WiFi-based indoor localization is essential for asset tracking, healthcare monitoring, and smart buildings. However, existing systems face challenges such as RSS variability, environmental noise, and difficulty in detecting floor and building levels, compounded by limited labeled data and the high costs of collecting received signal strength (RSS). This paper introduces quantum stochastic contrastive learning (QSCL), a novel framework grounded in rigorous theoretical foundations. We present four theorems and one lemma that establish bounded probabilistic augmentation, diversity of the strong view, the suitability of the symmetric contrastive objective under heterogeneous augmentation channels, and expected similarity stability under zero-mean perturbations, supported by formal proofs. Leveraging these foundations, QSCL uses quantum computing (QC) to generate strong data augmentations via stochastic perturbations, thereby enhancing data diversity, while classical weak augmentations provide subtle variations for robust feature learning. We propose a spatio-temporal encoder (STE) that integrates convolutional layers with channel and spatial attention modules (CBAM-style) to capture spatial and temporal dependencies in sequential data. Furthermore, a symmetric cross-view contrastive loss is introduced to capture forward and reverse relationships between augmented views, ensuring robust representations. Comprehensive evaluations on the UJIIndoorLoc and UTSIndoorLoc datasets validate QSCL, demonstrating superior performance with limited labeled data and resilience to quantum and measurement noise. The proposed framework significantly improves localization accuracy, floor and building detection, and generalizability in challenging indoor environments.

Modern supply chains must balance high-speed logistics with environmental impact and security constraints, prompting a surge of interest in AI-enabled Internet of Things (AIoT) solutions for global commerce. However, conventional supply chain optimization models often overlook crucial sustainability goals and cyber vulnerabilities, leaving systems susceptible to both ecological harm and malicious attacks.To tackle these challenges simultaneously, this work integrates a quantum-inspired reinforcement learning framework that unifies carbon footprint reduction, inventory management, and cryptographic-like security measures. We design a quantum-inspired reinforcement learning framework that couples a controllable spin-chain analogy with real-time AIoT signals and optimizes a multi-objective reward unifying fidelity, security, and carbon costs. The approach learns robust policies with stabilized training via value-based and ensemble updates, supported by window-normalized reward components to ensure commensurate scaling. In simulation, the method exhibits smooth convergence, strong late-episode performance, and graceful degradation under representative noise channels, outperforming standard learned and model-based references, highlighting its robust handling of real-time sustainability and risk demands. These findings reinforce the potential for quantum-inspired AIoT frameworks to drive secure, eco-conscious supply chain operations at scale, laying the groundwork for globally connected infrastructures that responsibly meet both consumer and environmental needs.

Federated Learning (FL) enables decentralized model training but faces critical challenges in jointly optimizing privacy, accuracy, and communication efficiency, essential for resource-constrained wireless IoT deployments. We introduce ADP-QFed, an Adaptive Differentially Private Quantized Federated Learning framework that addresses these challenges through layer-wise adaptive noise injection and dual-bit deterministic quantization. By computing layer-specific sensitivity and importance scores, ADP-QFed dynamically calibrates privacy noise to minimize accuracy loss while ensuring rigorous (ε, δ)-differential privacy guarantees. The framework employs n-bit quantization for local computation and m-bit quantization for transmission, reducing communication overhead by up to 75%. Experiments on MNIST, FMNIST, and CIFAR-10 achieve test accuracies of 99.41%, 91.06%, and 82.94%, respectively, outperforming existing privacy-preserving FL methods by an average of 3.5%. These results are obtained while maintaining a privacy budget under ε = 2.25, representing a 40% reduction compared to state-of-the-art methods at similar accuracy levels. ADP-QFed advances practical privacy-preserving FL for edge sensing in low-altitude IoT systems by simultaneously optimizing privacy guarantees, model utility, and energy efficiency in wireless environments.

Accurate indoor localization and navigation enable real-time, ubiquitous, location-based services. Over the past decade, data-driven approaches for inertial odometry have shown the potential to enhance indoor positioning accuracy. However, low-cost inertial measurement units (IMUs), commonly used in smartphones and IoT devices, are prone to significant noise, leading to drift and degraded performance in navigation algorithms. This article presents a novel, lightweight, and real-time end-to-end framework, DeepILS, designed to process raw inertial data for precise pedestrian localization in indoor environments. DeepILS utilizes a residual network enhanced with channel-wise and spatial attention mechanisms, enabling accurate velocity and position estimation across diverse motion dynamics. The framework’s effectiveness is validated using four benchmarks and two newly introduced datasets in real-time edge scenarios. These datasets were collected across diverse indoor environments at the KAIST campus and Incheon National Airport, using multiple hardware platforms, including the KAIST IoT positioning module and Android smartphones. Experimental results, including tests on unseen data and comprehensive ablation studies, demonstrate that DeepILS improves localization accuracy by 70% compared to state-of-the-art methods while effectively mitigating sensor noise and enhancing robustness in real-world environments. Specifically, DeepILS exhibits excellent edge performance on IoT devices, making it highly suitable for real-time applications.

Deep learning-based inertial navigation remains a formidable challenge due to the intricate temporal dynamics of human motion and the stringent computational constraints of edge devices. This study introduces NanoMST, a highly efficient multi-scale transformer architecture designed for precise pedestrian inertial motion tracking with minimal computational overhead. The proposed model integrates a hierarchical multi-scale embedding strategy with a scale-adaptive attention mechanism, effectively capturing motion patterns across diverse temporal resolutions while optimizing efficiency through hardware-aware quantization. With 298K parameters and 7.59M floating-point operations, NanoMST achieves performance comparable to substantially larger models while maintaining an exceptionally low computational burden. Extensive evaluations on benchmark datasets, including OxIOD, RoNIN, and RIDI, yield average trajectory errors of 2.68m on RoNIN, 1.64m on RIDI, and 1.80m on OxIOD. The quantized 8-bit implementation reduces the model size from 1.23MB to 0.41MB while retaining 94% of the original model’s accuracy. Profiling on edge devices confirms real-time feasibility, with inference latencies ranging from 0.18 milliseconds to 0.96 milliseconds across various smartphone generations and an average throughput exceeding 6,000 samples per second, surpassing contemporary architectures such as IMUNet and CTIN. This study illustrates an efficient engineering approach for deep learning-based inertial tracking, demonstrating that high-precision sequential motion estimation can be achieved with a minimal computational footprint. The efficiency and real-time capability of NanoMST make it particularly suitable for deployment in resource-constrained environments, including mobile, wearable, and Internet of Things (IoT) applications.