The rapid proliferation of Internet of Things (IoT) devices has led to a significant rise in security issues, as traditional centralized intrusion detection systems face difficulties in handling issues such as privacy concerns, communication bottlenecks, and heterogeneous data. Federated Learning (FL) is a new paradigm for collaborative learning that can be used to train intrusion detection systems without compromising user privacy. However, it is challenged by critical issues such as handling highly non-independent and identically distributed (non-IID) data, a common problem in heterogeneous IoT networks such as healthcare networks, financial networks, and industrial networks. In this paper, a novel CNN-LSTM architecture is proposed that is equipped with Explainable AI (XAI) to handle extreme cases of heterogeneous data in Federated Learning-based intrusion detection. Using the CIC-IDS2017 dataset and Dirichlet-based partitioning ([Formula: see text] our optimized CNN-LSTM model achieves a 97.36% centralized F1-score. We demonstrate that while architectural design provides highly stable robustness under moderate data heterogeneity ([Formula: see text]), extreme non-IID conditions ([Formula: see text]) trigger severe weight washing, dropping FedAvg performance to 71.53%. Through a comprehensive hyperparameter sweep, we prove that applying FedProx with a strong proximal penalty ([Formula: see text]) successfully mitigates this client drift, recovering the F1-score to 78.27%. Using SHAP, LIME, t-SNE, and PCA, we reveal that our model learns universal protocol-level features (Init_Win_bytes_forward, ACK Flag Count, and Fwd Packet Length Min) that remain invariant across heterogeneous networks. Finally, we demonstrate a highly stable detection is achieved for network-layer attacks, whereas application-layer intrusions (Web Attacks) suffer severe degradation, proving the fundamental limitations of flow-based features for payload-driven attacks.
Jallal-Eddine Moussaoui, Mehdi Kmiti, Yassine Maleh et al.· Scientific Reports· 0 citations
Visual foundation models are commonly adapted under the assumption that the appearance of incoming data may change while the semantic meaning of the prediction task remains fixed. In long-lived visual systems, however, taxonomies, policies, and concept definitions can themselves evolve, causing the same visual evidence to require a different interpretation. We study this setting as evolving semantic concept shift and introduce SemReWrite, a framework for selectively updating obsolete visual--semantic mappings while preserving knowledge that remains valid. SemReWrite represents changes between old and revised semantic specifications, combines semantic discrepancy with sparse revised supervision to localize affected visual regions, and uses an input-dependent low-rank rewriting mechanism together with structured semantic memory, preservation, and obsolete-decision suppression. We further introduce EvoShift-Bench, spanning ImageNet, iNaturalist, CUB-200-2011, and DomainNet, with semantic transitions including class split, merge, boundary revision, insertion, partial redefinition, recurrence, and mixed semantic--appearance shift. To explicitly evaluate selective semantic revision, we introduce Rewrite Accuracy (RA) and Preservation Accuracy (PA) for affected and unaffected regions, respectively, Obsolete Retention (OR) for measuring residual outdated semantic associations, and the Selective Revision Score (SRS), which jointly summarizes rewriting and preservation performance. Experiments show that SemReWrite achieves a stronger balance between learning revised semantics and retaining unaffected knowledge than prompt replacement, conventional fine-tuning, parameter-efficient adaptation, and continual-learning strategies.
Ismail Lamaakal, Chaymae Yahyati, Yassine Maleh et al.· 0 citations
Modern blockchain systems impose distinct cryptographic requirements, including pairing-based signatures for Ethereum, multi-scalar multiplication (MSM) for zero-knowledge proofs (ZKPs), and post-quantum migration, which existing hardware implementations largely overlook. In this paper, we present a PRISMA 2020-compliant systematic review that analyzes Field-Programmable Gate Array (FPGA) and Application-Specific Integrated Circuit (ASIC) implementations of Elliptic Curve Cryptography (ECC) and Post-Quantum Cryptography (PQC) from 2018 to 2025. We evaluate these across four blockchain node categories: full nodes, light clients, ZKP provers, and validators. From 688 initial records, rigorous screening yielded 56 FPGA implementations, 16 ASIC designs, and 3 hybrid architectures. Our findings reveal critical gaps between academic research and deployment reality: there is minimal hardware support for the pairing-friendly curves essential for Layer-2 scaling and zkRollups, as well as a lack of hybrid architectures preparing for the imminent post-quantum transition. Performance analysis shows that FPGAs achieve moderate throughput with high energy efficiency (hundreds to thousands of operations per joule), while ASICs reach significantly higher performance at substantially greater costs. PQC schemes consistently exhibit notable area overhead and memory consumption. Crucially, side-channel security remains underaddressed, with no implementations achieving both high performance and validated resistance. To bridge these gaps, we extract four security-aware design patterns, establish a platform selection framework integrating throughput-cost-security trade-offs, and identify five research priorities: standardized benchmarking, hybrid architectures, validated security-performance co-design, energy-proportional implementations, and protocol-hardware co-optimization.