In the era of increasingly complex and covert network attack on Internet of Things (IoT), existing anomaly detection paradigms often exhibit limitations in characterizing the dynamic evolution and cross-source dependencies inherent in multi-source threat scenarios. Most current approaches rely on static graph structures or single-modal features, failing to fully capture the temporal evolution of topology and heterogeneous information within network traffic. To address these challenges, this work has proposed a novel Anomaly Detection method based on Dynamic Graph structure and Multi-Modal Fusion (DGMF-Net) for IoT. First, we have adopted a Conditional Denoising Diffusion Probabilistic Model (CDDPM) to generate high-quality minority class samples, effectively mitigating data imbalance. Second, we have introduced a sliding time window mechanism to construct dynamic graphs, explicitly modeling the temporal dependencies and evolving correlations among traffic features. Furthermore, we have designed a multimodal feature extraction module that learns representations from three dimensions: statistical patterns (via 1D-CNN), temporal sequences (via GRU), and graph structures (via DGCN). Finally, a cross-modal attention fusion mechanism is adopted to dynamically integrate these heterogeneous features, thereby enhancing our model's perception of complex abnormal patterns. Experimental results on the UNSW-NB15 and CIC-IDS-2017 datasets have demonstrated that our proposed method has achieved superior performance, generalization, and robustness compared to state-of-the-art approaches for anomaly detection in a dynamic network environment of IoT with low false alarm rates.

Sequential recommendation aims to predict users’ next actions by modeling temporal dependencies in interaction sequences. However, existing attention- or graph-based models often struggle to jointly capture hierarchical semantics, long-range dependencies, and multi-relational knowledge, leading to suboptimal performance in sparse and cold-start scenarios. In this article, we propose KHARec, a knowledge-enhanced hyperbolic attention network that unifies geometric representation learning and temporal modeling for trustworthy sequential recommendation. KHARec consists of a hyperbolic fine-grain module that encodes hierarchical semantics from knowledge graphs, a Transformer-based coarse-grain module that models global dependencies with an FFT-based filter, and a learnable gating mechanism for adaptive feature fusion. Extensive experiments on three benchmark datasets demonstrate that KHARec consistently outperforms state-of-the-art sequential and knowledge-aware baselines in terms of HR@K and NDCG@K, while maintaining near-linear computational complexity. Further analyses verify the model’s robustness under sparsity and cold-start conditions and its superior capability in handling long-tail items. Overall, KHARec provides an interpretable and efficient framework for integrating hyperbolic reasoning and attention-based sequence modeling in large-scale recommendation environments.

Fast Gradient Sign Method (FGSM)-based ap proaches play a critical role in adversarial attacks, particularly in transfer-based attacks. Recent advancements aim to enhance adversarial transferability by refining the sign function in FGSMs using precise gradient information. However, due to the instability of gradient update directions, such precise gradient information may still converge to local minima, thereby limiting transferability. To address this issue, we propose Majority-vote Gradient Scaling (MVGS), a novel method designed to sta bilize gradient update directions. MVGS dynamically rescales momentum terms by evaluating the consistency of historical gradients at each pixel. Since MVGS operates solely through parameter-specific momentum rescaling, it can be seamlessly integrated with existing adversarial attack methods that lever age precise gradient information. Extensive comparative and ablation experiments demonstrate that MVGS significantly improves attack success rates (ASR), achieving up to a 6.1% increase over the state-of-the-art APAA method, confirming its effectiveness.

Abstract Scientific literature is growing in volume with time. The number of papers published each year by 28 100 journals is 2.5 million. The citation indexes and search engines are used extensively to find these publications. An individual receives many documents in response to a query, but only a few are relevant. The final documents lack structure due to inadequate indexing. Many systems index research papers using keywords instead of subject hierarchies. In the scientific literature classification paradigm, various multilabel classification methods have been proposed based on metadata features. The existing metadata-driven statistical measures use bag of words and traditional embedding techniques, like Word2Vec and BERT, which cannot quantify textual properties effectively. In this paper, we try to solve the limitations of existing classification techniques by unveiling the semantic context of the words using an advanced transformer-based recurrent neural networks (RNN) approach incorporating Dual Attention and layer-wise learning rate to enhance the classification performance. We propose a novel model, BioElectra-BiLSTM-Dual Attention that extracts the semantic features from the titles and abstracts of the research articles using BioElectra-encoder and then BILSTM layer along with Dual Attention label embeddings their correlation matrix and layer-wise learning rate strategy employed for performance enhancement. We evaluated the performance of the proposed model on the multilabel scientific literature LitCovid dataset and the results suggest that it significantly improves the macro-F1 and micro-F1 score as compared to the state-of-the-art baselines (ML-Net, Binary Bert, and LitMCBert).

Abstract Natural language processing models are widely acknowledged for their strong data fitting capabilities, diverse application scenarios, and adaptable learning methodologies. However, these models, including the large language models, exhibit sensitivity to adversarial example attacks. These examples are slightly perturbed from the pristine text but mislead the model classification. Nevertheless, the existing attack methods primarily focus on the attack effectiveness without semantics-preservation considered. Moreover, the trade-off between evasion effectiveness and concealment of perturbed texts is less investigated. In this study, we propose a multi-objective adversarial text generation framework (MOATG) that simultaneously optimizes attack success rate, imperceptibility, and semantic similarity. Tailored objective functions and dominance relations are designed for character-, word-, and sentence-level perturbations. MOATG is evaluated against five baselines across five benchmark datasets. Experimental results show that MOATG achieves a 5.38% average improvement in attack success rate and reduces word error rate by 1.48%, demonstrating its effectiveness in balancing attack strength and stealth.

The rapid growth of electric vehicle (EV) charging networks has exposed critical vulnerabilities to cybersecurity threats, while traditional intrusion detection systems (IDS) struggle with privacy requirements and distributed data challenges. This paper presents a novel knowledge distillation enhanced semi-supervised federated learning framework (KDSS-FLIDS) for intrusion detection in EV charging networks. By integrating federated learning with semisupervised techniques and a knowledge distillation mechanism, our framework effectively utilizes both labeled and unlabeled data distributed across EV charging stations while preserving data privacy. Our specialized knowledge distillation mechanism employs temperature scaling and soft label generation to improve knowledge transfer between clients with heterogeneous data distributions. Extensive experiments on the CICEVSE2024 dataset demonstrate that our method consistently outperforms traditional approaches across various Non-IID settings. In particular, the framework exhibits accelerated convergence rates and substantially lower communication overhead, making it practical for large-scale deployment. This research presents a significant advancement in securing EV charging networks while effectively balancing privacy protection, computational efficiency, and detection accuracy.

In the trend of continuously advancing urban intelligent transport construction, traditional traffic signal control (TSC) struggles to make effective decisions with complex traffic conditions. Although multi-agent deep reinforcement learning (MARL) shows promise in optimizing traffic flow, most existing studies ignore the complex relationships between signal lights and fail to communicate with neighbors effectively. Moreover, the deterministic strategies generated by Q-learning-based methods struggle to be extended to large-scale urban road networks. Therefore, this paper proposes a multi-agent graph-based soft actor-critic (MAGSAC) approach for TSC, which combines graph neural networks with the Soft Actor-Critic (SAC) algorithm and extends it to multi-agent environments to address the TSC problem. Specifically, we employ graph-based networks and attention mechanism to expand the receptive domain of agents, enable environmental information to be shared among agents, and utilize the attention mechanism to filter out unimportant information. The algorithm adheres to the Centralized Training Decentralized Execution (CTDE) paradigm to minimize the non-stationarity of MARL. Finally, a rigorous experimental evaluation was conducted using the CityFlow simulator on both synthetic traffic grids and real-world urban road networks. Experimental results show that MAGSAC outperforms other TSC methods in performance metrics, including average queue length and waiting time, and achieves excellent performance under complex urban traffic conditions.

In the rapidly evolving landscape of cloud-edge computing, efficient resource scheduling across Kubernetes clusters is essential for optimizing microservice deployment. Traditional scheduling methods,e.g.,heuristic and meta-heuristic algorithms, often struggle with the dynamic and heterogeneous nature of cloud-edge environments, relying on fixed parameters and lacking adaptability. We propose and implement DRKC, a novel deep reinforcement learning-based approach that addresses these challenges by improving resource utilization and balancing workloads. We model the scheduling problem as a Markov decision process, enabling DRKC to automatically learn optimal scheduling policies from real-time system data without relying on predefined heuristics. The work synthesizes state information from multiple clusters, using multidimensional resource awareness to effectively respond to changing conditions. We evaluate our performance in three Kubernetes clusters with thirteen nodes and ninety-six test applications with different resource requirements. Experimental results validate the effectiveness of DRKC in enhancing overall resource efficiency and achieving superior load balancing across cloud-edge environments.

Both predictive uncertainty estimation and visual explanation are crucial elements in helping humans understand the artificial intelligence (AI) decision-making process and in building trustworthy AI. However, there has been comparatively limited investigation into the intersection of these two domains in multimodal scenarios. In this article, we propose a universal explanation framework to evaluate counterfactual samples of predictive uncertainty in multimodal models. Inspired by multimodal representation learning, our framework leverages a shared latent space of multimodal variational autoencoders (MVAEs) to generate counterfactual explanations (CEs) of predictive uncertainty, enabling us to identify the input features contributing to high predictive uncertainty. To further evaluate the quality of counterfactual samples, we propose a Bayesian local linear approximation (BLLA) method. This method models the overall linear space as an inverse chi-square distribution while representing feature importance and the error term as normal distributions. By doing so, it captures the uncertainty and feature importance of each modality. Through a comprehensive suite of experiments conducted on multimodal classification and regression tasks, we demonstrate that our framework successfully generates accurate CEs of predictive uncertainty, establishes the consistency of feature importance, and comprehensively facilitates users' comprehension of multimodal model behavior.

Saliency maps have proven to be a highly efficacious approach for explicating the decisions of convolutional neural networks (CNNs). However, extant methodologies predominantly rely on gradients, which constrain their ability to explicate complex models. Furthermore, such approaches are not fully adept at leveraging negative gradient information to improve interpretive veracity. In this study, we present a novel concept, termed positive and negative excitation (PANE), which enables the direct extraction of PANE for each layer, thus enabling complete layer-by-layer information utilization sans gradients. To organize these excitations into final saliency maps, we introduce a double-chain backpropagation procedure. A comprehensive experimental evaluation, encompassing both binary classification and multiclassification tasks, was conducted to gauge the effectiveness of our proposed method. Encouragingly, the results evince that our approach offers a significant improvement over the state-of-the-art methods in terms of salient pixel removal, minor pixel removal, and inconspicuous adversarial perturbation generation guidance. In addition, we verify the correlation between PANEs.

Various correlations hidden in crowdsourcing annotation tasks bring opportunities to further improve the accuracy of label aggregation. However, these relationships are usually extremely difficult to be modeled. Most existing methods can merely make use of one or two correlations. In this paper, we propose a novel graph neural network model, namely LAGNN, which models five different correlations in crowdsourced annotation tasks by utilizing deep graph neural networks with convolution operations and derives a high label aggregation performance. Utilizing the group of high quality workers through labeling similarity, LAGNN can efficiently revise the preference among workers. Moreover, by injecting a little ground truth in its training stage, the label aggregation performance of LAGNN can be further significantly improved. We evaluate LAGNN on a large number of simulated datasets generated through varying six degrees of freedom and on eight real-world crowdsourcing datasets in both supervised and unsupervised (agnostic) modes. Experiments on data leakage is also contained. Experimental results consistently show that the proposed LAGNN significantly outperforms six state-of-the-art models in terms of label aggregation accuracy.

Abstract Currently, the accurate prediction of social relationships can effectively reduce the decision‐making burden of users in various service platforms. However, in the big data environment, the users' data information used for the relationship prediction is highly fragmented distribution, so it is a non‐trivial challenge to integrate the users' sequence data information from different platforms while preventing sensitive information leakage. To this end, based on the microservice environment, we devise a cross‐platform social relationship prediction approach (CPSRP) to address the above problems. Briefly, the improved Simhash method aggregates similar users into the common bucket. Then the embedding technique converts the users' sparse data information into the low‐dimensional dense continuous feature vectors; the redefined Gated Recurrent Unit (r‐GRU) network and the Multilayer Perceptron (MLP) network are employed to extract the overall temporal sequence features of users. The relationship prediction is finally executed according to the users' sequential features. Extensive experiments are conducted on Epinions, and the experimental results further prove the benefits of our proposal in terms of relationship prediction while protecting users' sensitive information.

Collaborative filtering (CF) is dedicated to learning the representations of users and items based on interactive data. Regrettably, the lack of fine-grained modeling of interactive motivation makes the model less interpretable. A feasible solution is to combine the disentangling idea with the graph neural network (GNN) and capture different types of interaction relationships by using a message propagation mechanism on the graph of user–item interaction. However, this process typically relies on the disentangling of users’ hidden intents, ignoring the significance of item features to user engagement. This fact leads to the inadequate interpretability of existing models. To make up for the deficiency, this article proposes a new feature-aware disentangled GNN (FDGNN) for the recommendation. By learning the relationship between user behavior and important features of items, the model aims to achieve better recommendation performance and model interpretability. In the end, we first realize the feature partition based on mutual information and then design an attention-based graph disentangling model to realize the fine-grained disentangling of user intents. In addition, to further ensure the independence of the disentangled intents, we augment the model with disagreement regularization. Through multilayer embedding propagation, FDGNN can display a capture CF effect in feature semantics. The interpretability and efficiency of our proposed approach are demonstrated by numerous pertinent experiments.

Recent years have seen rapid progress in network representation learning, which removes the need for burdensome feature engineering and facilitates downstream network-based tasks. In reality, networks often exhibit heterogeneity, which means there may exist multiple types of nodes and interactions. Heterogeneous networks raise new challenges to representation learning, as the awareness of node and edge types is required. In this article, we study a basic building block of general heterogeneous networks, the heterogeneous networks with two types of nodes. Many problems can be solved by decomposing general heterogeneous networks into multiple bipartite ones. Recently, to overcome the demerits of non-metric measures used in the embedding space, metric learning-based approaches have been leveraged to tackle heterogeneous network representation learning. These approaches first generate triplets of samples, in which an anchor node, a positive counterpart, and a negative one co-exist, and then try to pull closer positive samples and push away negative ones. However, when dealing with heterogeneous networks, even the simplest two-typed ones, triplets cannot simultaneously involve both positive and negative samples from different parts of networks. To address this incompatibility of triplet-based metric learning, in this article, we propose a novel quintuple-based method for learning node representations in bipartite heterogeneous networks. Specifically, we generate quintuples that contain positive and negative samples from two different parts of networks. And we formulate two learning objectives that accommodate quintuple-based learning samples, a proximity-based loss that models the relations in quintuples by sigmoid probabilities and an angular loss that more robustly maintains similarity structures. In addition, we also parameterize feature learning by using one-dimensional convolution operators around nodes’ neighborhoods. Compared with eight methods, extensive experiments on two downstream tasks manifest the effectiveness of our approach.

In digital twin systems for freeways, it is essential to track individual vehicles. When sensing devices cannot fully cover an entire road, it is necessary to accurately predict the travel time of individual vehicles. Therefore, this paper proposes a dual-state traffic factor state network (DS-TFSN), which combines macro traffic states and micro vehicle travel states. Based on the DS-TFSN, a digital twin framework is proposed for freeways. This framework can realize long-distance freeway supervision and vehicle tracking by predicting the travel time of specific vehicles in unsupervised road sections to ascertain their driving process. As the core of digital twin frameworks of freeways, the freeway section travel time prediction model based on the DS-TFSN considers the interactions among macro factors, micro factors, and environmental factors. The model divides the macro traffic state and micro vehicle travel state, and adds them as inputs to the LSTM model. A new vehicle-specific deep learning method is proposed to improve the prediction accuracy in terms of the freeway section travel time. The results show that, for freeways, more accurate prediction results are achieved during both normal hours and holidays. The MAPE of the prediction results using the dual-state traffic factor state network decreases by 6.2%, at most, and the proportion of vehicles with a prediction error of less than 1 second per kilometer increases by 54%, at most.

Exploring user-item interaction cues is crucial for the performance of recommender systems. Explicit investigation of interaction cues is made possible by using graph-based models, where each user-item relationship is described by an edge, and the introduction of user-user social network. While existing graph-based recommendation methods use only a single-value edge to define the relationship between a pair of user and item, which limits the ability to represent complex user-item interactions. Furthermore, some social recommendation methods overlook the heterogeneous user behavior patterns in social and interaction relationships, resulting in the suboptimal performance of existing systems. In this paper, we propose a novel Social-aware Multi-dimensional Edge Feature-based Graph Representation Learning method, called SMEF. It represents all users and items as a graph and deep learns a multi-dimensional edge feature to explicitly describe the task-specific relationships of each user-item pair. Specifically, the proposed SMEF focuses on two distinct user behavior patterns toward social friends and interactive items, which explore the underlying heterogeneous relationship cues within them. This way, the learned multi-dimensional edge features encode user information from both social and interaction aspects. The proposed SMEF is a plug-and-play module that can be combined with different recommendation frameworks and Graph Neural Networks (GNNs) backbones to generate high quality user representations. The experimental results achieved on three publicly accessible datasets show that our SMEF-based method outperforms strong baselines.

Deep neural network is very vulnerable to adversarial examples, which add subtle perturbations on the original image that are difficult for human to perceive, but can make network produce wrong classification results. The current advanced adversarial attack methods can achieve satisfactory results under the white-box setting, but when attacking the black-box model, especially for the defense models, they show poor transferability. It can mainly improve the transferability of adversarial attacks under black-box settings from two perspectives of gradient optimization and image transformation. We propose a new image transformation method, which is different from treating each pixel equally in previous works. We consider using the size of gradient value to reflect the importance of pixels, assigning different scaling factors to each gradient unit, and conducting heuristic random transformation on the images input in each iteration to achieve data enhancement, obtain more stable update direction and escape from local optimal values. Extensive experiments on ImageNet Dataset show that the proposed method has better performance than the existing methods. In addition, our method can also be combined with other attack methods to further improve the transferability of adversarial attacks. Besides, our approach also has excellent performance on the defense models.

In image, video and even real physics domains, adversarial examples can mislead deep models to produce wrong predictions. Transfer-based attacks against black-box models are more in line with realistic scenarios, but adversarial examples made on surrogate model have a low success rate when transferred to the target model due to overfitting the source model. We study the Stochastic Weight Averaging strategy in the domain generalization process and propose a Stochastic Perturbation Averaging method (SPA). Specifically, we add stochastic perturbations to the examples during the gradient descent attack, and we design a Central Amplification method (CAM) to enhance this random variation, then SPA stabilizes the iteration direction by computing the gradient average of the perturbed examples to find a relatively flat local minimum of the loss function. SPA is an efficient and general strategy which can significantly improve the transferability of the gradient-based attack methods. For instance, the average attack success rate of the adversarial examples produced based on four single models against seven pre-trained models reached 90.10%, which is the best result so far. Code is available at https://github.con yangrongbo/SPA.

In recent years, the field of unsupervised graph representation learning has witnessed the emergence of graph contrastive learning (GCL) as a highly successful approach. GCL excels in learning graph representations by effectively bringing positive sample pairs into proximity while simultaneously pushing negative sample pairs apart in the representation space, without any manual labels. Graph-structured data to be learned inherently exhibits a critical hierarchical structure, which is crucial for organizing and managing graphs. Leveraging this attribute enhances the accuracy of graph representation outcomes. However, current GCL methods tend to overlook the hierarchical structure, which can result in sampling bias during node selection. Nodes of the same semantics can potentially be sampled as negative pairs. To overcome these limitations, we present a novel framework, Hierarchical Graph Contrastive Learning via Debiasing Noise Samples with Adaptive Repelling Ratio (HGClear). Our framework enables the simultaneous learning of node representations and the graph hierarchy in an end-to-end manner. During the process of method design, we discovered that the accuracy of node category prediction significantly affects representation results. To remove the bias caused by noise samples, we introduced a module to handle boundary nodes (i.e., noise samples) that are vulnerable to mislabeling. Specifically, we introduce a hierarchy detection module that captures both coarse-grained views and category attributes of nodes. Leveraging these results, we can identify boundary nodes and establish varying repelling ratios based on category labels, replacing the conventional temperature coefficient in the contrastive loss. Simultaneously incorporating an intra-view node contrast module not only eliminates the bias resulting from noise samples but also enhances the uniqueness of node representations. Numerous experiments on node classification datasets show that HGClear produces encouraging results and outperforms some state-of-the-art methods.

Federated Learning has been widely used due to its ability to train models while ensuring data privacy and security. However, the presence of non-i.i.d. (independent and identically distributed) data among different participating entities leads to significant performance disparities in Federated Learning. In addition, as the essence of Federated Learning involves model training on local devices without centralized data storage, it poses certain challenges in personalized tasks. Traditional centralized machine learning methods can perform deep learning and personalized model training on centralized data, while distributed data in Federated Learning may not provide enough information to support personalized requirements. In this paper, we propose a personalized Federated Learning method called FedFC, which adopts parameter decoupling to address the data domain shift issue in Federated Learning. We experimentally validate our proposed solutions on the OrganCMNIST and COVID-19 datasets. The experimental results show that compared to existing methods, FedFC achieves average accuracy improvements of 4.93% and 6.31% on the two datasets, respectively. Additionally, by employing partial gradient uploading, we successfully reduce the communication overhead by 22.67% and 26.91% for each dataset.

In the current era of information overload, service recommendations have emerged as a valuable tool for enhancing the user experience. Among them, social recommendation models have shown promising results by incorporating social relationships to improve representation learning. However, most of these models lack fine-grained modeling of social user behavior, leading to a unified representation of users and a loss of expressiveness in user representations. To address this issue, we propose DisenHGCF, a new social recommendation approach based on disentangled hypergraph collaborative filtering, to disentangle the representations of users and items at the granularity of social users’ intents. This approach aims to provide a more nuanced understanding of the user, leading to more accurate and personalized recommendations. To be specific, DisenHGCF leverages hypergraphs to represent the complex relationships among users, friends, and items. By using the hypergraph disentangling module based on attention, it is able to disentangle the user’s intents and generate users representations in cooperating their intents for recommendation tasks. Additionally, a contrastive learning task based on intent-weight perturbation is designed to enhance representational learning. The experimental results obtained from the BeiBei and Beidian datasets demonstrate the superiority of our proposed approach in comparison to previous baseline methods, as evidenced by higher Recall and NDCG scores.

Transfer-based attacks against black-box neural network models have received increasing attention because they are more realistic scenarios, but how to produce highly transferable adversarial examples on the surrogate model becomes critical. In this work, we find that if the attack direction of the original example is controlled from the beginning, the produced adversarial examples will be more transferable. Specifically, we propose the Output Direction Controller (ODC) to initialize the example direction so that the example starts off with a deviation from the true direction or toward the target direction. ODC is a simple and extensible component that can be combined with various transfer-based attack methods and significantly improve the transferability of the adversarial examples. On the ImageNet dataset, we optimize the baseline method by ODC to improve the success rate of untargeted attacks by an average of 11.79% and targeted attacks by an average of 3.38%. Code is available at https://github.com/yangrongbo/ODC.

Abstract Nowadays, there is a significant increase in information, resulting in information overload. Recommendation systems have been widely adopted, and they can help users find information relevant to their interests. However, a malicious attacker can infer users' private information via recommendations. To solve problems of data sparseness, enormous high‐dimensional data, the cold start problem and privacy protection in an intelligent recommender system, this study proposes a privacy‐preserving collaborative filtering recommendation method with clustering and locality‐sensitive hashing. First, we cluster users according to their characteristic information to obtain sub‐rating matrices. We use the latent factor model to predict and fill in the missing ratings in those matrices. Second, we combine the sub‐rating matrices into a complete rating matrix, subsequently, we obtained the neighbors of the target user by analyzing the similarity of the users. We use a locality‐sensitive hashing algorithm to reduce the dimensionality of the user rating data and build an index that could quickly obtain the neighbors of the target user. Finally, we predict the target user's ratings and provide recommendations to the target user. Through experiments, our study shows that our method can deal with the problems of data sparseness and cold start problems well and the accuracy of the intelligent recommendation system has been improved. In addition, we use hash techniques to search for the neighbors, which effectively protects the privacy of the user.

With the rapid development of Internet of Things (IoT), big data analytics (BDA) has gradually stepped into the spotlight of IoT research. However, existing research mostly focused on improving the efficiency of data mining, the wishes and profits of the participants in BDA are largely ignored. In real systems, it is always supposed that the IoT data owners accept the big data analytics in IoT (BDA-IoT), the theoretical modeling of BDA-IoT is needed to be constructed urgently. In this paper, aiming to provide theoretical modeling of the practical application of BDA-IoT, we prove the feasibility of participants voluntarily participating in BDA-IoT for the first time. Subsequently, a non-cooperative game theory model with incentive and payment mechanisms is constructed, and the multi-parties interaction process in BDA-IoT is simulated. The overall benefits of all participants are then discussed, the rationality and feasibility of our study are proved. Simulation results show the feasibility of our model in improving the benefits of all stakeholders in a non-cooperative game, and the best choice for all participants is to accept BDA-IoT.

The application of recommendation systems online services is becoming more and more extensive. However, most existing recommendation algorithms centralize multi-party information into a central processor, which may lead to the risk of privacy leakage. And many enterprises or institutions still have the problem that data cannot be shared. Federated learning has been introduced into recommendation algorithms for privacy- aware distributed learning. A typical federated learning is that each client uses local data to train a shared model, the server uses their gradient information to form a global model, and then each client updates. In this paper, we propose a federated deep recommendation algorithm called FedHe-mlp that applies a federated deep learning for data privacy protection, and combines heterogeneous information network (HIN) and matrix factorization technique for better prediction performance. First, each client obtains heterogeneous information through meta- paths, then we combine matrix factorization and heterogeneous information to mine the latent features and heterogeneous features of each client. Finally, We propose a deep neural network that considers features from multiple views. Extensive experiments on three public datasets demonstrate that FedHe- mlp can provide excellent convergence speed, recommendation accuracy, and communication efficiency while preserving data privacy.

Deep neural networks are vulnerable to adversarial examples generated by black-box attacks with tiny perturbations. However, black-box based transferability attacks usually add perturbation to the whole image. It is easy for the defender to detect the adversarial examples. Inspired by attention modules, we propose a method named Gradient-mask and Attention-whey (GM&AW) to reduce redundant noise and maintain attack effect of them in this work. During Gradient-mask iterations, we only choose the regions with greater gradient and update them along the direction of gradient. Then, we utilize Attention-whey optimization combining attention mechanism and query to further decrease noise. Extensive experiments on Imagenet demonstrate that our approach can achieve enormous decrease on noise in$\ell_{2}$norm and keep attack success rate for black models.

Deep Neural Network (DNN) models are usually trained with tremendous data and computation resources. Thus, DNN models are now regarded as important assets, however facing a great risk of being stolen and illegal distribution. In recent years, watermark is introduced to protect the ownership of DNN models. The watermark can be extracted in a relatively simple way to declare the ownership of the model. However, watermark is vulnerable to be attacked. In this work, we propose a watermark defense method for DNN model based on pruning. Inspired by the pruning methods, we design a fused channel-wise pruning strategy which selects important filters for watermarks embedding. Specifically, we introduce a novel method to enhance the watermark robustness by selecting important filters as the watermark carrier based on multiple pruning methods, including network slimming, efficient filter and entropy. We conduct experiments on the VGG-19 model with the CIFAR-10 dataset. The experimental results show that this method is robust against fine-tuning attack, pruning attack and overwriting attack. In addition, our method does not significantly change the distribution of model weights so that the watermark is hard to be detected.

Attributed graph clustering is a fundamental yet challenging task for graph analysis, which is of significance in the real world. Recently, graph contrastive learning have demonstrated excellent performance on clustering task. However, we observe that existing methods have the following three defects: a) an end-to-end graph clustering framework based on contrastive clustering has not yet been implemented, b) they require a pretext task to obtain initial labels, c) and most methods lack a specific loss function for clustering processes. To address above issues, we innovatively propose Attributed Graph Contrastive Clustering (AGCC), which is a novel self-supervised attributed graph clustering framework. AGCC can directly produce a clustering assignment matrix without calculating target distribution matrix. Specifically, in our method, an effective module based on contrastive graph representation learning is being leveraged to extract node embeddings by node features and topological structure. In the phase of data augmentation, the module can adaptively mask node features and drops edges. Meanwhile, another reliable contrastive clustering module generates clustering results through contrasting the representation of each cluster. Comprehensive empirical analysis, compared with various types of classic and deep methods, demonstrates the efficacy and the scalability of the proposed method.

Federated learning enables local nodes to train a global model together by uploading only training updates to the parameter server without exchanging private data. However, as the complexity of the federated learning task increases, the communication volume of the training process becomes extremely large, hence the huge communication traffic becomes a serious bottleneck in current federated learning application. Existing methods reduce communication overhead from two aspects, the number of communications and the traffic per communication. But these methods usually lead to more consumption of computing resources or a decrease in model accuracy. To handle these problems, this paper proposes a data fitting based weight compression algorithm, FWC, which includes four sequential stages: sparsification, polynomial fitting, encoding, reconstruction and two mechanism: warm-up and accumulation. In particular, the warm-up mechanism can well address the problem of slow convergence in early training period. Experimental results on models with different scales show that FWC is able to provide more than 600x traffic compression at the cost of only millisecond-level computational time cost and less than 1% accuracy loss.

Real-world networks often show heterogeneity. A frequently encountered type is the bipartite heterogeneous structure, in which two types of nodes and three types of edges exist. Recently, much attention has been devoted to representation learning in these networks. One of the essential differences between heterogeneous and homogeneous learning is that the former structure requires methods to possess awareness to node and edge types. Most existing methods, including metapath-based, proximity-based and graph neural network-based, adopt inner product or vector norms to evaluate the similarities in embedding space. However, these measures either violates the triangle inequality, or show severe sensitivity to scaling transformation. The limitations often hinder the applicability to real-world problems. In view of this, in this paper, we propose a novel angle-based method for bipartite heterogeneous network representation. Specifically, we first construct training sets by generating quintuples, which contain both positive and negative samples from two different parts of networks. Then we analyze the quintuple-based problem from a geometry perspective, and transform the comparisons between preferred and non-preferred samples to the comparisons of angles. In addition, we utilize convolution modules to extract node features. A hinge loss, as the final objective, is proposed to relax the angular constraint for learning. Extensive experiments for two typical tasks show the efficacy of the proposed method, comparing with eight competitive methods.

User authentication remains a challenging issue, despite the existence of a large number of proposed solutions, such as traditional text-based, graphical-based, biometrics-based, Web-based, and hardware-based schemes. For example, some of these schemes are not suitable for deployment in an Internet of Things (IoT) setting, partly due to the hardware and/or software constraints of IoT devices. The increasing popularity and pervasiveness of IoT equipment in a broad range of settings reinforces the importance of ensuring the security and privacy of IoT devices. Therefore, in this article, we conduct a comprehensive literature review and an empirical study to gain an in-depth understanding of the different authentication schemes as well as their vulnerabilities and deficits against various types of cyberattacks when applied in IoT-based systems. Based on the identified limitations, we recommend several mitigation strategies and discuss the practical implications of our findings.

Industrial Internet of Things (IIoT) facilitate private data collecting via (a broad range of) sensors, and the analysis of such data can inform decision making at different levels. Federated learning (FL) can be used to analyze the collected data, in privacy-preserving manner by transmitting model updates instead of private data in IIoT networks. The FL framework is, however, vulnerable because model updates are easily tampered with by malicious agents. Motivated by this observation, we propose a novel chameleon hash scheme with a changeable trapdoor (CHCT) for secure FL in IIoT settings. Our scheme imposes various constraints on the use of trapdoor. We give a rigorous security analysis on our CHCT scheme. We also instantiate the CHCT scheme as a redactable medical blockchain (RMB). The experimental evaluations demonstrate the practical utility of CHCT in terms of accuracy and efficiency.

While encryption can prevent unauthorized access to a secret message , it does not provide undetectability of covert communications over the public network. Implementing a highly latent data exchange, especially with low eavesdropping/discovery probability, is challenging for practical scenarios, such as social and political movements in authoritarian regimes , military operations, and privacy preservation . Moreover, the current literature suffers from a low embedding capacity and monolingual applicability, limiting the amount of hiding secret data within short text messages using state-of-the-art algorithms, e.g., linguistic-based, structural-based, or coverless-based solutions. In this paper, we present a systematic covert communication technique called CovertSYS that enables a multilingual secure end-to-end conversation via messaging or social network platforms. The CovertSYS functions by encrypting a confidential message using a multi-factor authentication scheme and converting the encoded binary data into hidden Unicode symbols to be transmitted under cover of short text messages. We then conduct extensive experiments to confirm the security and validity of the proposed technique against state-of-the-art approaches. Our experimental results show that the CovertSYS provides a superior mean performance of 91.53% by improving the criteria scores: embedding capacity rate of 100%, imperceptibility rate of 76.4%, and distortion robustness rate of 98.2%. Finally, we discuss the practical implications of the proposed technique compared to the existing text steganography methods.

With the advancements of the Health 2.0 technology, large-scale healthcare services are available online. Recommender systems for healthcare services have emerged for decision assistance. Most existing collaborative recommendation algorithms only mine global interactions while failing to capture the local different information of users or items. Besides, privacy concern is another significant problem to be considered in recommender systems for healthcare services. In this article, a privacy-aware factorization-based hybrid method is proposed for healthcare service recommendations. For better modeling of user preferences and service features, multiview embeddings of users and healthcare services are learned. Besides, we address the privacy problem by integrating local differential privacy and locality-sensitive hashing techniques into the recommendation model for privacy-aware neighbor searching. The final prediction is made based on a stochastic gradient descent learning-based hybrid collaborative model. Experiments demonstrate the effectiveness of the proposed method in both recommendation performance and privacy concerns.

The deep learning approach to detecting malicious software (malware) is promising but has yet to tackle the problem of dataset shift, namely that the joint distribution of examples and their labels associated with the test set is different from that of the training set. This problem causes the degradation of deep learning models without users’ notice. In order to alleviate the problem, one approach is to let a classifier not only predict the label on a given example but also present its uncertainty (or confidence) on the predicted label, whereby a defender can decide whether to use the predicted label or not. While intuitive and clearly important, the capabilities and limitations of this approach have not been well understood. In this paper, we conduct an empirical study to evaluate the quality of predictive uncertainties of malware detectors. Specifically, we re-design and build 24 Android malware detectors (by transforming four off-the-shelf detectors with six calibration methods) and quantify their uncertainties with nine metrics, including three metrics dealing with data imbalance. Our main findings are: (i) predictive uncertainty indeed helps achieve reliable malware detection in the presence of dataset shift, but cannot cope with adversarial evasion attacks; (ii) approximate Bayesian methods are promising to calibrate and generalize malware detectors to deal with dataset shift, but cannot cope with adversarial evasion attacks; (iii) adversarial evasion attacks can render calibration methods useless, and it is an open problem to quantify the uncertainty associated with the predicted labels of adversarial examples (i.e., it is not effective to use predictive uncertainty to detect adversarial examples).

Keyphrase annotation task aims to retrieve the most representative phrases that express the essential gist of documents. In reality, some phrases that best summarize documents are often absent from the original text, which motivates researchers to develop generation methods, being able to create phrases. Existing generation approaches usually adopt the encoder-decoder framework for sequence generation. However, the widely-used recurrent neural network might fail to capture long-range dependencies among items. In addition, intuitively, as keyphrases are likely to correlate with topical words, some methods propose to introduce topic models into keyphrase generation. But they hardly leverage the global information of topics. In view of this, we employ the Transformer architecture with the pre-trained BERT model as the encoder-decoder framework for keyphrase generation. BERT and Transformer are demonstrated to be effective for many text mining tasks. But they have not been extensively studied for keyphrase generation. Furthermore, we propose a topic attention mechanism to utilize the corpus-level topic information globally for keyphrase generation. Specifically, we propose BertTKG, a keyphrase generation method that uses a contextualized neural topic model for corpus-level topic representation learning, and then enhances the document representations learned by pre-trained language model for better keyphrase decoding. Extensive experiments conducted on three public datasets manifest the superiority of BertTKG.

Graph neural networks have shown superior performance in representing graph-structured data. However, directly applying most of existing methods to heterogeneous graph learning tasks may be ill-suited or infeasible, since multiple different types of nodes, edges and attributes challenge the methods originally developed for homogeneous networks. Given this situation, heterogeneous graph embedding recently has gained much research attention. Existing heterogeneous graph embedding methods often adopt random walk-based sampling techniques to model both heterogeneity and rich attribute information. These methods strongly rely on customized sample generation strategies and usually consume a lot of memory. To avoid these restrictions, we present a novel model based on edge-aware neighborhood convolution which is designed to extract features from the neighborhood of different types of nodes. The first and second-order proximity-based objectives are formulated for learning node representations. On the task of link prediction and recommendation, experiments on five heterogeneous graph datasets show the effectiveness of our model.

As a representative classification model, decision tree has been extensively applied in data mining. It generates a series of if-then rules based on the homogeneity of class distribution. In a society where data spreads everywhere for knowledge discovery, the privacy of the data respondents is likely to be leaked and abused. Based on this concern, we propose an overview of the rapidly evolving research results focusing on privacy preserving decision tree induction. The research results are summarised according to the characteristics of related privacy preservation techniques, which include data perturbation, cryptography, and data anonymisation. In addition, we demonstrate the comparison between the merits and demerits of these methods considering the specific property of decision tree induction. At last, we conclude the future trend of privacy preserving techniques.

With the rapid development of intelligent manufacturing and Industrial Internet of Things, many industrial control systems have high requirements for the security of the system itself. Failures of industrial control equipment will cause abnormal operation of industrial control equipment and waste of resources. It is very meaningful to detect and identify potential equipment abnormalities and failures in time and implement effective fault tolerance strategies. In the Industrial Internet of Things environment, the instructions and parameters of industrial control equipment often change due to changes in actual requirements. However, it is impractical to customize the learning method for each parameter value. Aiming at the problem, this paper proposes a fault diagnosis model based on ensemble learning and proposes a method of updating voting weights based on dynamic programming to assist decision-making. This method is based on Bagging strategy and combined with dynamic programming voting weight adjustment method to complete fault type prediction. Finally, this paper uses different loads as dynamic conditions; the diagnostic capability of the Bagging-based fault diagnosis integrated model in a dynamically changing industrial control system environment is verified by experiments. The fault diagnosis model of industrial control equipment based on ensemble learning effectively improves the adaptive ability of the model and makes the fault diagnosis framework truly intelligent. The voting weight adjustment method based on dynamic programming further improves the reliability of voting.

Nowadays, large number of cloud-based techniques have been used in industrial control systems (ICS), which also brings many security threats. The emergence of security-aware industrial control has paved the way of security-aware scheduling in cloud-based industrial applications. Actually, most cloud-based industrial applications are time sensitive, which need real-time processing. Edge cloud computing paradigm extends the computing ability of traditional cloud model with low-latency local resources. Thus, heterogeneous clouds that consist of both centralized resources and edge resources may be a promising resource model to provide both scalable and low-latency resources for cloud-based industrial applications. In view of these challenges, in this article, we propose a security-aware dynamic scheduling method for real-time resource allocation in ICS. First, a three-level security model is designed for both tasks and cloud resources in ICS, and a two-tier heterogeneous cloud architecture is introduced. Accordingly, a security-aware scheduling method based on distributed particle swarm optimization is presented for resource allocation with security concerns. To deal with the dynamics of edge resources and the mobility of mobile industrial applications, a dynamic scheduling mechanism based on dynamic workflow model is proposed for real-time optimization. Experimental results validate that the scheduling control policy proposed in this article can achieve a good balance between scheduling performance and security performance.

Volunteer computing uses computers volunteered by the general public to do distributed scientific computing. Volunteer computing is being used in high-energy physics, molecular biology, medicine, astrophysics, climate study, and other areas. These projects have attained unprecedented computing power. However, with the development of information technology, the traditional defense system cannot deal with the unknown security problems of volunteer computing . At the same time, Cyber Mimic Defense (CMD) can defend the unknown attack behavior through its three characteristics: dynamic, heterogeneous, and redundant. As an important part of the CMD, the dynamic scheduling algorithm realizes the dynamic change of the service centralized executor, which can enusre the security and reliability of CMD of volunteer computing . Aiming at the problems of passive scheduling and large scheduling granularity existing in the existing scheduling algorithms, this article first proposes a scheduling algorithm based on time threshold and task threshold and realizes the dynamic randomness of mimic defense from two different dimensions; finally, combining time threshold and random threshold, a dynamic scheduling algorithm based on multi-level queue is proposed. The experiment shows that the dynamic scheduling algorithm based on multi-level queue can take both security and reliability into account, has better dynamic heterogeneous redundancy characteristics, and can effectively prevent the transformation rule of heterogeneous executors from being mastered by attackers.

The rapid development of pay-as-you-go cloud services motivates the increasing number of cloud resource demands. However, the volatile demands bring new challenges for current techniques to minimize the cost of cloud capacity planning and VM provisioning while satisfying the customer demands. The service vendors will incur enormous revenue loss within the long-term inappropriate planning, especially when the demands fluctuate abruptly and frequently. In this paper, we cast the cloud capacity planning as a classification problem and propose an integrated framework, which effectively predicts the abrupt changing demands, to reduce the cost of cloud resource provisioning. In this framework, we first apply Piecewise Linear Representation to segment the time series of cloud resource demands for labeling the changing trend of each period. Second, Weighted SVM is leveraged to fit the statistical information and the label of each period and predict the changing trend of the following period. Finally, an incremental learning strategy is utilized to ensure the low cost of updating the model using the upcoming requests. We evaluate our framework on the IBM Smart Cloud Enterprise (SCE) trace data and the experimental results show the effectiveness of our proposed framework.

Summary With the explosive growth of cloud services, how to design effective recommendation models has become more and more important. Temporal information has been proved to be an important factor affecting recommendation performance. Actually, both of user behaviors and QoS performance of services are time‐sensitive, especially in dynamic cloud environment. However, most existing collaborative recommendation methods seldom consider temporal influence to QoS performance. Furthermore, with the ever‐increasing number of security threats in clouds, privacy preservation becomes an important problem to be addressed in recommender systems. Based on these observations, in this paper, we propose a temporal‐aware and sparsity‐tolerant hybrid collaborative recommendation method with privacy preservation, where tensor factorization‐based CF model is integrated into neighborhood‐based CF model to achieve improved recommendation performance. Specifically, temporal influence is considered into recommendation model by distinguishing short‐term QoS metrics from long‐term QoS metrics. A privacy‐aware time aggregation mechanism is adopted to preserve the sensitive time information of users. Moreover, to deal with the sparsity problem, a sparsity‐oriented data QoS rating prediction mechanism based on tensor factorization technique is applied to predict the missing or unrated QoS rating data. Finally, predictions are made based on both stable and temporal nearest neighbors. Experiments are conducted to demonstrate the prediction performance of our proposal.

Current trends of mobile technology have seen a tremendous growth in its application in smart healthcare. This has resulted in the adoption and implementation of mobile health (m-health) systems by providing health assistance to the aging population. Despite its advantageous benefits, its computational complexities cannot be overlooked. M-health devices are portable processing tiny equipment with limited computational capabilities thereby making them complex for the implementation of public key cryptosystems. In spite of this, an Offline-Online signature scheme called the MHCOOS has been proposed to solve the difficulties in the computational ability. The scheme enjoys the following benefits by splitting the signing part into both offline and online phases. The offline phase performs heavy computations when a message is absent, whereas lighter computations are performed at the online stage when a message is present. Secondly, the online computations are extremely fast due to the already computed offline signature value and lighter pairings involved. Our performance analysis demonstrates how the proposed scheme outperforms other schemes. Finally, the hardness of the scheme is proven under the Bilinear Diffie–Hellman (BDH) and Computational Diffie–Hellman (CDH) problem in the random oracle model.

Malware remains a big threat to cyber security, calling for machine learning based malware detection. While promising, such detectors are known to be vulnerable to evasion attacks. Ensemble learning typically facilitates countermeasures, while attackers can leverage this technique to improve attack effectiveness as well. This motivates us to investigate which kind of robustness the ensemble defense or effectiveness the ensemble attack can achieve, particularly when they combat with each other. We thus propose a new attack approach, named mixture of attacks, by rendering attackers capable of multiple generative methods and multiple manipulation sets, to perturb a malware example without ruining its malicious functionality. This naturally leads to a new instantiation of adversarial training, which is further geared to enhancing the ensemble of deep neural networks. We evaluate defenses using Android malware detectors against 26 different attacks upon two practical datasets. Experimental results show that the new adversarial training significantly enhances the robustness of deep neural networks against a wide range of attacks, ensemble methods promote the robustness when base classifiers are robust enough, and yet ensemble attacks can evade the enhanced malware detectors effectively, even notably downgrading the VirusTotal service.

With the development of location-based social networks (LBSNs), Point-of-Interest (POI) recommendation has attracted lots of attention. Most of the existing studies focus on recommending POIs to users based on their recent check-ins. However, the recent check-ins may contain some daily check-ins that users are not really interested in. If a model treats the recent check-ins equally, it is non-trivial to capture the actual preference of users. To address the issue of mining the actual preferences of users in the POI recommendation, we propose an attention-based deep learning POI recommendation framework (ADPR), which consists of a latent representation method and an attention-based deep convolutional neural network. To learn the embedding of users and POIs, we propose a latent representation method, which incorporates the geographical influence and the categories of POIs to capture the relationships between POIs better. Further, we propose an attention-based deep convolutional neural network, which employs the attention mechanism to filter the important information in the recent check-ins, to recommend POIs to users based on the latent representations of users and the recent check-ins. We conduct experiments on a real-world LBSN dataset to evaluate our framework, and the experimental results show the effectiveness of our framework.

Abstract Various studies have demonstrated that convolutional neural networks (CNNs) can be directly applied to different levels of text embedding, such as character‐, word‐, or document‐levels. However, the effectiveness of different embeddings is limited in the reported result and there is a lack of clear guidance on some aspects of their use, including choosing the proper level of embedding and switching word semantics from one domain to another when appropriate. In this paper, we propose a new architecture of CNN based on multiple representations for text classification, by constructing multiple planes so that more information can be dumped into the networks, such as different parts of text obtained through named entity recognizer or part‐of‐speech tagging tools, different levels of text embedding, or contextual sentences. Various large‐scale, domain‐specific datasets are used to validate the proposed architecture. Tasks analyzed include ontology document classification, biomedical event categorization, and sentiment analysis, showing that multi‐representational CNNs, which learns to focus attention to specific representations of text, can obtain further gains in performance over state‐of‐the‐art deep neural network models.

With the ceaseless usage of web and other online services, it has turned out that copying, sharing, and transmitting digital media over the Internet are amazingly simple. Since the text is one of the main available data sources and most widely used digital media on the Internet, the significant part of websites, books, articles, daily papers, and so on is just the plain text. Therefore, copyrights protection of plain texts is still a remaining issue that must be improved in order to provide proof of ownership and obtain the desired accuracy. During the last decade, digital watermarking and steganography techniques have been used as alternatives to prevent tampering, distortion, and media forgery and also to protect both copyright and authentication. This paper presents a comparative analysis of information hiding techniques, especially on those ones which are focused on modifying the structure and content of digital texts. Herein, various text watermarking and text steganography techniques characteristics are highlighted along with their applications. In addition, various types of attacks are described and their effects are analyzed in order to highlight the advantages and weaknesses of current techniques. Finally, some guidelines and directions are suggested for future works.

In this paper, a call admission control algorithm based on inter-network cooperation is proposed via a Stackelberg game framework for cognitive heterogeneous networks. The call admission control problem is subject to the variable bandwidth rate traffic, network service selection, feasible subchannel allocation, and call blocking probability. The call admission control algorithm is based on spectrum price at primary heterogeneous networks and subchannel allocation price and network selection at cognitive heterogeneous networks. In order to determine the call blocking probability, a probability upper bound of exceeding the maximum admission number for secondary mobile terminals (MTs) is analyzed based on M/M/∞ model. Then, the subchannel allocation price and network selection are designed via the dual decomposition method, and the vacant spectrum price is determined with Bertrand game theory. Finally, a call admission control algorithm is proposed. Simulation results demonstrate that the proposed algorithm not only improves quality of service at each secondary MT, but also reduces the call blocking probability for cognitive heterogeneous networks.

Bag-of-words (BOW) is now the most popular way to model text in statistical machine learning approaches in sentiment analysis. However, the performance of BOW sometimes remains limited due to some fundamental deficiencies in handling the polarity shift problem. We propose a model called dual sentiment analysis (DSA), to address this problem for sentiment classification. We first propose a novel data expansion technique by creating a sentiment-reversed review for each training and test review. On this basis, we propose a dual training algorithm to make use of original and reversed training reviews in pairs for learning a sentiment classifier, and a dual prediction algorithm to classify the test reviews by considering two sides of one review. We also extend the DSA framework from polarity (positive-negative) classification to 3-class (positive-negative-neutral) classification, by taking the neutral reviews into consideration. Finally, we develop a corpus-based method to construct a pseudo-antonym dictionary, which removes DSA's dependency on an external antonym dictionary for review reversion. We conduct a wide range of experiments including two tasks, nine datasets, two antonym dictionaries, three classification algorithms, and two types of features. The results demonstrate the effectiveness of DSA in supervised sentiment classification.

In this paper, a kind of traffic prediction and congestion control policy based on RFNN (rough-fuzzy neural network) is proposed for ATM (asynchronous transfer mode). Congestion control is one of the key problems in high-speed networks, such as ATM. Conventional traffic prediction method for congestion control using BPN (back propagation neural network) has suffered from long convergence time and dissatisfying precision and it is not effective. The fuzzy neural network scheme presented in this paper can solve these limitations satisfactorily for its good capability of processing inaccurate information and learning. Finally, the performance of the scheme based on BPN is compared with the scheme based on RFNN using simulations. The results show that the RFNN scheme is effective.

Based on the basic theory of cellular automata (CA), using genetic algorithm to find the rules of CA, one-dimensional extended non-uniform CA is constructed to generate high quality random bit numbers. A stringent test of randomness is applied to the bit numbers. The results show that our extended CA generator outperforms other generators such as shift register generator, congruential generator and lagged Fibonacci generator. This generator can meet the requirement of the security for cryptographic modules and provide a new method to generate pseudo-random numbers in cryptography. An example of image encryption is also provided for illustration.