyama homomorphic encryption technology. The proposed system is evaluated via standard blockchain and security evaluation metrics, demonstrating efficient utilization of computational costs and realization of security, with a high scalability rate. The VGG16 deep learning model is employed in the proposed system to detect and classify criminal activities in surveillance videos. Owing to its ability to identify patterns and anomalies, the model achieved an accuracy of 94%, demonstrating a high level of performance in crime detection and prevention. Overall, the use of VGG16 provides an efficient and reliable approach for improving the security of public spaces and reducing the risk of criminal activity.

SecureXGB: A Secure and Efficient Multi-party Protocol for Vertical Federated XGBoost

Extreme Gradient Boosting (XGBoost) demonstrates excellent performance in practice and is widely used in both industry and academic research. This extensive application has led to a growing interest in employing multi-party data to develop more robust XGBoost models. In response to increasing concerns about privacy leakage, secure vertical federated XGBoost is proposed. It employs secure multi-party computation techniques, such as secret sharing (SS), to allow multiple parties holding vertically partitioned data, i.e., disjoint features on the same samples, to collaborate in constructing an XGBoost model. However, the running efficiency is the primary obstacle to the practical application of existing protocols, especially in multi-party settings. The reason is that these protocols not only require the execution of data-oblivious computations to protect intermediate results, leading to high computational complexity, but also involve a large number of SS-based non-linear operations with high overheads, e.g., division operations in gain score calculation and comparison operations in best split selection. To this end, we present a secure and efficient multi-party protocol for vertical federated XGBoost, called SecureXGB, which can perform the collaborative training of an XGBoost model in an SS-friendly manner. In SecureXGB, we first propose a parallelizable multi-party permutation method, which can secretly and efficiently permute all samples before model training to reduce the reliance on data-oblivious computations. Then, we design a linear gain score that can be evaluated without involving division operations and has equivalent utility to the original gain score. Finally, we develop a synchronous best split selection method to secretly identify the best split with the maximum gain score using a minimal number of comparison operations. Experimental results demonstrate that SecureXGB can achieve better training efficiency than state-of-the-art protocols without the loss of model accuracy.

Measurement‐Device‐Independent Continuous‐Variable Quantum Secret Sharing

AbstractAlthough continuous‐variable quantum secret sharing (CVQSS) is theoretically proven to be secure, it may still be subject to various practical attacks due to the imperfections of devices. In this study, measurement‐device‐independent CVQSS (MDI‐CVQSS) is proposed in which measurement is no longer the duty for the dealer but is performed by an untrusted party Charlie, so that all detector side‐channel attacks can be eliminated, greatly enhancing the practical security of CVQSS system. The security bound is derived for the proposed MDI‐CVQSS, and its numerical simulation shows that MDI‐CVQSS outperforms conventional CVQSS in terms of both secret key rate and maximal number of users. It is found that MDI‐CVQSS possesses several unique properties, such as local stability of the secret key rate, low total excess noise and high utilization rate of raw data, which indicate that MDI‐CVQSS has the potential for efficiently building a large‐scale quantum communication network. Moreover, it is also suggested three variations of MDI‐CVQSS to satisfy the requirements of different multi‐point quantum communication scenarios.

Low-complexity User Scheduling Framework for mmWave Hybrid Beam-forming OFDM System under Non-ideal CSI

Introduction: User scheduling in millimeter-wave (mmWave) multi-user hybrid beam-forming Orthogonal Frequency Division Multiplexing (OFDM) systems involves the joint optimization of resource block group (RBG) allocation, beam pairing, and user selection. However, choosing the optimal scheduled User Equipment (UE), allocated RBGs, and communicating beams in practical mmWave hybrid beam-forming systems with non-ideal Channel State Information (CSI) remains challenging. Methods: In this paper, we propose a low-complexity user scheduling framework. On one hand, two user classification methods are proposed under non-ideal CSI, assisted by beam search and RBG allocation respectively. On the other hand, in order to ensure both the throughput and fairness performance, a novel user selection scheme, called weighted user selection based on feedback threshold (WUSFT) scheme is proposed, and approximate closed-form expression of fairness index for both full feedback and feedback based on threshold are derived. Our proposed methods consist of three steps. First, RBGs are allocated based on the maximum received signal power (MRSP) of each user on each RBG, utilizing quasi-omnidirectional beams. In the second step, the communicating beams are further searched to achieve the MRSP with the allocated RBG. Results: Finally, the allocated RBG, or the determined beam information obtained through the beam search, is used to represent the correlation of user channels and classify the UEs into groups. Only UEs whose reported factor is not less than the feedback threshold will report received information. This simplifies the operation of user scheduling. Moreover, simulation results demonstrate that our proposed schemes can achieve more than 92.3% of the sum rate performance of exhaustive search methods while maintaining relatively low complexity. Conclusion: In addition, the user feedback overhead (FO) reduces obviously, especially with large UE number.

Tennis Legend Venus Williams to Share Secrets of Success at Annual Meeting

Tennis Legend Venus Williams to Share Secrets of Success at Annual Meeting

Coupling Secret Sharing with Decentralized Server-Aided Encryption in Encrypted Deduplication

Outsourcing storage to the cloud can save storage costs and is commonly used in businesses. It should fulfill two major goals: storage efficiency and data confidentiality. Encrypted deduplication can achieve both goals via performing deduplication to eliminate the duplicate data within encrypted data. Traditional encrypted deduplication generates the encryption key on the client side, which poses a risk of offline brute-force cracking of the outsourced data. Server-aided encryption schemes have been proposed to strengthen the confidentiality of encrypted deduplication by distributing the encryption process to dedicated servers. Existing schemes rely on expensive cryptographic primitives to provide a decentralized setting on the dedicated servers for scalability. However, this incurs substantial performance slowdown and can not be applied in practical encrypted deduplication storage systems. In this paper, we propose a new decentralized server-aided encrypted deduplication approach for outsourced storage, called ECDedup, which leverages secret sharing to achieve secure and efficient key management. We are the first to use the coding matrix as the encryption key to couple the encryption and encoding processes in encrypted deduplication. We also propose a acceleration scheme to speed up the encryption process of our ECDedup. We prototype ECDedup in cloud environments, and our experimental results based on the real-world backup datasets show that ECDedup can improve the client throughput by up to 51.9% compared to the state-of-the-art encrypted deduplication schemes.

Materials Secure Computation with Secret Sharing: A Bayesian Optimization Scheme and Its Performance

Materials Secure Computation with Secret Sharing: A Bayesian Optimization Scheme and Its Performance

Advanced Physical Layer Metasurface‐Empowered Cryptography with Robustness, High Capacity, and Enhanced Security

AbstractPhysical layer metasurface‐empowered cryptography offers a novel approach to secure encryption. This study proposes an encryption method with robustness, high capacity, and enhanced security based on the dual‐band complex‐amplitude metasurface. By leveraging the American Standard Code for Information Interchange (ASCII), both the secret message and the compressed digital signature are encrypted and transmitted together, eliminating codebook transmission and reducing security risks. The modified Visual Secret Sharing (VSS) scheme is employed to strengthen the security, fidelity, and robustness. Additionally, a three‐dimensional wavefront reconstruction algorithm is used to increase the encryption capacity, and a multivariate encryption key provides an extra layer of protection. A prototype metasurface sample is designed, fabricated, and characterized as a proof‐of‐concept demonstration. The measured results closely match the numerical ones and the design objective, confirming the viability and resilience of this high‐capacity, enhanced security encryption method for applications in anti‐counterfeiting and secure data transmission.

Haven++: Batched and Packed Dual-Threshold Asynchronous Complete Secret Sharing with Applications

Asynchronous complete secret sharing (ACSS) is a foundational primitive in the design of distributed algorithms and cryptosystems that require confidentiality. ACSS permits a dealer to distribute a secret to a collection of N servers so that everyone holds shares of a polynomial containing the dealer's secret. This work contributes a new ACSS protocol, called Haven++, that uses packing and batching to make asymptotic and concrete advances in the design and application of ACSS for large secrets. Haven++ allows the dealer to pack multiple secrets in a single sharing phase, and to reconstruct either one or all of them later. For even larger secrets, we contribute a batching technique to amortize the cost of proof generation and verification across multiple invocations of our protocol. The result is an asymptotic improvement in the worst-case amortized communication and computation complexity, both for ACSS itself and for its application to asynchronous distributed key generation. Our ADKG based on Haven++ achieves, for the first time, an optimal worst case amortized communication complexity of κN without a trusted setup. To show the practicality of Haven++, we implement it and find that it outperforms the work of Yurek et al. (NDSS 2022) by more than an order of magnitude when there are malicious, faulty parties.