HIFN-Mamba: hybrid information flow networks with selective state-space sequence modeling for intrusion detection in the Internet of Medical Things
Year: 2026
Type: article
Abstract: The exponential growth of the Internet of Medical Things (IoMT) — a multi-billion-device infrastructure of clinically critical sensors, monitors, and connected medical equipment — has dramatically expanded the attack surface for ransomware, lateral-movement, and protocol-specific exploits against healthcare networks. Existing deep-learning intrusion detection systems (IDSs) for this domain suffer from three coupled limitations: fixed information-bottleneck hyperparameters that cannot adapt to heterogeneous protocol distributions, dependence on costly post-hoc explainability methods (SHAP, LIME) that violate sub-millisecond clinical latency budgets, and quadratic-cost attention mechanisms that are infeasible for edge deployment. This paper proposes HIFN-Mamba, a hybrid architecture that couples a two-layer Hybrid Information Flow Network (HIFN) trunk with a stack of selective state-space (Mamba/S6) sequence blocks. Each HIFN layer learns four information-theoretic regulators end-to-end: a static gate vector providing intrinsic sample-invariant feature interpretability, a magnitude scaling factor preventing latent collapse, an information retention coefficient replacing the fixed beta of the Variational Information Bottleneck, and an entropy budget enforcing a hinge constraint on latent differential entropy. We prove that HIFN strictly generalises VIB and that all four regulators remain bounded for any finite training dynamics. Empirical evaluation across the IoMT-focused CICIoMT2024 dataset and two cross-domain benchmarks (UNSW-NB15, CICIDS-2017) in six experimental settings yields 99.999% accuracy with zero false alarms on CICIoMT2024 binary, 99.995% accuracy and 0.9999 macro-F1 on CICIoMT2024 19-class, and sub-millisecond inference latency (0.0129–0.0170 ms/sample) on all six experiments. The static gates provide global feature rankings at zero inference cost, eliminating the exponential complexity of post-hoc SHAP. To our knowledge, HIFN-Mamba is the first IoMT IDS to combine learned information-theoretic regularisation with linear-time selective state-space modelling and to report verifiable sub-millisecond throughput on a nineteen-class IoMT benchmark.
Source: Scientific Reports
Institutions: Sana'a University, Ajloun National University, Thamar University
Language: English
Subfield: Computer Networks and Communications
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $2,690
OpenAlex ID: W7215753268
FWCI:
Cites:
Cited by:
Related to: