With the integration of intelligent medical devices and wireless sensor networks, telemedicine queries have significantly enhanced the efficiency and accessibility of medical services. However, the security and privacy of medical query records remain a critical challenge. While existing privacy-preserving query frameworks in Wireless Body Area Networks (WBANs) primarily focus on protecting the privacy of medical records, they often lack robust mechanisms for authenticating query users. This deficiency may enable malicious entities to gain unauthorized access, thereby resulting in data leakage. To address this issue, we propose a privacy-preserving query scheme with unlinkability by integrating double-Schnorr-based anonymous authentication and 1-out-of-n oblivious transfer (OT). Specifically, anonymous authentication is achieved via a double-Schnorr-based credential construction with session-dependent pseudonym blinding, enabling the server to verify user legitimacy without revealing users’ real identities. OT is implemented through a short-hash-based keyword partition mechanism and masked encryption, ensuring that the curious server cannot infer user queries while fulfilling them correctly. Additionally, query privileges are restricted to legitimate users by verifying anonymous credentials issued during the registration phase. The formal security analysis confirms that the proposed scheme achieves essential security attributes, including anonymous authentication, query privacy, and unlinkability. Experimental results demonstrate that the proposed scheme not only integrates anonymous authentication but also outperforms baseline schemes in both computational and communication efficiency. Therefore, the proposed scheme is more suitable for real-world deployment within IoT-based healthcare ecosystems.

In the last few years, face recognition technology gains a huge development because of the popularity of intelligent devices and the development of computer vision. However, due to the fact that face images often contain important sensitive information, the leakage of sensitive information will have a serious impact on both data owners and users. Recently, a series of excellent privacy-preserving face recognition schemes has been proposed. However, existing face recognition schemes focus on the completion of recognition, but ignore the security issues during the recognition process, and the recognition results need to be transmitted through a secure additional channel. Although some schemes have addressed several security issues in the recognition process and protected face information through authentication, the additional identification or random numbers are needed to assist authentication. To address these issues, we propose a lightweight privacy-preserving face authentication scheme based on local sensitive hashing. First, in order to protect face information, the proposed scheme encrypts the face information and stores the encrypted face on the server. Second, the local sensitive hashing function can be used for fast retrieval of face images, which can help servers quickly match encrypted eigenface in the ciphertext domain. Third, the lightweight authentication scheme ensures that users only need to use face information to access the terminal without any additional authentication credentials. Meanwhile it can also achieve user anonymity, prevent internal server attacks, and so on. The comparison with related schemes indicates that the proposed one has fewer computation overhead and communication overhead.

Recently, deep reinforcement learning (DRL) has emerged as a promising approach for robotic control. However, the deployment of DRL in real-world robots is hindered by its sensitivity to environmental perturbations. While existing whitebox adversarial attacks rely on local gradient information and apply uniform perturbations across all states to evaluate DRL robustness, they fail to account for temporal dynamics and statespecific vulnerabilities. To combat the above challenge, we first conduct a theoretical analysis of white-box attacks in DRL by establishing the adversarial victim-dynamics Markov decision process (AVD-MDP), to derive the necessary and sufficient conditions for a successful attack. Based on this, we propose a selective state-aware reinforcement adversarial attack method, named STAR, to optimize perturbation stealthiness and state visitation dispersion. STAR first employs a soft mask-based state-targeting mechanism to minimize redundant perturbations, enhancing stealthiness and attack effectiveness. Then, it incorporates an information-theoretic optimization objective to maximize mutual information between perturbations, environmental states, and victim actions, ensuring a dispersed state-visitation distribution that steers the victim agent into vulnerable states for maximum return reduction. Extensive experiments demonstrate that STAR outperforms state-of-the-art benchmarks

In rice breeding, improving both crop yield and quality is of paramount importance. This study investigates key factors directly influencing yield, particularly the number of rice panicles and leaf width. We hypothesize that increases in panicle count and leaf width correlate positively with photosynthetic efficiency, thereby significantly affecting overall crop yield. To test this hypothesis, we aim to evaluate and select rice varieties exhibiting desirable phenotypes through targeted detection techniques. We utilize an enhanced DINO (self-distillation with no labels) model for detecting and segmenting rice panicles and leaves. The upstream component of our model functions as an unsupervised general feature extractor, learning rich visual features from a large dataset of unlabeled rice images. The downstream task consists of two branches: one for detecting the number of rice panicles and another for segmenting leaf areas. By combining these two branches, we are able to accurately assess the photosynthetic potential and reproductive capacity of rice plants. Experimental results demonstrate that our model outperforms traditional methods in both panicle detection and leaf area segmentation, achieving higher accuracy and robustness. We conduct experiments on a newly curated dataset, RiceVar, which comprises over 50,000 images covering three rice cultivars captured under varied angles and backgrounds. Our proposed method achieves a mean average precision of 81.401 in panicle detection, a 14.7 point improvement over ResNet50, and a Dice coefficient of 84.322 and intersection over union of 82.186 in leaf segmentation, outperforming the EAPT model by 14.08 and 2.45 points, respectively. Moreover, our model remains stable under varying environmental conditions, highlighting its practical value for rice breeding applications. By precisely evaluating panicle count and leaf width, our model supports the selection of high-yield, high-efficiency rice varieties, contributing to the advancement of sustainable agricultural practices. The relevant code and data are available at https://github.com/xiaobeial/Semi-supervised-detection-and-segmentation-algorithm-for-efficient-rice-breeding.

Deep Reinforcement Learning (DRL) suffers from uncertainties and inaccuracies in the observation signal in real-world applications. Adversarial attack is an effective method for evaluating the robustness of DRL agents. However, existing attack methods targeting individual sampled actions have limited impacts on the overall policy distribution, particularly in continuous action spaces. To address these limitations, we propose the Distribution-Aware Projected Gradient Descent attack (DAPGD). DAPGD uses distribution similarity as the gradient perturbation input to attack the policy network, which leverages the entire policy distribution rather than relying on individual samples. We utilize the Bhattacharyya distance in DAPGD to measure policy similarity, enabling sensitive detection of subtle but critical differences between probability distributions. Our experiment results demonstrate that DAPGD achieves SOTA results compared to the baselines in three robot navigation tasks, achieving an average 22.03% higher reward drop compared to the best baseline.

Deep reinforcement learning (DRL) has emerged as a promising approach for robotic control, but its real-world deployment remains challenging due to its vulnerability to environmental perturbations. Existing white-box adversarial attack methods, adapted from supervised learning, fail to effectively target DRL agents as they overlook temporal dynamics and indiscriminately perturb all state dimensions, limiting their impact on long-term rewards. To address these challenges, we propose the Adaptive Gradient-Masked Reinforcement (AGMR) Attack, a white-box attack method that combines DRL with a gradient-based soft masking mechanism to dynamically identify critical state dimensions and optimize adversarial policies. AGMR selectively allocates perturbations to the most impactful state features and incorporates a dynamic adjustment mechanism to balance exploration and exploitation during training. Extensive experiments demonstrate that AGMR outperforms state-of-the-art adversarial attack methods in degrading the performance of the victim agent and enhances the victim agent’s robustness through adversarial defense mechanisms.

Federated learning (FL), as an effective method to solve the problem of “data island”, has become one of the hot and widespread concern topics in recent years. However, with the using of FL technology in the practical applications, an increasing number of FL tasks make the training management be more complex and the trade-off of multi-task becomes difficult. To overcome this weakness, this work proposes a privacy-preserving FL framework with multi-tasks using partitioned blockchain, which can run several different FL tasks by multiple requesters. First, a temporary committee is formed for an FL task to facilitating visualization, organization and management of security aggregation. Second, the proposed framework combines Paillier homomorphic encryption with Pearson correlation coefficient to protect users' privacy and ensure the accuracy of global model. Finally, a new blockchain-based reward method is presented to inspire participants to share their valuable data. The experimental results show that the global model accuracy of our proposed framework is able to reach 98.43%. Obviously, the proposed framework is more suitable for practical application environment, especially in industrial application field.

Federated learning (FL) emerges as a promising paradigm, relentlessly pursuing excellence in efficiency, privacy preservation, and security—the holy trinity that underpins the fundamental philosophy and practical implementation of FL. However, previous scholarly works have predominantly focused on the privacy protection and secure aggregation in the realm of FL. Undoubtedly, efficiency still plays a pivotal role in determining FL’s viability in real-world applications. To improve efficiency while ensuring privacy protection and secure aggregation, this work proposes a verifiable privacy-preserving federated learning framework tailored for practical applications. Firstly, a novel aggregation rule, constrained M maximum security aggregation, forces the server to securely aggregate the local gradients from M users without relying on any auxiliary servers, thereby considerably decreasing the communication overhead. Secondly, regardless of the dimension of the aggregated gradient being verified, our scheme performs a single verification per user per round. Through security analysis, our scheme could guarantee some given security requirements. Besides, extensive experiments show that under the proportion that the gradient dimension$(d)$to the number of users$(n)$is 1:2 and 2:1, ours is$\times 1.3$and$\times 1$faster for total runtime and is$\times 3$and$\times 7$lower for total communication cost compared with a state-of-the-art framework, respectively. Therefore, our FL framework is more suitable for practical application in real life.

Blockchain-based authentication, as a distributed system, is a significant method to achieve secure service access and provision for the distributed mobile cloud computing (MCC) environment. However, owing to the transparency of blockchain, it remains a challenge to protect users’ access behavior from disclosure. Besides, billions of users in the MCC system may cause storage bottlenecks to the blockchain network. To overcome these challenges, this paper designs two blockchain-based privacy-preserving authentication schemes supporting hierarchical access control for the MCC environment. Both schemes allow users to access multiple services with different permissions after a single registration. To address the challenges of privacy disclosure, we use polynomial commitment to replace the plaintext on the blockchain. Meanwhile, a new verification and updating of the access permission method is proposed using the homomorphic property of polynomial commitment. The first scheme works toward reducing computation costs, which is more suitable for systems with a limited number of service providers (SPs). On the other hand, the second scheme aims to reduce the storage requirements of blockchain, and it provides more efficient hierarchical access control for large-scale scenarios without requiring more storage space. Then, the security analysis demonstrates that the two schemes satisfy multiple security requirements. Finally, a comparative summary is presented to show that our schemes have good performance in computation and communication efficiency and are well suited to the MCC system.

With the worldwide spreading of Coronavirus disease 2019 (Covid-19) pandemic, besides the traditional diagnosing approach, Artificial Intelligence provides additional support for the pre-diagnosis of Covid-19 by using data such as patients' images, and sounds, etc. Being able to recognize Covid-19 positive patients quickly and correctly is the key to preventing the expansion of the disease. However, the existing Covid-19 diagnosis models still face challenges due to the complex network structure and additional medical examination. It takes much time to return a diagnosis result. In this paper, a diagnostic model is proposed as an early work for Covid-19 diagnosis using sound samples. The features of sound signals are expressed by Mel Frequency Cepstral Coefficients, which are input into the Online Sequential Extreme Learning Machine for normal/abnormal detection. Data from an open-source database were used to train the proposed model, the experiments show that using vowel pronunciations the model can achieve an accuracy of 96.4% on average, with about 10 times faster for testing than the Support Vector Machine.

In the last few years, mobile cloud computing (MCC) gains a huge development because of the popularity of mobile applications and cloud computing. User authentication and access control are two indispensable security components in the MCC environment. To the best of our knowledge, they are generally designed in different procedures. Access control can be executed after the authentication completes successfully. In order to improve efficiency, this article constructs an integrated scheme of authentication and hierarchical access control using self-certified public key cryptography (SCPKC) and the Chinese remainder theorem (CRT) for MCC environment. The proposed scheme can achieve mutual authentication while determining the access rights of mobile users without storing any access control list in the MCC service provider side. Besides, we also give a dynamic adding or deletion of MCC service provider to efficiently address potential changes in the hierarchy. The security of our proposed scheme is proved by the random oracle model. Compared with recently related multi-server authentication schemes for the MCC environment, the proposed scheme not only adds a new function of hierarchical access control but also has better computation and communication efficiencies. Therefore, the proposed scheme is more suitable for real-life MCC applications.

By regularly broadcasting information about the road and weather conditions around vehicles, a Vehicular Adhoc Network (VANET) is expected to improve traffic congestion and relieve traffic pressure. To guarantee the safe communication of vehicles, the security and privacy problems are urgent to be solved in VANETs. Until now, numerous conditional privacy-preserving authentication (CPPA) schemes have been proposed to solve these problems. However, cross-region authentication still not being taken seriously in CPPA scheme. Although several blockchain-based CPPA scheme have being proposed to realize cross-region authentication, there still exist some imperfection such as frequent interactions or unlinkability. Therefore, this paper proposes a novel blockchain-assisted authentication scheme to solve these existing issues. First, we adopt a puncturable pseudorandom function to complete identity authentication of vehicles. Second, a key derivation algorithm is used to achieve unlinkability of messages. Third, we achieve cross-region authentication by taking advantage of the properties of blockchain. The security analysis shows that the proposed scheme is suitable for VANETs.

In recent decades, as a green industry, electric vehicles (EVs) has a broader development perspective. Vehicle-Smart Grid ecosystem (V2G) intelligently and efficiently configures the electric energy between the electric vehicle and the grid, which can not only improve the efficiency of resource allocation, but also bring economic benefits to every entity in the system. Although the smart grid technology has become increasingly mature, security and privacy issues should not be ignored. To achieve these, a series of privacy-preserving authentication schemes have been designed for the V2G system. However, there are still some security problems like the problem of user anonymity. In order to address these issues, this work proposes a lightweight privacy-preserving authentication scheme using physical unclonable function (PUF). The proposed scheme can resist various known attacks such as impersonation attack and replay attack. Compared with previous related schemes in V2G system, our scheme has obvious advantages in terms of computing and communication overhead.

Until now, the pseudonym ID and one-time hash chain as the key technologies solve anonymous and forward secrecy properties, respectively, in lightweight authentication protocols for wireless sensor networks (WSNs) environment. However, both techniques have to face the limitation of asynchronization attack, which will cause the pseudonym ID and one-time hash chain value between participants to be loss of synchronization when the transmitted message is blocked. Most recently, a serial of lightweight anonymous authentication scheme with forward secrecy (LAASFS) have been built to resolve this problem. Unfortunately, they failed to handle this question satisfactorily. Besides, most of them are still vulnerable to smart card loss attack (SCLA). In this article, to address these issues and make more effective, we combine pseudonym ID, one-time hash chain and tag techniques to construct an LAASFS for WSNs environment. The proposed LAASFS can be adaptive resilience of asynchronization attacks. In addition, it is also able to resist a variety of known attacks like SCLA and wrong password login attack. Formal analyses are taken by BAN logic and ProVerif tool to demonstrate the security of properties of our LAASFS. Compared with several previous related schemes, the proposed LAASFS possesses obvious advantages in computation and communication costs.

With the advances in wireless communication and Internet of things, wireless body area networks (WBANs) have attracted more and more attention because of the potential in improving the quality of health care services. With the help of WBANs, the user can access the patient's life‐critical data generated by miniaturised medical sensors, and remote health care monitoring services are provided. Since the open nature of wireless channel and sensitivity of transmitted information, the security and privacy of such personal data are becoming important issues that must be dealt with. In the past few years, a large number of authentication schemes had been proposed to solve these issues. However, most of the existing schemes are not secure enough. As a step toward this direction, in this study, the authors present a privacy‐preserving authentication scheme with adaptive resilience of desynchronisation attacks for WBANs, in which lightweight crypto‐modules are adopted to pursue the best efficiency. The proposed scheme adopts the pseudonym identity technique to provide user anonymity, and one‐way hash chain technique and serial number method are employed to ensure forward secrecy and resist desynchronisation attack, respectively. Analysis and comparison results demonstrate that the proposed scheme achieves a delicate balance between security and efficiency.

The multiserver architecture authentication (MSAA) protocol plays a significant role in achieving secure communications between devices. In recent years, researchers proposed many new MSAA protocols to gain more functionality and security. However, in the existing studies, registered users can access to all registered service providers in the system without any limitation. To ensure that the system can restrict users that are at different levels and can access to different levels of service providers, we propose a new lightweight hierarchical authentication protocol for multiserver architecture using a Merkle tree to verify user’s authentication right. The proposed protocol has hierarchical authentication functionality, high security, and reasonable computation and communication costs. Moreover, the security analysis demonstrates that the proposed protocol satisfies the security requirements in practical applications, and the proposed protocol is provably secure in the general security model.

On-body wireless networks (oBWNs) play a crucial role in improving the ubiquitous healthcare services. Using oBWNs, the vital physiological information of the patient can be gathered from the wearable sensor nodes and accessed by the authorized user like the health professional or the doctor. Since the open nature of wireless communication and the sensitivity of physiological information, secure communication has always been the vital issue in oBWNs-based systems. In recent years, several authentication schemes have been proposed for remote patient monitoring. However, most of these schemes are so susceptible to security threats and not suitable for practical use. Specifically, all these schemes using lightweight cryptographic primitives fail to provide forward secrecy and suffer from the desynchronization attack. To overcome the historical security problems, in this paper, we present a lightweight and secure three-factor authentication scheme for remote patient monitoring using oBWNs. The proposed scheme adopts one-time hash chain technique to ensure forward secrecy, and the pseudonym identity method is employed to provide user anonymity and resist against desynchronization attack. The formal and informal security analyses demonstrate that the proposed scheme not only overcomes the security weaknesses in previous schemes but also provides more excellent security and functional features. The comparisons with six state-of-the-art schemes indicate that the proposed scheme is practical with acceptable computational and communication efficiency.