Internet of Things (IoT) botnets are networks of infected smart devices controlled by attackers and posing a serious cybersecurity challenge. Developing detection approaches that maintain high accuracy while protecting privacy presents considerable challenges, particularly in large and heterogeneous IoT networks. This paper empirically compares three modelling approaches on Bot-IoT and N-BaIoT in binary and multiclass settings: handcrafted machine learning with random forest (RF), centralised deep learning (CDL) with DNN/LSTM/BiLSTM, and federated deep learning (FDL) with the same architectures. Model hyperparameters are selected via randomised search on stratified subsets and then fixed for final training. Results show near-perfect performance for all approaches in binary detection: on Bot-IoT, CDL-DNN attains perfect accuracy, and RF is virtually perfect (only four benign-to-attack false positives), while FDL models are similarly strong with only small false-positive and false-negative counts. On N-BaIoT, RF and CDL (especially LSTM) are near-perfect, and FDL is very close to CDL. For multiclass detection, CDL-DNN leads on Bot-IoT, RF remains near perfect with minimal cross-class confusion, and FDL trails slightly; on N-BaIoT, FDL-BiLSTM and RF are essentially perfect, with CDL-LSTM close behind. Overall, the findings validate RF as a competitive classical approach, show where centralised representation learning adds value, and demonstrate that federated training preserves most of the centralised accuracy while avoiding raw data centralization (data locality) for scalable deployment.

Nonlinear loads are widely applied, making the generation mechanism of grid harmonics increasingly intricate. However, high-precision monitoring devices suffer from high deployment costs and limited coverage. This poses a major challenge to directly acquiring harmonic voltages at some nodes. To solve this problem, this paper proposes a harmonic state estimation method based on a Dual-Stream Adaptive Fusion Generative Adversarial Network (DSAF-GAN), with an innovative design in its generator architecture. A dual-path generator is developed to extract multi-scale features through heterogeneous network branches collaboratively. The ResNet-GRU path integrates convolutional residual modules with Bidirectional Gated Recurrent Units (Bi-GRUs). It effectively captures local spatial patterns and temporal dynamic characteristics of time-series data. The multi-layer perceptron (MLP) path focuses on mining global nonlinear correlations, thereby enhancing the overall feature-expressing capability. An adaptive weight fusion module (Attention Weight Net) fuses the outputs of the two paths. It dynamically allocates contribution weights, improving the model’s flexibility and generalization performance. Experimental results show that the proposed DSAF-GAN can accurately reconstruct the harmonic voltage component content rate of missing nodes.

The fast-changing Cloud-IoT environment requires effective and dynamic intrusion detection systems (IDS). In order to overcome the weaknesses of the current solutions in managing high-dimensional, heterogeneous network traffic, this paper presents an improved structure with an Enhanced Feature Pyramid Network (EFPN) and a Quantum-Enhanced Child Drawing Development Optimizer (Q-CDDO). The EFPN has been scaled to tabular network data based on multi scale features extraction as pseudo-images. The quantum rotation gates are used in the Q-CDDO for better adjusting the hyperparameters in the EFPN structure with high-dimensional spaces. The model is validated on standard benchmark functions including the CIC-IDS-2017 and Bot-IoT datasets. Th results show an accuracy of 96.3% and 94.6% on the CIC-IDS-2017 and Bot-IoT datasets, respectively. The synergistic effect of both constituents, and the discriminative power of the model are verified by the studies of the ablation and the visualization, respectively. This paper contributes to the progress of IDS and proves the potential of quantum-inspired metaheuristics in cybersecurity.

While the energy levels of nodes in these networks vary, there is still a need for further advancements to fully optimize the performance of heterogeneous Wireless Sensor Networks (WSNs). Efficient data routing plays a vital role in improving the IoT-enabled WSNs overall performance. Existing routing protocols face difficulties in addressing issues like frequent node movement, optimizing energy efficiency, scalability, and adapting to changing network conditions. These challenges need to be addressed to ensure effective routing in IoT-based heterogeneous WSNs. Thus, a new approach for optimal routing in IoT-based heterogeneous WSNs that utilizes the STGOA is proposed. In the Network Modeling stage, various components of the IoT-enabled heterogeneous WSN, including sensor nodes, servers, the internet, BS, and the network topology, are defined. The interactions between these elements are established to define the communication structure and flow within the network. The Optimal Routing stage begins with the clustering of sensor nodes via the Fuzzy C Means (FCM) method. After clustering, the STGOA approach is employed to choose the optimal paths while considering critical constraints like energy consumption, trust, distance, security, and delay. The proposed STGOA method achieved maximum energy rate of 0.368 J at node 100 and 0.375 J at node 200 as compared to the other methods like GWO, SHO, GOA, STO, BOA, NBO, JFO, EEMCM and GMPSO.

Polymer-based composites are receiving more and more attention in the field of radiation shielding due to their good recyclability and processability; however, the poor compatibility between the organic polymer matrices and inorganic fillers usually appears as a bottleneck to their full shielding effectiveness. Although compatibilizers are widely utilized to suppress the agglomeration of inorganic fillers, their impact on the overall performance of composites is less concerning. In this study, we systematically investigate the effects of three common compatibilizers, i.e. 3-aminopropyltriethoxysilane (KH550), 3-methacryloxypropyltrimethoxysilane (AM), and maleic anhydride–grafted polypropylene (PP-g-MAH), on the properties of WO3/PP composites, including rheological performance, thermal properties, crystallization behaviors, mechanical properties, and radiation shielding effectiveness. It was found that in addition to dispersing WO3 particles, KH550 and AM contribute to the formation of a heterogeneous network structure in the PP matrix, and especially for KH550, it can effectively promote the structural stability, thermal stability, crystallinity, and toughness of WO3/PP composites. More importantly, WO3/PP-S-K exhibits the best shielding performance, with linear attenuation coefficients of 3.71 cm−1 and 0.279 cm−1 under 241Am (59.6 keV) and 60Co (1332 keV) irradiation, respectively. Compared with the pristine WO3/PP composite, the half-value layer is markedly reduced to 0.187 cm and 2.48 cm, representing decreases of 25.5% and 60.6%. These findings underscore the enhancements in shielding performance and structural integrity achieved through modification with KH550, thereby providing valuable insights for the future development of environmentally friendly radiation shielding materials.

ncFN: a comprehensive non-coding RNA function annotation framework based on a global and heterogeneous biomolecular network.

Context-aware heterogeneous graph neural network for multi-level description and invasiveness prediction in renal cell carcinoma.

Cancer is a complex and heterogeneous disease, where even patients with the same cancer type may respond differently to treatment regimens. Predicting the therapeutic effects of drugs on cancer based on cancer characteristics is a critical aspect of precision oncology. Currently, most anticancer drug response(CDR) prediction methods rely on extracting features from the cell line-drug bipartite composition. However, these methods often fail to adequately capture the features of both drugs and cell lines, ignoring the homogeneous features of cell lines and drugs and their correlation with deep heterogeneous features. To address these challenges, we propose a novel prediction framework that leverages a heterogeneous and homogeneous hybrid graph neural network named H$^{3}$CDR. H$^{3}$CDR learns the similarity features of cancer cell lines and drugs by fusing their multi-omics data. Additionally, a multi-branch network is employed to extract features from both cell lines and drugs, enabling the identification of potential features. Extensive experiments on the GDSC and CCLE databases demonstrate the superiority of our model. Evaluated by five-fold cross-validation, H$^{3}$CDR achieves an area under the ROC curve (AUC) of 0.8772 and an area under the precision-recall curve (AUPRC) of 0.8819 on the GDSC dataset.

Dual-level dynamic heterogeneous graph network for video question answering.

A unified gradient regularization method for heterogeneous graph neural networks.

Reducing size bias in epidemic network modelling.

MT-RoadNet: A heterogeneous network with local–global joint enhancement for road surface and centerline extraction

Spatial heterogeneity of burial-related fracture networks controls compartmentalized fluid flow in volcano-sedimentary sequences

Metapath-based feature aggregated heterogeneous graph neural network for adverse drug reactions prediction

Drug repositioning is an efficient drug discovery method for identifying associations between present drugs and new diseases, offering considerable development time and cost savings. Although existing methods have been widely applied, they fail to fully capture the complex semantics between drugs and diseases, and are deficient in terms of model interpretability. In this paper, we propose a novel method using a hierarchical attention mechanism aggregating meta-path information for drug-disease association prediction (MPHAM), aiming at effectively integrating heterogeneous information from various sources to enhance prediction accuracy and model interpretability. First, considering the wide range of biological interactions between drugs and diseases, we construct a heterogeneous information network (HIN) to utilize data on drugs, proteins, and diseases. Then, we introduce a meta-path-based feature fusion strategy designed to effectively capture the complex semantics between nodes in the network. By defining meta-paths of multiple lengths and types, information about different relationship types is systematically integrated to generate high-quality node feature representations. Furthermore, the feature fusion strategy incorporates a multi-layer attention mechanism that dynamically assigns weights to the contributions of various meta-paths in the feature aggregation process, significantly improving the model's capacity to capture important semantic information. Experimental results demonstrate that MPHAM can effectively predict drug-disease association by integrating complex meta-path information, and the prediction accuracy is better than five state-of-the-art methods. The case studies of three classical drugs further demonstrate the more accurate predictive performance of MPHAM in drug-candidate disease association prediction.

Adaptive reinforcement learning algorithm for real-time energy optimization in building digital twins with heterogeneous IoT sensor networks

Multi-semantic fusion of heterogeneous graph neural network

A heterogeneous Hopfield neural network with discrete memristor: modeling, dynamics, and application in medical image encryption

Adaptive computation control and model aggregation for parallel federated learning in heterogeneous edge networks

HetMS-AMRGNN: Heterogeneous multi-scale graph neural network for antimicrobial drug recommendation in electronic health records