Automated respiratory disease classification from auscultation sounds holds transformative potential for early clinical screening, yet existing approaches remain constrained by the quadratic complexity of Transformer-based sequence encoders, the limited expressiveness of conventional multi-layer perceptron classifiers, and the persistent challenge of scarce annotated medical audio data. This paper presents MambaResp-KAN, a novel architecture that unifies Bidirectional Mamba state space models, Kolmogorov–Arnold Network classifiers with learnable B-spline activation functions, multi-modal gated cross-attention fusion of WavLM, BEATs, and handcrafted spectral features, and class-conditional denoising diffusion probabilistic model augmentation into a single end-to-end framework for explainable respiratory sound analysis. The Bidirectional Mamba encoder achieves linear-time sequence modeling through input-dependent selective state space discretization, processing forward and reverses temporal streams with gated aggregation to capture both causal and anti-causal dependencies in respiratory waveforms. The Kolmogorov–Arnold Network classifier replaces fixed-activation neurons with learnable univariate B-spline functions on each network edge, directly grounded in the Kolmogorov–Arnold representation theorem, yielding a classifier that is both more parameter-efficient and intrinsically interpretable than standard multi-layer perceptrons. A gated cross-modal attention mechanism fuses embeddings from the self-supervised WavLM and BEATs audio encoders with handcrafted MFCC and spectral features, while a class-conditional denoising diffusion probabilistic model synthesizes high-fidelity respiratory audio to alleviate class imbalance. Integrated Gradients attribution and KAN concept bottleneck analysis provide clinician-interpretable explanations of model decisions. Evaluated on two benchmark datasets, Asthma Detection V2 (five classes, 1,211 samples) and KAUH (four classes, 940 samples), MambaResp-KAN achieves classification accuracies of 99.6% and 99.4%, respectively, surpassing the prior state-of-the-art E-RespiNet by 0.7 and 0.6 percentage points while using 62% fewer parameters and reducing inference latency by 56.3%. Cross-dataset evaluation yields an average accuracy of 84.0% with a generalization gap of 15.8%, compared to 23.3% for E-RespiNet, confirming improved transferability across clinical institutions.

IoT networks face persistent security challenges due to limited compute, heterogeneous hardware, and weak threat-detection coverage. Classical machine-learning methods struggle with high-dimensional traffic and novel attack patterns. This paper proposes a hybrid framework combining Fractional Generalized Laguerre (FrGL) moment-based feature extraction with a Residual Network augmented by Squeeze-and-Excitation attention (ResNet-SE). FrGL moments yield compact, noise-resistant descriptors via simple recurrence relations, while ResNet-SE mitigates degradation in deep networks through identity shortcuts and adaptively recalibrates channels to highlight attack-relevant features. On the Bot-IoT and Leopard Mobile IoT benchmarks the method reaches 99.78 % accuracy and 99.37 % F1, exceeding KNN (84.7 %), MLR (87.5 %) and a baseline CNN (99.3 %); cross-dataset tests on UNSW-NB15 and IoT-Bot give 96.34 % and 97.12 % accuracy. The framework additionally delivers per-sample inference latency on server- and edge-class hardware (3.9 ms on an NVIDIA V100 and 27.4 ms on a Raspberry Pi 4B with a Coral USB accelerator), an energy cost of 0.42 J per inference on the edge platform, a sensitivity analysis over learning rate, batch size, fractional order λ and reduction ratio r, and an adversarial-robustness evaluation under FGSM and PGD attacks, supporting real-time deployment on resource-constrained IoT gateways.