Linearly homomorphic signatures (LHS) allow users to outsource signed datasets to a server for computing linear functions. Users can verify the correctness of results returned by the server using public keys. Constructing a LHS scheme with short public keys in the standard model is a challenging and useful open problem. Most existing standard-model LHS schemes have public key sizes of O(N) that are linearly proportional to the dataset size N. Only one scheme achieves sublinear public keys of size ON12. This limits LHS schemes from efficiently processing large-scale datasets in the most secure way; for a dataset with ≥1014 data items (e.g., LHC dataset), the public key size of the existing standard-model LHS schemes may exceed one GB, imposing significant storage costs.In this paper, we propose a new LHS scheme with adaptively sublinear public keys of size ON1c in the standard model, where the value of c can be adaptively selected by users according to different dataset sizes N to minimize N1c. For a large dataset with N=1015 data items, compared to the state-of-the-art LHS scheme, our scheme reduces the public key size and key generation time by approximately a factor of 105, and the offline verification time by approximately a factor of 104, while maintaining comparable performance in the other algorithms. Our scheme enables verifiers to verify the correctness of results calculated on large datasets with extremely low storage overhead (KB level) and in the most secure way.

The rapid development of the automotive industry and the Internet of Vehicles (IoV) has led to an exponential growth of distributed vehicular data, driving the need for secure and efficient collaborative machine learning solutions. However, existing distributed collaborative machine learning (DCML) approaches, such as federated learning and split learning, face significant challenges in IoV scenarios, including limited training efficiency, centralized aggregation vulnerabilities, and constrained privacy and model protection. To address these issues, we propose a blockchain-based split federated learning (BSFL) scheme for IoV applications. BSFL non-trivially combines federated learning and split learning to enable vehicles with low computational power to participate in parallel training, improving both model accuracy and training efficiency. By utilizing blockchain as a decentralized infrastructure, BSFL eliminates the risks of single points of failure and ensures model consistency through Byzantine fault-tolerant consensus. Furthermore, we design a noise addition mechanism based on differential privacy to safeguard client data privacy and model security. Formal security analysis and extensive experiments demonstrate that BSFL achieves enhanced privacy, security, and training performance. Comparing to related DCML schemes, BSFL reduces computational overhead by up to 88.84% and client training time by up to 29.49% while maintaining comparable accuracy. When training on ResNet-50 based on CIFAR10, BSFL achieved an accuracy of 93.15%. And the verification process for each model’s training results on the blockchain requires 1.49 ms.

The vigorous development of the Internet has spurred exponential data growth, yet data is predominantly stored in isolated user entities, hampering its full value realization. In large-scale deployment of “AI+industries” such as smart medical care, intelligent transportation and smart homes, the gap between data supply and demand continues to widen, and establishing an effective data sharing mechanism is the core of promoting high-quality industrial development. However, data sharing faces significant challenges in security, performance, and functional adaptability. Privacy-enhancing encryption technologies, including Attribute-Based Encryption (ABE), Proxy Re-encryption (PRE), and Searchable Encryption (SE), offer promising solutions with distinct advantages in enhancing security, improving flexibility, and enabling efficient sharing. Statistical analysis of relevant literature from 2020 to 2025 reveals a rising research trend in ABE, PRE and SE, focusing on their data sharing applications. Firstly, this work proposes a data sharing process framework and identifies 20 potential attacks across its stages. Secondly, this work integrates ABE, SE, PRE with 12 enhancement technologies and examines their multi-dimensional impacts on the security, performance, and functional adaptability of data sharing schemes. Lastly, this work outlines key application scenarios, challenges, and future research directions, providing valuable insights for advancing data sharing mechanisms based on privacy-enhancing encryption technologies.

Postquantum key-encapsulation mechanisms (KEMs) usually use the well-known Fujisaki-Okamoto (FO) transformation during key decapsulation to achieve chosen ciphertext attack (CCA) security. In the FO transformation, the re-encryption procedure depends on the message. The side-channel leakage of re-encryption can be exploited to recover the coefficients of the secret key under chosen ciphertexts, even if the KEM scheme is CCA-secure. However, it still requires a large amount of trace during the profiling and attacking phase. In this work, we introduce the first nonprofiled deep learning-based attack (NPDLA) on lattice-based KEMs. We propose a chosen ciphertext method that is suitable for NPDLAs. The horizontal chunking strategy is employed to partition the coefficients into chunks, enabling the independent recovery of multiple secret key coefficients within each chunk. Then, we adopt an incremental learning strategy to allow the deep learning model to gradually learn the knowledge of each chunk. Moreover, we push the limits of traditional NPDLAs by combining the unsupervised domain adaptation with the correlation distinguisher, eliminating the necessity of neural network training for each possible key guess. The feasibility of the attack is verified by practical experiments for the unprotected and masked implementations of Kyber on the ARM Cortex-M4.

Publicly verifiable secret sharing (PVSS) allows a dealer to share a secret among a set of shareholders so that the secret can be reconstructed later from any set of qualified participants. In addition, any public verifier should be able to check the correctness of the sharing and reconstruction process. PVSS has been demonstrated to yield various applications, such as e-voting, decentralized random number generation protocols, and secure computation on distributed networks. Although many concrete PVSS protocols have been proposed, their security is either proven in the random oracle model or relies on quantum-vulnerable assumptions such as factoring or discrete logarithm. In this work, we propose a generic construction of a PVSS from any public key encryption scheme and non-interactive zero-knowledge arguments for suitable gap languages. We then provide lattice-based instantiations of the underlying components, which can be proven in the standard model. As a result, we construct the first post-quantum PVSS in the standard model, with a reasonable level of asymptotic efficiency.

API keys are essential authentication credentials for accessing online services, yet insufficient protection can cause severe breaches, from unauthorized data exfiltration to largescale service abuse. Existing research has investigated API key leakage in platforms like GitHub, mobile applications, and cloud storage, but the browser extension ecosystem remains largely neglected. Previous work relies largely on predefined patterns. This limitation is especially problematic for browser extensions, where API keys are prevalent and often embedded in novel or transformed forms. We introduce an entropy-guided heuristic detection that performs API key inference within the program context of network requests. By constructing program dependence graphs, our method traces key propagation, reconstructs transformed keys, and heuristically correlates multiple API requests to uncover diverse key formats without relying on strict pattern definition. Applied to 21,192 real-world Chrome extensions, our system detected 359 leaked API keys from 286 extensions and 125 distinct services, including 57 previously undocumented formats (176 leaks) invisible to the regex-based tool, achieving a 31.8% coverage gain. The empirical assessment verified the exploitability of many keys, allowing data leakage, service exploitation, data manipulation, and phishing attacks. We responsibly reported these findings to the Google security team, who classified the issue as P2 priority, highlighting the critical risks of insecure API key usage in browser extensions.

The widespread adoption of public cloud services, such as Dropbox and Google Drive, is driven by their efficiency in storage and sharing. However, the risk of providers accessing user data, even under the guise of encrypted protection, remains a significant barrier to trust. Most attempts to implement end-to-end encryption fail to achieve usability, largely due to their inability to integrate seamlessly or their requirement for complex server-side changes. In this paper, we propose CloudMirror, the first solution to offer truly seamless end-to-end encryption for authenticated personal storage across devices. A distinguishing feature of CloudMirror is its ability to function atop existing cloud infrastructure without necessitating any new programmable operations on either the server or client side. Beyond guaranteeing confidentiality and integrity against curious cloud providers, our system maintains the utility of native cloud functions, including data sharing and fast synchronization. It also enables the simultaneous use of multiple clouds within a single view. From a security perspective, CloudMirror mitigates various attacks by incorporating the buffer offset value as Associated Data (AD) in AEAD encryption. Our evaluations with Dropbox, OneDrive, and Google Drive demonstrate that CloudMirror causes no significant performance degradation and maintains full compatibility with unencrypted files.

Dynamic proactive secret sharing (DPSS) is essential for distributed systems, enabling long-term key escrow, BFT protocol reconfiguration, and confidential state machine replication. Yet existing asynchronous schemes, while crucial for realistic settings, suffer from high communication overhead and poor practicality, limiting real-world deployment. We propose PlainDrop, a concise and efficient DPSS protocol designed specifically for asynchronous networks. PlainDrop achieves optimized communication complexity of $O(n<^>{2})$ via commitment-share decoupling combined with homomorphic threshold encryption techniques. PlainDrop also eliminates the need for expensive distributed key generation (DKG) and complex bivariate polynomial structures by introducing a lightweight silent setup framework and employing direct share processing based on univariate polynomials. We formally prove that PlainDrop provides secrecy, integrity, and termination in asynchronous networks against a mobile adversary corrupting up to one third of the parties. We implement PlainDrop and evaluate it on Amazon EC2 with up to 100 nodes. Our experimental results demonstrate average reductions of 37% and 67% in completion time, and 61% and 89% in communication volume, compared to DyCAPS and LongLive, respectively.

As blockchain technology advances, an increasing number of applications require interactions between smart contracts across multiple blockchains. However, existing cross-chain solutions exhibit limited scalability due to heterogeneous blockchain environments and diverse application requirements. A fundamental challenge lies in the absence of a unified resource definition for cross-chain processes, impeding moderate resource allocation and effective conflict resolution. Specifically, when extended to general cross-chain transactions involving invocations among multiple contracts, these methods lack the capability to correctly handle state transitions for all related contracts. This paper proposes AtomXross, a novel cross-chain scheme that supports arbitrary combinations of smart contracts during the cross-chain process. We build a scalable cross-chain architecture based on a relay chain and a cluster of cross-chain nodes to provide better scalability. We propose a unified definition for cross-chain resources within the system and implement an adaptive resource management mechanism on the relay chain, enabling it to record the invocation relationships of contract functions. When a cross-chain transaction involves multiple contract calls, AtomXross can index the calls and generate the corresponding call tree. To address the challenges posed by potential mutual invocations between smart contracts, we design an atomic transaction protocol based on buckle-lock, an ordered two-tier pessimistic locking mechanism. AtomXross ensures that cross-chain transactions do not conflict with each other while remaining compatible with non-cross-chain calls that may occur at any time. Furthermore, we propose a universal programming template for on-chain smart contracts, which enables ordinary smart contracts to acquire cross-chain capabilities. We implement AtomXross based on Hyperledger Fabric and FiscoBCOS. In comparison to WeCross, AtomXross lowers the gas cost on system initialization and incurs only a 14% increase in transaction latency while supporting inter-contract calls.

The importance of tight security has been increasingly recognised in recent years, as it helps avoid the inflation of security parameters caused by reduction loss, thereby improving efficiency. However, achieving tight security also incurs additional overhead compared to traditional designs. As a result, the efficiency gains offered by tight reductions do not always fully outweigh the costs required to realise them. This trade-off has motivated considerable efforts to design practical schemes that achieve this property without sacrificing appreciable efficiency. In this work, we show this goal can be further advanced by leveraging the Algebraic Group Model (AGM). Our first contribution is a tightly secure digital signature scheme with multi-user security against adaptive corruptions. The construction is proven in the non-programmable ROM (NPROM) + AGM and can be viewed as a simplified variant of Appendix A of AOS02 [2]. Compared to the most efficient known scheme by Diemert et al. [12], our scheme achieves a tight reduction to the DL assumption (instead of DDH), reduces the public key size from 4 to 2 elements and the computational cost from 14 to 5 exponentiations. Building on this, we further present a pairing-free identity-based signature (IBS) scheme that also achieves tight EUF-ID-CMA security under the DL assumption. Compared to the only existing IBS with the same property by Loh et al. [29], our scheme reduces the signature size from 8 to 4 elements and the computational cost from 10 to 6 exponentiations. Moreover, our security proof is in the NPROM + AGM, in contrast to prior works proven in the ROM + AGM. These results demonstrate the practical feasibility of tightly secure signature schemes in real-world applications.

In IoT-enabled electricity markets, trading is often conducted in a forward-style manner, where parties agree on prices and quantities in advance and settle upon future delivery. However, deploying such forward-style exchanges across chains is incompatible with execution-restricted blockchains and typically incurs significant cross-chain gas overhead. Moreover, existing low-overhead cross-chain mechanisms, particularly HTLC-based schemes, inherently rely on pairwise locking and thus cannot efficiently support one-to-many settlement. We propose an inter-chain forward-style exchange framework based on a dual-track validation architecture. The framework integrates a double-spend fraud-proof mechanism to ensure correctness and a proof-of-acceptance (PoAc) mechanism to reduce cross-chain overhead for successful settlements. We further introduce a novel k-directional hash lock enabling one-to-many settlement, allowing a single buy order to be split across multiple sellers. Our analysis shows that the framework resists griefing attacks, prevents adversarial losses, and preserves cross-chain atomicity. Experiments demonstrate support for execution-restricted blockchains such as Bitcoin, scalability to 862,000 sellers per order, and a 78.25% reduction in settlement cost compared to light-client approaches.

Federated Learning (FL) enables collaborative machine learning training while preserving data privacy. However, reliance on a central server of the typical FL confronts the risk of single server failure. Decentralized Federated Learning (DFL) emerges as a promising distributed framework, allowing clients to directly share models without server intervention, thereby addressing this challenge. Nevertheless, due to its decentralized nature, DFL is highly susceptible to Byzantine attacks orchestrated by malicious clients. Existing Byzantine-resilient DFL algorithms, though few, remain vulnerable to adaptive attacks due to their heavy reliance on gradient checks of local models, which can be adaptively manipulated by intelligent adversaries. To tackle this issue, we propose a DFL aggregation scheme called FORCE (Byzantine-Resilient Decentralized Federated Learning via Game-Theoretic Contribution Aggregation). Drawing inspiration from the Shapley value in game theory, FORCE shifts from gradient-checking approaches to employ a universal metric, the loss of the local model-independent of specific gradients, to identify potentially malicious clients. To reduce the computational overhead of FORCE as the number of neighboring clients scales up, we propose a computation-lightweight variant, FORCE ${}<^>{-}$ , which is optimized through approximating Shapley value computation. This variant becomes more scalable for resource-restricted DFL clients that are also aggregators. Experimental results on four diverse datasets under three attacks demonstrate that FORCE outperforms existing state-of-the-art Byzantine-resilient DFL aggregation methods, effectively defending against Byzantine attacks.

With the development of mobile crowdsensing systems (MCSs), wireless network transmission efficiency has attracted widespread attention. Network coding can be used in wireless communication to improve network throughput and robustness, which allows intermediate nodes to perform arbitrary coding operations on data packets. However, the data packet in network coding systems is vulnerable to pollution attacks. The special operation of intermediate nodes makes some security protocols in traditional store-and-forward networks unavailable in network coding systems. To address this problem, an efficient certificate-based linearly homomorphic signature scheme against pollution attacks in network coding systems is presented. A novel homomorphic contraction mapping technique is introduced to reduce the computational cost of signature generation. In the proposed scheme, the computational cost of both signature generation and verification is independent of the data packet size. Furthermore, a construction is provided to simultaneously defend against both eavesdropping attacks and pollution attacks in unicast networks. The security of the certificate-based linearly homomorphic signature scheme is formally proved in the random oracle model (ROM), and the scheme is implemented using the Java Pairing-Based Cryptography (JPBC) library. Simulation results demonstrate that the scheme is efficient and practical for real-world deployments in public environments without requiring secure channels.

With the development of air transportation, Space-Air-Ground Integrated Network (SAGIN) are playing an increasingly important role in optimizing air traffic management and enhancing flight safety for billions of passengers and trillions dollars of aviation industry. As the key technology of SAGIN, the Automatic Dependent Surveillance-Broadcast (ADS-B) system is widely used due to its simple operation, low construction cost, and high information accuracy. However, the security problems in ADS-B system, including lack of identity authentication between all communication links, crucial information transmitted in plaintext, and susceptibility to the single point of failure, have been serious obstacle to its wide application. Existing solutions fail to account for the unique characteristics of ADS-B and SAGIN, leading to inadequate security and poor performance in these specialized contexts. Aiming to solve the above issues and provide security and scalability for ADS-B system, we conduct the following research. Firstly, an enhanced identity-based broadcast signcryption (e-IBBSC) scheme is designed to keep crucial information confidential and all messages authenticated simultaneously. Secondly, we propose an efficient batch message authentication method combined with the Merkle tree and proposed e-IBBSC, significantly improving the ADS-B message utilization ratio from 1.35% to 74.10%. Thirdly, we utilize the sharding blockchain and Byzantine fault tolerance protocol to design the first sharding-based distributed management system for SAGIN that realizes fault tolerance and scalability. Finally, after a detailed security analysis and comprehensive performance evaluation, we demonstrate that our solution can achieve all proposed system goals including security, scalability, and high performance of 1s flight transaction processing latency and 62KTPS throughput.

Graph Neural Networks (GNNs) have powerful representation capabilities for graph data, achieving excellent performance across various fields. Considering the scarcity of labels in real-world scenarios, graph self-supervised learning (GSSL) has gained increasing attention due to its ability to train without relying on labels. However, recent studies have revealed that GNNs are vulnerable to stealthy backdoor attacks in GSSL scenarios, enabling the encoder to learn backdoor features simply by injecting triggers. Existing graph backdoor defense methods mainly focus on supervised settings and cannot be directly transferred to self-supervised scenarios due to the lack of label guidance. To bridge this gap, we propose GDetox, the first backdoor defense approach against backdoored encoders in GSSL. GDetox aims to eliminate backdoor logic in encoders while maintaining the encoder's original performance. Specifically, GDetox can purify the graph backdoor encoder based on the self-supervised distillation approach without relying on label information. Further, we introduce an adversarial contrastive learning that augments node representations without relying on labels to enhance teacher model performance, thereby improving distilled encoder performance. We evaluate the defense performance of GDetox on four node classifications and four graph classification datasets by comparing with four state-of-the-art (SOTA) defense methods against seven latest backdoor attack methods on GSSL. Extensive experiments demonstrate that GDetox far outperforms the SOTA defense methods, reducing the attack success rate to 4% with negligible degradation in encoder performance (within 2%) in both node-level and graph-level tasks.

The convergence of the Internet of Things (IoT) and Software-Defined Networking (SDN) has enabled scalability, dynamic resource allocation, and centralized management of SDN-IoT infrastructures. However, the integration introduces critical security challenges such as cross-layer threats and IoT-specific exploits. While Machine Learning (ML)–based Intrusion Detection Systems (IDS) offer promising defense, their effectiveness is hindered by inadequate and outdated datasets that lack realistic SDN control-plane dynamics and IoT heterogeneity within integrated SDN-IoT environments. Critically, no existing dataset simultaneously captures three dimensions, including SDN control-plane behaviour, IoT device traffic, and cross-layer attack vectors essential for securing modern converged networks. To address the gap, we propose a novel dataset generated from a real-world SDN-IoT testbed. The dataset is validated using five ML models with two classification schemes. The experimental results show high detection performance, with 100

Public key authenticated encryption with keyword search (PAEKS) is a cryptographic primitive applicable in cloud storage systems. It empowers cloud servers to conduct searches on encrypted data without decryption while safeguarding against the brute-force attack known as insider-keyword-guessing attacks (IKGAs). In contrast to the pioneering primitive PEKS, which is vulnerable to IKGAs, PAEKS incurs additional computational and communication overhead due to the sender keys' involvement in encryption and trapdoor-generation processes. Although the recent work improves the efficiency of PAEKS by re-encrypting received ciphertexts, the requirement of a fully trusted proxy is rather costly for users to implement in practice. To reduce the economic cost and to keep a high efficiency, we propose a new primitive of Proxy-free Public-key Authenticated Updatable and Searchable Encryption (PF-AUKS). The key concept is to let the cloud server, instead of the proxy, directly convert different-source ciphertexts into a uniform format securely. We propose a concrete PF-AUKS scheme that supports fast search, constant trapdoor generation, and secure ciphertext update. Theoretical evaluation and experimental results illustrate high algorithm running speed and retrieval efficiency. We formally define the security model of PF-AUKS and prove that our scheme is secure under this model.

We revisit Group Encryption (GE)—an encryption analogue of group signatures introduced by Kiayias et al. (Asiacrypt 2007). A GE system simultaneously provides anonymity and traceability for receivers who are certified group members, enabling a range of privacy-preserving applications. While prior work has extensively addressed how to trace receivers in GE, the question of why a ciphertext should be traceable remains unexplored. Unlike group signatures, where opening can be justified by the signed content, tracing in GE poses a dilemma because the underlying plaintext is confidential. To address this gap, we introduce Group Encryption with Oblivious Traceability (GEOT), an enhanced form of GE in which the traceability of a ciphertext ψ intended for receiver and containing message w is governed by a public tracing policy P(,w) ∈0,1 . Here, P(,w)=0 denotes traceability, whereas P(,w)=1 ensures non-traceability. The traceability status is known to the sender but remains hidden from all parties except the opening authority, which learns nothing about in the non-traceable case. GEOT further supports message filtering and dynamic membership, following Nguyen et al. (PKC 2021). Filtering enforces that valid ciphertexts satisfy a public policy F(w)=1 , while dynamicity enables users to join and leave the system over time. We formalize GEOT with concise syntax and rigorous security notions, and present a modular construction based on standard cryptographic primitives: signatures, public-key encryption, and non-interactive zero-knowledge proofs. We also give a concrete instantiation from code-based assumptions supporting arbitrary tracing and filtering policies represented by polynomial-size Boolean circuits. In addition to expressive filtering and tracing functionalities, our scheme achieves significant efficiency improvements over existing post-quantum GE constructions.