As large language models (LLMs) become increasingly capable, concerns over the unauthorized use of copyrighted and licensed content in their training data have grown, especially in the context of code. Open-source code, often protected by open source licenses (e.g, GPL), poses legal and ethical challenges when used in pretraining. Detecting whether specific code samples were included in LLM training data is thus critical for transparency, accountability, and copyright compliance. We propose SynPrune, a syntax-pruned membership inference attack method tailored for code. Unlike prior MIA approaches that treat code as plain text, SynPrune leverages the structured and rule-governed nature of programming languages. Specifically, it identifies and excludes consequent tokens that are syntactically required and not reflective of authorship, from attribution when computing membership scores. Experimental results show that SynPrune consistently outperforms the state-of-the-arts. Our method is also robust across varying function lengths and syntax categories.

Extensive research has been done on the security of symmetric encryption algorithms in the black-box attack contexts, where the execution platforms are supposed to be secure. Recent studies intend to secure encryption algorithms in a more challenging but widely adopted scenario, the white-attack context (WBAC), where the adversaries have full visibility of the implementations of cryptosystems and full control over execution platforms. Various of approaches for protecting symmetric encryption algorithms in the WBAC have been proposed. Unfortunately, most existing approaches have been broken. This paper proposes a novel approach for securing symmetric encryption algorithms based on substitution-permutation network (SPN). Our key idea is to incorporate additional secret components into lookup tables to expand and dramatically change the inner states of encryption. Unlike existing approaches, these components do not need to be annihilated in adjacent rounds, and the ciphertext remains essentially unchanged. To recover the plaintext, only simple operations and the standard decryption algorithm are required. Our approach can be applied to protect SPN-based symmetric encryption algorithms such as AES. Security analysis indicates that the approach is expected to be resistant to both existing and unknown attacks. Furthermore, experimental evaluation shows that our approach performs well on various platforms.

With the rapid advancement of cloud computing and the exponential growth of data, the demand for secure data querying and sharing has become increasingly prominent. Identity-Based Encryption with Keyword Search (IBEKS) and Hierarchical IBEKS (HIBEKS) address the issue of secure data querying, as it enables resource-constrained clients to effectively search for encrypted data stored in the cloud. However, existing HIBEKS schemes lack flexible data-query mechanisms among users in the same level, are highly vulnerable to attacks launched by quantum computers and Keyword Guessing Attacks (KGA), and lead to relatively high end-to-end latency. To meet more complex and diverse application requirements and address these vulnerabilities, we introduce a novel primitive called Hierarchical Identity-Based Puncturable Encryption with Keyword Search (HIBPEKS). The decryption keys held by higher-level users are capable of generating decryption keys for lower-level users, thereby enhancing the ability to perform multi-level encrypted data queries within the user group. In addition, to control the query of encrypted data, higher-level users can use specific tags to puncture decryption keys (for lower-level users), so that lower-level users will no longer be able to query the parts of the data associated with the punctured tags. Technically, we have improved the previous lattice-based IBEKS schemes and implemented an efficient and flexible data query mechanism in a hierarchical setting by exploiting Puncturable Encryption (PE) techniques. Moreover, we formalize the security model of HIBPEKS and prove its security within the framework of the random oracle model. Finally, we experimentally evaluate HIBPEKS and show that HIBPEKS is computationally efficient and practical.

Graph Neural Networks (GNNs) have marked significant impact in traffic state prediction, social recommendation, knowledge-aware question answering and so on. As more and more users move towards cloud computing, it has become a critical issue to unleash the power of GNNs while protecting the privacy in cloud environments. Specifically, the training data and inference data for GNNs need to be protected from being stolen by external adversaries. Meanwhile, the financial cost of cloud computing is another primary concern for users. Therefore, although existing studies have proposed privacy-preserving techniques for GNNs in cloud environments, their additional computational and communication overhead remain relatively high, causing high financial costs that limit their widespread adoption among users. To protect GNN privacy while lowering the additional financial costs, we introducePanther, a cost-effective privacy-preserving framework for GNN training and inference services in cloud environments. Technically,Pantherleverages four-party computation to asynchronously executing the secure array access protocol, and randomly pads the neighbor information of GNN nodes. We prove thatPanthercan protect privacy for both training and inference of GNN models. Our evaluation shows thatPantherreduces the training and inference time by an average of 75.28% and 82.80%, respectively, and communication overhead by an average of 52.61% and 50.26% compared with the state-of-the-art, which is estimated to save an average of 55.05% and 59.00% in financial costs (based on on-demand pricing model) for the GNN training and inference process on Google Cloud Platform.

Ring Confidential Transaction (RingCT) protocols are widely used in cryptocurrencies to protect user privacy. Consequently, a corresponding digital signature scheme, such as a ring signature scheme that hides the signers’ identities, is required. Accordingly, the security of a RingCT protocol depends on the confidentiality of the secret signing keys of the underlying ring signature scheme. However, existing solutions like hardware wallets, Trusted Execution Environments (TEEs), and threshold signature schemes have limitations such as specified expensive hardware, targeting attacks at CPUs on insufficiently secure hardware, and overheads caused by multiple parties. On the contrary, program obfuscation for signature schemes offers advantages over these existing approaches. Concretely, we propose a novel obfuscator that secures the secret keys of the concise linkable spontaneous anonymous group (CLSAG) signature scheme, which is the latest ring signature scheme used in Monero’s RingCT protocol. To achieve enhanced security, the proposed obfuscator leverages Paillier homomorphic encryption to transform secret keys into an obfuscated form resistant to attacks. The security of the proposed obfuscator has been formally proved. Computational efficiency has been both theoretically analyzed and experimentally evaluated with positive results on various testing platforms.

With the development of the industrial Internet, industrial Internet data has been growing rapidly, and so has the need for secure communications. In the context of industrial communication, it is essential to establish an effective session key between untrusted nodes. The prevailing key management schemes concentrate on key negotiation with the assistance of a central node through a man-in-the-middle approach. However, industrial field environments are typically characterised by harsh conditions, and any node may be damaged or subject to malicious compromise. This can result in the complete paralysis of communication within the entire system. Consequently, existing key management schemes are unable to fulfil the requisite performance requirements. In contrast to previous centralized or polycentric schemes, we propose an asymmetric bivariate polynomials-based novel efficient decentralized key management scheme (ABP-DKM). ABP-DKM achieves threshold switching through a twice-distribution method, which is more secure than other existing schemes. During the whole key negotiation process, ABP-DKM is decentralized. ABP-DKM is capable of not only peer-to-peer communication but also intra-group communication with forward and backward secrecy. The proposed scheme is more secure and efficient than the existing schemes.

Machine learning has profoundly influenced various aspects of our lives. However, privacy breaches have caused significant unease and concern among the general public. Preserving the privacy of sensitive data during the training and inference phases of machine learning is a key challenge. Cryptography-based privacy-preserving machine learning (crypto-based PPML) offers a viable solution to this challenge. In this article, we studied over 100 publications on crypto-based PPML frameworks published between 2016 and 2024, including 55 client-server architecture frameworks and 64 multi-party architecture frameworks. We provide a comprehensive overview of these frameworks, highlighting their features across various dimensions. Furthermore, we conduct an in-depth analysis, delving into scenarios, privacy goals, threat models, and optimization techniques that underpin these innovative solutions. We also discuss the challenges in the field of crypto-based PPML, including aspects of security and privacy , efficiency , and availability and usability . Finally, we offer an outlook on future research directions, aiming to provide valuable insights for both scholars and practitioners.

Video has become the primary channel for people to access information and engage in leisure and entertainment activities. Video service platforms have emerged to meet these demands and create significant profits. These platforms aggregate diverse videos and offer subscription services, aiming to provide convenient and personalized viewing experiences. To safeguard the copyright of video content against unauthorized access, these platforms employ digital rights management (DRM) technologies. However, the use of DRM faces significant security challenges due to the presence of malware or malicious users, which can compromise the decryption program running on user devices and lead to illegal video content distribution. This paper proposes a high-performance encryption scheme for video DRM protection. We establish the security of our proposed scheme through rigorous mathematical proof and analyze its effectiveness in the context of the video DRM scenario. Furthermore, the experimental results demonstrate that our scheme outperforms other schemes. Overall, our proposed scheme offers a promising solution to the security challenges faced by DRM technologies in protecting video content. Its efficiency, security, and resilience to various attacks make it a viable option for video service platforms seeking to enhance the security of their digital content.

Digital identities and credentials are important for authentication and authorization. Most contemporary digital identity systems rely on a central service provider, which may lead to privacy and centralization issues, such as data lost and information misuse. To address the problems, selective disclosure approaches within the self-sovereign identity have been proposed. However, these approaches primarily focus on data minimization, often overlooking the need for presentation unlinkability and identity supervisibility. In this paper, we propose a blockchain-based selective disclosure approach that supports presentation unlinkability, attribute aggregation, and identity traceability. We implement our approach and evaluate its efficiency by comparing it with approaches including atomic credentials, hashed values and selective disclosure signatures. The results demonstrate that our approach successfully achieves all intended objectives with high efficiency, rendering it applicable to realistic scenarios such as cross-chain authentication and access control in IoT.

In cloud computing environments, virtual machines (VMs) running on cloud servers are vulnerable to shared cache attacks, such as Spectre and Foreshadow. By exploiting memory sharing among VMs, these attacks can compromise cryptographic keys in software modules. Program obfuscation serves as a promising countermeasure against key compromises by transforming a program into an unintelligent form while preserving its functionality. Unfortunately, for certain cryptographic algorithms such as the digital signature schemes, it is extremely difficult to construct provably secure obfuscators using traditional obfuscation approaches. To address such a challenge, this study proposes a novel approach to construct obfuscators for cryptographic algorithms named space-hard obfuscation, which can mitigate the threats from adversaries with the capability of acquiring a limited size of memory in shared cache attacks. Considering the extensive use of the Elliptic Curve Digital Signature Algorithm (ECDSA) in cloud-based Blockchain-as-a-Service (BaaS) and its potential vulnerability to shared cache attacks, we construct an exemplary scheme with provable security using space-hard obfuscation for ECDSA. Experimental results have demonstrated the scheme's high efficiency on cloud servers, as well as its successful integration with Hyperledger Fabric and Ethereum, two widely used blockchain systems.

Blockchain systems, such as Bitcoin and Ethereum, enable new applications, such as cryptocurrencies and smart contracts, using decentralized consensus without trusted authorities. Since the most widely used technique, proof-of-work, suffers from the costs of high latency and huge energy consumption, a number of blockchain systems based on proof-of-stake techniques have been proposed in recent years, many of which use verifiable random functions as fundamental building blocks, such as Ouroboros, Algorand, and Dfinity, etc. The secret key of a verifiable random function scheme, similar to that of a digital signature scheme, is critical to the security of a verifiable random function and the entire blockchain system built on it. To protect the secret keys of verifiable random functions and maintain the efficiency of the proof-of-stake protocol, we extend the objective of cryptographic program obfuscation to verifiable random functions and propose a novel obfuscatable verifiable random function scheme. In particular, we propose an obfuscator that can transform the implementation of the scheme's random string generation algorithm and the given secret key into an unintelligible form. Obfuscated implementations of the random string generation algorithm are deployed on peers of a blockchain for supporting normal routines of the proof-of-stake protocol. Even if a hacker has controlled a peer's host, the owner's secret key will not be compromised because the key has been hardwired into the obfuscated implementation in an “encrypted manner”. We formally prove the correctness and the security of the proposed verifiable random function and obfuscator. Since the proposed scheme supports the general semantics of verifiable random functions, it can be used as a building block for all blockchain systems that adopt proof-of-stake protocols based on Verifiable Random Functions (VRFs). The extensive experimental result indicated that the scheme performs well on various platforms, such as cloud servers, workstations, PCs, smartphones, and embedded devices.

With the rapid growth of wireless sensor networks, secure data transmission, storage, and distribution in such networks has become an urgent demand. To defend against security risks such as data leakage, key compromise, and unauthorized misuse of data simultaneously, we propose a novel obfuscatable ciphertext-policy attribute-based re-encryption scheme with a specially designed obfuscator. The proposed scheme leverages program obfuscation to transform the re-encryption program codes into an unintelligible form and embed the private keys into the obfuscated implementation. Consequently, the proposed scheme protects data confidentiality and keeps the secrecy of the private key while providing fine-grained access control. Formal proofs for the security of the proposed re-encryption scheme and the obfuscator are provided. Extensive experiments have been conducted on representative platforms, including cloud servers, workstations, and embedded devices, to evaluate the computational efficiency and energy consumption of the scheme. Experimental results indicate that the scheme achieves high efficiency on various platforms and economical energy consumption on typical embedded devices with constrained resources.

White-box cryptography is widely used in Internet of Things (IoT) devices to ensure data confidentiality. However, the traditional implementations of white-box cryptographic algorithms (WBCAs) are inefficient and impractical for IoT devices that require fast encryption and decryption of large amounts of data. To address this challenge, we propose a framework that can accelerate WBCAs employed by IoT devices with Graphics Processing Units (GPUs). Our framework leverages the inherent multithreading capabilities of GPUs to simultaneously perform a large number of table lookups required in WBCAs. Additionally, we employ pipelined execution strategies to enhance performance further. To demonstrate the effectiveness of our framework, we apply it to two well-known WBCAs and evaluate its performance on two IoT devices. The experimental results show that our framework can improve performance by up to 70 times compared with the execution on a single-threaded CPU.

Nowadays, with the development and massive use of cloud services, protecting the information security on untrusted cloud servers is becoming an important issue. Since the signing key is one of the most important parts of secret information and the Schnorr signature is frequently used in on-cloud systems, this study proposes an obfuscatable proxy signature scheme based on the Schnorr signature and the Elgamal encryption, and presents a provably secure obfuscator for the scheme. Moreover, the application scenario of the proposed scheme for untrusted cloud servers has been discussed. Experimental results have indicated that the proposed scheme is computationally efficient.

The threshold digital signature technique is important for decentralized finance (DeFi) applications such as asset custody and cross-chain interoperations. The Edwards-curve digital signature algorithm (EdDSA) is widely used in blockchains, e.g., Libra/Diem; however, no suitable threshold solution exists. Therefore, to bridge this gap, we propose a threshold EdDSA that allows n parties to generate keys in a decentralized and distributed manner. Any t + 1-of-n parties can generate standard EdDSA signatures. This scheme supports an arbitrary threshold (t, n) and has been proven to be secure against at most t malicious adversaries. The theoretical analysis (computation complexity and communication footprints) and experimental results demonstrate that the proposed scheme performs efficiently on cloud servers and embedded devices. Furthermore, the proposed scheme is integrated with Tendermint, a blockchain framework that uses EdDSA, to generate keys and sign transactions in a decentralized manner, which indicates that this scheme is compatible with blockchains for supporting DeFi applications.

Deep learning has become a key technology on modeling large amounts of multi-sourced data. For privacy concerns, the data sharing among companies and organizations is increasingly difficult. In this paper, we present a crowd-sourced federated learning solution to train neural networks with a hybrid blockchain architecture. Smart contracts are used to share data authentications on the main chain, where the proxy re-encryption is for the privacy preserving. A consensus-based asynchronous practical byzantine federated learning (APBFL) algorithm is proposed on the side chains, to improve the model reliability and security. Experiments show that our solution is efficient, secure and robust.

In cloud computing, since virtual machines (VMs) running on the same physical server share CPU caches, adversaries can exploit CPU's vulnerabilities to launch shared cache attacks (e.g., Spectre vulnerability) for illegally accessing sensitive data (e.g., key of symmetric encryption) on other VMs. Since it is difficult to fix such vulnerabilities, in this paper, we propose a novel solution that leverages two salient features of white-box encryption to protect data against such attacks: white-box encryption turns the keys and code into unintelligible programs; it is provably secure even if part of its critical data is accessed by adversaries. Although there are many white-box schemes, they cannot be used in our solution due to their limitations. Therefore, we propose a new white-box encryption scheme with highly efficient instances. These instances are parameterized, and can be configured according to the tradeoff between security margin and storage cost. Moreover, our scheme is provably secure in the space-hardness model. The evaluation shows that our solution works well in public clouds and outperforms other methods.

Malware authors are abusing packers (or runtime-based obfuscators) to protect malicious apps from being analyzed. Although many unpacking tools have been proposed, they can be easily impeded by the anti-analysis methods adopted by the packers, and they fail to effectively collect the hidden Dex data due to the evolving protection strategies of packers. Consequently, many packing behaviors are unknown to analysts and packed malware can circumvent the inspection. To fill the gap, in this paper, we propose a novel hardware-assisted approach that first monitors the packing behaviors and then selects the proper approach to unpack the packed apps. Moreover, we develop a prototype named Happerwith a domain-specific language named behavior description language (BDL) for the ease of extending Happerafter tackling several technical challenges. We conduct extensive experiments with 12 commercial Android packers and more than 24k Android apps to evaluate Happer. The results show that Happerobserved 27 packing behaviors, 17 of which have not been elaborated by previous studies. Based on the observed packing behaviors, Happeradopted proper approaches to collect all the hidden Dex data and assembled them to valid Dex files.

In this article, with the rapid development of industrial Internet of Things, a large number of embedded devices, such as sensors and tag readers, have been widely deployed for gathering and sending data. These devices are commonly unreliable and vulnerable to many threats, because they are located in unattended areas which are vulnerable to device capture attacks. Such environments can be regarded as white-box attack contexts, in which the adversary has total visibility and full control of the implementations. White-box cryptography (WBC) aims to protect implementations of symmetric encryption algorithms in white-box attack contexts. Unfortunately, existing WBC schemes are vulnerable to various attacks, and most of them are insufficiently secure in strict white-box attack contexts. Based on the investigation of existing designs and the corresponding cryptanalysis, we propose a novel design approach for securing WBC schemes, which is named state-dependent selectable random substitutions (SDSRS). It uses SDSRSs to defeat various related white-box cryptanalytic approaches. With special considerations for IIoT systems, such as high performance for supporting real-time applications and small block size for fitting industrial protocols, a concrete WBC scheme designed with the proposed approach has been provided. Our theoretical analysis shows that the proposed scheme is secure. Additionally, experimental results indicate that the scheme performs well in practice, and it is significantly efficient in time and energy consumptions compared with existing secure white-box cryptographic schemes.

Distributed embedded devices are widely used in sensor networks and the Internet of Things for gathering and sending data. Many of them are deployed in an unattended manner (e.g., sensor nodes and tag readers), while others may be easily lost (e.g., smart wristbands and watches). These distributed embedded devices could be potentially captured and accessed in an unauthorized manner due to their physical natures. From a security perspective, they are typically working in the white-box attack context, where adversaries have total visibility on the implementations of built-in cryptosystems and full control over their execution processes. It is undoubtedly a significant challenge to deal with white-box attacks on these devices. Existing encryption algorithms for white-box attack contexts require large memory footprint and thus are not suitable for resource- constrained embedded devices. To address this challenge, we propose a novel light-weight encryption scheme for protecting data confidentiality. The encryption is conducted with specialized secret components, and the encryption algorithm requires a small volume of static data for storing critical information. In addition, this scheme uniquely supports efficient key-updating at very small cost. The security and the cost of the proposed scheme have been theoretically analyzed with positive results, and the extensive experimental evaluations indicate that the new scheme satisfies the requirements of distributed embedded devices in terms of limited memory usage and low computational cost.

Mobile agents are useful in collaborative business systems due to their mobility and autonomy, which can roam over the Internet to purchase goods and services on behalf of their owners. However, given attacks from a malicious host, it is a challenge to securely sign a contract on behalf of the owner (the original signer). In this paper, we propose an intrusion-resilient undetachable digital signature (IR-UDS) approach to mitigate the security risk of signing key leakage on the signer's host, base device, and potentially malicious remote hosts, as well as the risk of misusing the signing algorithm on remote hosts. An attacker will be unable to forge the past and future signatures as long as the base device is secure, even if the current signing key of the original signer has been gained. When the base device is compromised, although the future signatures could be forged, all past signatures remain secure. Furthermore, the encrypted signing function has been combined with the original signer's requirement to prevent the misuse of signing algorithm and the exposure of original signing key on malicious hosts. Security analysis has indicated that our scheme can defeat a variety of attacks, and experimental evaluations have demonstrated the good performance of the scheme.

Real-time collaborative programming allows programmers to concurrently edit shared source code over communication networks. To support semantic conflict prevention, prior work has proposed a bask dependency-based automatic locking (DAL) approach to automatically grant locks on source code regions with dependency relationships, under the assumptions that there exists no locking-scope overlapping among concurrent editing operations, and the source code structure remains static during the collaboration process. To address major restrictions of the basic DAL scheme, this paper presents a shard-locking approach and techniques to fully support unconstrained real-time collaborative programming with semantic conflict prevention. The approach allows multiple programmers to concurrently edit source code regions with overlapping locking scopes in the presence of concurrent editing operations and dynamic source code structures. The techniques and solutions have been implemented in a research prototype for evaluations.

Signcryption is a cryptographic technique for simultaneously performing both digital signature and data encryption. It is effective for protecting the confidentiality and unforgeability of communications in Internet of Things (IoT) systems, especially when a number of generated ciphertexts can be aggregated into a compact form. However, device capture attacks are commonly threatening the implementations of signcryption on unattended devices by enabling an attacker to extract the cryptographic key from a captured device. Motivated by this issue, we propose a novel and specialized obfuscatable aggregatable signcryption scheme (OASC) together with an obfuscator for the signcryption algorithm, which has been designed by taking into account that the computational and communication costs should be sufficiently small (light-weighted) to fit applications in resource-constrained embedded devices. The proposed obfuscator can protect signcryption programs from key-extraction attacks by transforming the programs into unintelligible obfuscated programs. To the best of our knowledge, this is the first OASC in the community. The scheme's security features with respect to obfuscation, confidentiality, and unforgeability have been theoretically proved. Moreover, in comparison with other (nonobfuscatable) aggregatable signcryption schemes, the scheme's computational efficiency is positioned at a medium level while the communication cost is also relatively small, with extra unique security features benefiting from obfuscation. Experiments on different devices indicated that the proposed scheme performs reasonably well as expected. The scheme is widely applicable for various scenarios of IoT, where information is sent from unattended leaf nodes to a sink point.

The development of location-based applications raises a new challenge to manage and visualize large amounts of geo-tags presented on a web map. The visualization of the geo-tags often leads to a clutter problem, especially in web-mapping systems. We present a new clustering method to reduce the amount of visual clutter. A split smart swap strategy, which has the advantage that it can be applied to a certain data only once at all map scales, is employed in the method. We compare the proposed method to several other methods. Taking the advantage of the one-time running offline, the proposed method is more applicable for the clutter problem.

Wireless Sensor Networks (WSNs) are often deployed in hostile environments and an adversary can potentially capture sensor nodes. This is a typical white-box attack context, i.e., the adversary may have total visibility of the implementation of the build-in cryptosystem and full control over its execution platform - the sensor nodes. Existing encryption algorithms for white-box attack contexts require large memory footprint and hence are not applicable for wireless sensor networks scenarios. As a countermeasure against the threat in this context, a lightweight secure implementation of the symmetric encryption algorithm SMS4 is proposed. The basic idea of our solution is to merge several steps of the round function of SMS4 into table lookups, blended by randomly generated mixing bijections. Its security and efficiency are analyzed. Evaluation shows our solution satisfies the requirement of sensor nodes in terms of limited memory size and low computational costs.