Sana'a University Journal of Applied Sciences and Technology

Smart agriculture and industrial Internet-of-Things
(IoT) systems increasingly rely on distributed sensors, edge gate-ways, and networked control services, making privacy-preserving and communication-efficient anomaly detection essential. Centralized learning can expose operational telemetry, increase communication overhead, and degrade under heterogeneous non-independent and identically distributed (non-IID) client conditions. This paper presents SEAF-FL-MLP, a leakage safe federated edge-AI framework for binary anomaly detection across five heterogeneous IoT clients: AGRI, SWAT, WADI, WUSTL-IIoT, and TON-IoT. The framework integrates split-first preprocessing, train-only imputation/encoding/scaling/feature selection, client- specific processing, a compact multilayer perceptron (MLP), FedAvg aggregation, validation-only threshold selection, and test-only final evaluation. In the audited V9 output, the balanced validation-selected threshold achieved Accuracy=96.32%, Precision=96.67%, Recall = 93.74%, F1-score = 95.18%, ROC-AUC = 98.86%, and PR-AUC = 98.03% on the global held-out test set, with a model size of 0.0372 MB and an estimated total communication cost of 4.4682 MB across 12 federated rounds and five clients. Client-wise analysis revealed substantial heterogeneity: WUSTL-IIoT, TON-IoT, and SWAT achieved strong F1-scores, whereas AGRI and WADI remained challenging stress cases. WUSTL shortcut-reduced sensitivity showed that the WUSTL client remained strong after removing high-risk shortcut-sensitive features, although the global pipeline was partially sensitive to replacing WUSTL with the shortcut-reduced version. FedProx did not materially improve performance under the tested configuration, and the ablation results did not support claiming mutual-information feature selection as a performance-improving component in this run. The results position SEAF-FL-MLP as a compact, communication-efficient, and reproducible federated edge-AI baseline while emphasizing the necessity of transparent client-level, fairness, and shortcut-robustness analysis.

We present a quantum formulation of the Fundamental Speed Theory (QFST) built from the classical FST Lagrangian, where the dimensionless speed field ν^μ plays a dual role: it modifies gravity classically in galaxies, while its quantum excitations behave as ultra-light dark matter on cosmological  scales.
On galactic scales, the classical FST equation fits rotation-curve data without dark matter (171 SPARC galaxies, χ² _ν = 0.170). Linearization yields a screening scale μ = 1 /λ_screen, implying a particle mass m = ¯h/(c λ_screen) = 3.88 × 10⁻²⁴ eV, with no free parameters. Quantum corrections to the galactic equation are suppressed by (ℓ_Pl/λ_screen)² ≈ 10⁻¹⁰³, ensuring classical-quantum separation.
Dark energy arises from the logarithmic potential V( ˜ν) = −V0 ln1 + ˜ν²/ν² ₀ . We obtain a positive vacuum energy density ρΛ = [c⁴/(16πGL² ₀)] · U( ˜νnow) and determine ˜νnow = 8.21 × 10⁻¹² by matching the observed _ρΛ. At the background level, QFST reproduces the standard expansion history as tested by Cosmic Chronometer H(z) data. Full discrimination from ΛCDM requires a Boltzmann-code implementation at the perturbation level.
In a spatially flat FLRW spacetime, we derive the reduced kinetic scalar including Hν terms, construct the stressenergy tensor T^(ν) _μν , and obtain closed-form expressions for ρ_ν and p_ν. The equation of state w(z) = p_ν/ρ_ν is determined by the coupled background system (Friedmann equations plus the ν(t) equation of motion). Detailed derivations and dimensional checks are provided in the appendices.