A wearable system enabling acute stress monitoring and closed-loop mitigation through transcutaneous median nerve stimulation.

Conventional occupational radiation exposure monitoring relies on cumulative dose data from personal dosimeters without providing information on when, where, or under what conditions exposure occurs. This lack of context limits analysis of causal factors, evaluation of protective behaviors, and the effectiveness of safety education. This study aimed to develop and clinically implement an integrated information system for occupational radiation exposure by combining dose data, spatiotemporal movement records, and angiography-related radiation information. We also assessed its utility and potential for improving radiation safety management. The system was implemented for 1 mo in a clinical angiography suite. It integrated (1) personal digital dosimeters recording dose and time, (2) Bluetooth Low Energy beacons tracking healthcare workers' positions and movements, and (3) Radiation Dose Structured Reports providing exposure details. Data were synchronized to reconstruct when, where, and under what conditions exposure occurred. The system identified high-risk positions near x-ray tubes (Beacon IDs 1-3), where exposure was greatest. Avoidance behaviors were also detected, such as movement to low-risk areas (e.g., Beacon ID 8) before irradiation. We successfully developed, implemented, and evaluated the system, demonstrating its utility for improving radiation safety management. The insights gained support targeted interventions and the refinement of safety protocols, with potential for broader use in diverse radiation-controlled settings.

Rapid localisation of trapped victims after urban disasters is essential but challenging because Bluetooth Low Energy (BLE) beacons are intermittent, radio propagation is obstructed by rubble, UAVs are energy-constrained, and real-world multi-UAV training is impractical in high-risk search-and-rescue (SAR) environments. This study formulates post-disaster victim localisation as a cooperative Dec-POMDP and adapts a model-aided federated multi-agent reinforcement learning framework based on FedQMIX. The proposed pipeline combines a lightweight LoS/NLoS surrogate channel model, PSO-based victim-position estimation, return-to-base and map-feasibility safety checks, an SAR-aligned shaped reward, and a leakage-free centralised training state based on estimated rather than ground-truth victim locations. Each UAV trains locally inside a learned digital-twin simulator and periodically shares only QMIX network parameters, avoiding the exchange of raw trajectories or RSSI logs. The framework is evaluated on two synthetic post-earthquake urban maps representing a compact return-to-base scenario and a larger reach-to-destination scenario. Across five independent seeds per method and map, Model-Aided FedQMIX achieves the highest and most stable victim-localisation performance, with the clearest advantage observed in the larger long-horizon scenario. Additional diagnostic tests examine reward-weight sensitivity, RF channel-shift robustness, BLE/smartphone hardware heterogeneity, non-IID client-data variation, and partial-client FedAvg under missing client updates. The results indicate that combining model-aided localisation cues, decentralised value factorisation, SAR-aligned objective design, and federated parameter sharing can improve the robustness of UAV-based victim-localisation policies. The framework also clarifies deployment considerations for federated SAR coordination, including communication payload, privacy boundaries, heterogeneous client experience, device variability, and intermittent connectivity. This study remains simulation-based, and future validation with real UAVs, BLE devices, and rubble-inspired testbeds is required before operational deployment.

Abstract: Communication systems today are overwhelmingly dependent on centralized internet infrastructure, which introduces vulnerabilities such as network outages, censorship, surveillance, and infrastructural dependency. This research introduces BlueMesh, a novel decentralized messaging architecture that enables users to communicate using Bluetooth Low Energy (BLE) without requiring internet connectivity. Unlike traditional messaging platforms, BlueMesh leverages a hybrid peer-to-peer and mesh-based communication protocol built on top of Bluetooth to enable message propagation beyond direct device range. Existing implementations of Bluetooth messaging are typically limited to point-to-point communication, where two devices connect and exchange messages directly. However, such systems suffer from a severe limitation in scalability and communication range. Research indicates that Bluetooth typically operates within a range of 10–100 meters, depending on environmental conditions . To overcome this constraint, BlueMesh introduces a dynamic relay mechanism, where nearby devices act as intermediate nodes, forwarding encrypted messages across a distributed network. The system is implemented using the MERN stack (MongoDB, Express.js, React, Node.js), but with a critical architectural twist: instead of relying on cloud servers, the backend logic is partially executed on local device nodes and edge services, enabling decentralized processing. Each device maintains a lightweight local database and sync layer, ensuring message persistence and delivery even in disconnected environments. A key innovation in this research is the introduction of a “Proximity-Aware Routing Algorithm”, which intelligently determines message paths based on device density, signal strength, and mobility patterns. Unlike traditional mesh networks, which broadcast blindly, BlueMesh optimizes message delivery to reduce redundancy and energy consumption. Security is addressed through end-to-end encryption, ephemeral identity keys, and local-only data storage, ensuring that no centralized authority can access user data. The system also introduces a store-and-forward mechanism, allowing messages to persist and propagate even when recipients are temporarily offline. This research demonstrates that decentralized, infrastructure-free communication is not only feasible but also scalable when combined with modern web technologies and Bluetooth mesh networking. BlueMesh has potential applications in disaster communication, remote areas, military operations, and censorship-resistant communication systems.

ABSTRACT The Internet Blackout Survival System is a decentralized communication platform designed to enable communication during internet outages, natural disasters, or network failures. The system operates without relying on traditional infrastructure such as cellular networks or centralized servers. Instead, it utilizes peer-to-peer communication technologies like Wi-Fi Direct and Bluetooth Low Energy (BLE) to form a resilient mesh network. The platform allows users to send messages, broadcast emergency alerts (SOS), and share critical information across nearby devices. It incorporates secure communication using encryption techniques such as Elliptic Curve Diffie-Hellman (ECDH) for key exchange and AES encryption for data transmission. The system ensures reliability through multi-layer communication, where BLE acts as a fallback mechanism when Wi-Fi Direct is unavailable. The application is built using modern Android technologies such as Jetpack Compose for UI, MVVM architecture for state management, and Room Database for offline storage. The system also includes intelligent routing, message propagation control, and real-time synchronization across devices. This platform is highly useful in disaster scenarios, remote areas, and situations where internet access is restricted or unavailable, providing a reliable and secure communication alternative. Keywords: Mesh Network, Offline Communication, Wi-Fi Direct, BLE, Decentralized System, Disaster Communication, Peer-to-Peer Network, AES Encryption

Ankle–foot orthoses (AFOs) are widely used in the rehabilitation of patients with neurological or musculoskeletal disorders. However, treatment outcomes may be influenced by incorrect use of the device or by inappropriate orthosis selection. Since many types of AFOs are available, differing in materials, stiffness, and geometry, an objective evaluation tool can support clinical decision-making. This work presents the design, development, and characterization of an instrumented AFO able to quantify relevant gait parameters in an objective way. The proposed device integrates three measurement modalities in a compact wearable structure. Two longitudinal strain gauges estimate ankle plantar- and dorsiflexion angles. Two force-sensitive elements detect foot–ground contact and allow identification of stance and swing phases of the gait cycle. A single inertial measurement unit (IMU) is used to measure lateral shank inclination. The strain-gauge-based angle estimation was validated against a gold-standard motion capture system, achieving a root mean square error of approximately 1.6 degrees and showing higher accuracy than the IMU for plantar/dorsiflexion measurement, while maintaining a simple electronic architecture. The force sensors were validated using a force platform and demonstrated reliable detection of loading and unloading events. Monitoring lateral inclination through the single IMU provides additional information related to balance and potential fall risk. Data are transmitted via Bluetooth Low Energy (BLE) to a custom Python-based application for real-time visualization and recording. Overall, the results validate the electronic instrumentation and demonstrate reliable system performance, indicating that the proposed instrumented AFO represents a promising platform for objective gait assessment and future clinical applications.

In this article, a low-power low-intermediate-frequency (Low-IF) receiver front-end is presented for Bluetooth Low Energy (BLE) body area network (BAN) applications. The receiver employs an input matching network, an inductorless self-biased inverter-based low-noise transconductance amplifier (LNTA), a single-balanced passive mixer, a common-gate transimpedance amplifier (TIA), and a Gm-C complex filter for image suppression. Native MOS devices are adopted to support low-voltage operation and reduce static power consumption. The interstage on-chip coupling capacitor between the RF front-end and the TIA is removed by aligning the DC operating points of the two stages. The receiver front-end is implemented in a 55 nm standard CMOS process and occupies an active area of 0.081 mm2, excluding bonding pads. Post-layout simulations show that the proposed design achieves 45.2 dB gain, 7.2 dB noise figure, and 28.1 dB image rejection ratio over the 2.4–2.48 GHz band, while consuming 537 μW. The proposed front-end is suitable for low-power BLE BAN sensor nodes.

The current implementation of Remote Patient Monitoring (RPM) still faces crucial challenges related to the accuracy and integrity of medical data. Many healthcare IoT devices rely on generic sensors that require rigorous manual calibration and exhibit unstable error rates, failing to meet international clinical standards. This study aims to design and implement an integrated backend architecture that bridges certified commercial medical devices with digital health systems. The main contribution is a six-layer IoT architecture specifically designed to integrate the Omron HEM-7142T1 device to ensure data validity in remote blood pressure monitoring. Following the Design Science Research Methodology (DSRM), the system was developed using Python, the Bleak library for Bluetooth Low Energy (BLE) communication, and FastAPI to provide interoperable REST API services. Functional testing in Postman demonstrated that the system successfully extracts medical data, producing JSON output with an HTTP 200 OK status under single-access conditions. However, load testing using Apache JMeter with 10 virtual users revealed limitations in the hardware’s point-to-point BLE protocol. The /scan endpoint showed stable performance with a 0% error rate and an average response time of 5.04 seconds. In contrast, endpoints /connect-and-read and endpoint /latest-bp-records recorded error rates of 100% and 90%, respectively, with an average response time of 23.29 seconds when accessed simultaneously, due to the Omron device’s locking mechanism. This study concludes that while the six-layer architecture effectively ensures medical data integrity in single-access scenarios, it requires a database caching module in the Logic Tier to overcome parallel access constraints. The implementation provides a foundation for developing secure, standardized professional RPM systems for medical use.

Introduction/Objective: This work presents a fully integrated Integer-N frequency synthesizer targeting multi-standard short-range wireless protocols operating in the 2.4 GHz ISM band, including Bluetooth LE (BLE 5.4), ZigBee, Thread, and Wi-Fi coexistence. The design addresses the need for fast lock time, low residual FM, excellent spectral purity, and low-cost implementation in resource-constrained IoT applications. Methods: The synthesizer is implemented in 0.35 µm CMOS and centers on a resistorless discretetime loop filter (DT-LPF), the core innovation of this work, which eliminates on-chip resistors entirely by replacing them with a delayed charge injection mechanism using switched-capacitor networks. This approach improves stability of the loop, eliminates thermal noise associated with passive resistors, and by avoiding on-chip resistors entirely, reduces sensitivity to resistor mismatch and process variation, while enabling a more compact loop filter topology in standard CMOS processes. Complementing this, the design employs a gm/Id-optimized differential LC-VCO, a dead-zone-free tri-state PFD with 1 ns reset-path delay, and a current-scaled CML multi-modulus 2/3 divider that reduces highfrequency power dissipation. All blocks are co-optimized for 1 MHz reference frequency operation, and the system is fully designed and simulated using industry-standard EDA tools. Results: The PLL achieves a 230 MHz tuning range (2.28–2.51 GHz) with 1 MHz channel resolution, fully covering the 79-channel Bluetooth band, ZigBee/Thread channels (2.405–2.480 GHz), and the entire 2.4 GHz ISM band used by Wi-Fi. It locks in approximately 100 µs, satisfying the <150 µs fast-hopping requirement for BLE 5.4 and enabling low-latency operation in ZigBee/Thread networks. The proposed architecture exhibits –117.5 dBc/Hz phase noise at 1 MHz offset, exceeding ZigBee’s –110 dBc/Hz specification and supporting robust Wi-Fi coexistence, along with 5 kHz RMS residual FM (well below the 10 kHz Bluetooth budget), reference spurs below –67 dBc (within ZigBee’s –60 dBc and Bluetooth’s –55 dBc limits), and 32.6 mW total power consumption from a 3.3 V supply. The Figure of Merit (FoM) is –170.1 dB. Discussion: The DT-LPF enhances integrability, reduces thermal noise, and assures PLL lock stability, while the current-scaled CML divider cuts prescaler power by >30%. Despite using a lowcost 0.35 µm process, the architecture rivals advanced-node PLLs in FoM and protocol compliance. The 1 MHz reference frequency and explicit residual FM validation ensure real-world suitability for Bluetooth and ZigBee, metrics often omitted in recent works. The design demonstrates that architectural innovation can compensate for technological constraints in cost-sensitive IoT. Conclusion: The proposed synthesizer meets all essential performance metrics for ISM-band wireless protocols and demonstrates architectural efficiency and scalability. Its design offers a competitive and low-cost solution for power-sensitive short-range communication systems, particularly in cost-constrained IoT applications where advanced CMOS nodes are economically prohibitive.

Perishable supply chains lose substantial value because product deterioration is rarely measured directly during distribution, while static date labels and temperature records provide only indirect assurance of quality. This review surveys nano-enabled sensor elements that translate headspace chemistry and microbial activity into actionable freshness information, prioritizing studies evaluated on real foods and under realistic cold chain excursions. We synthesize progress across three technology clusters. Optical indicators combine pH responsive dyes, polydiacetylene assemblies, quantum dots, and carbon dots to deliver visible or camera-readable outputs that track amine accumulation and ripening-related metabolites. Gas sensors and electronic noses employ semiconducting metal oxides, graphene, and carbon nanotube networks to quantify volatile organic compounds (VOCs) such as ammonia, trimethylamine, ethylene, and carbon dioxide, with increasing attention to room temperature operation and pattern-based classification. Biosensors extend specificity through antibodies, aptamers, and molecularly imprinted polymers on conductive nanostructures, frequently using electrochemical impedance spectroscopy (EIS) to detect bacteria, toxins, or enzymatic byproducts at low loads. Integration pathways are reviewed from passive color labels to wireless tags using radio frequency identification (RFID), near field communication (NFC), and Bluetooth Low Energy (BLE), enabling package-level data capture and cloud analytics for remaining shelf-life estimation. We also consolidate cross-cutting constraints including calibration drift, humidity cross sensitivity, limited independent validation, added system cost, and safety and regulatory requirements around nanomaterial migration. Taken together, the literature indicates that multi-parameter designs coupled to data models are central to translating laboratory sensitivity into deployable decision support in practice.

ESPiezometer: ESP32-based field tool for installation and validation of piezometric sensors for groundwater level monitoring.

Abstract The Global Positioning System (GPS) uses signals from satellites to specialised devices to detect an entity’s location on Earth’s surface. However, GPS is ineffective for indoor navigation due to high power consumption and signal obstruction by walls. This study proposes a low-cost, energy-efficient alternative using Bluetooth 4.0-enabled Bluetooth Low Energy (BLE) beacons to enable Indoor Positioning and guide users in confined spaces, such as libraries. Tiny, low-energy BLE beacons transmit Advertising Packets at regular intervals. In contrast, Anchor Nodes send and receive signals to construct a Received Signal Strength Indication (RSSI)-based signal-to-distance model, enabling triangulation-based localisation. The system monitors behaviour and environmental interaction through RSSI analysis and user interaction with Beacons. Proven in a library setting, the BLE-based system proved extremely accurate in guiding users to destinations, bridging GPS shortcomings at minimal energy expenditure and cost.

We present MW-FlexNIRS, a wearable, low-cost, LED-based, multi-wavelength near-infrared spectroscopy (NIRS) system designed for continuous monitoring of cerebral oxygenation and metabolic dynamics in neonates. The device extends the original FlexNIRS platform by integrating custom eight-wavelength LED sources, a new analog front-end, and Bluetooth low energy 5 communication onto a single flexible printed circuit board encapsulated in medical-grade silicone. To support quantitative analysis with broadband LED sources, we introduce a wavelength-weighted fitting (WWF) algorithm that explicitly accounts for LED emission spectra, photodiode responsivity, and tissue optical properties. System performance was characterized through spectral calibration, stability testing, and noise equivalent power measurements, yielding an average NEP of 123 ± 25 fW/√Hz. Validation experiments using solid silicone phantoms demonstrated accurate recovery of effective attenuation coefficients with a mean error of 2.1% using a calibrated multi-distance approach. Liquid phantom studies incorporating blood, Intralipid, and yeast were used to evaluate recovery of hemoglobin and to investigate cytochrome c oxidase (CCO) related spectral sensitivity, parameter coupling, and model limitations. Finally, a pilot measurement in a term infant with transient respiratory distress demonstrates the preliminary feasibility of stable, continuous cerebral oxygenation monitoring in the neonatal intensive care unit. Together, these results support MW-FlexNIRS as a promising platform for multi-wavelength wearable NIRS in neonates and highlight both the potential and current challenges of oxCCO quantification.

Continuous respiratory rate monitoring is essential for early detection of clinical deterioration, but conventional methods like capnography are often impractical outside high-acuity settings. We developed a low-cost, Bluetooth Low Energy (BLE) respiratory rate monitor incorporating a MEMS temperature-humidity sensor (AHT21) and Nordic nRF52840 microcontroller, designed to attach externally to a standard oxygen mask. The device detects breaths using rapid, cyclic changes in exhaled temperature and humidity via a slope-based time-domain algorithm. Bench validation over 48 hours using a breathing simulator (5-60 breaths/min, tidal volume 300 mL, 34 °C, 95-100% RH) demonstrated 96.4% overall breath detection accuracy and strong correlation with programmed rates (r = 0.996, P < 0.0001). Sensor saturation occurred at high respiratory rates (60 breaths/min), and high ambient humidity levels (90% RH) both of which reduced measurement accuracy. Human testing in three healthy adults showed strong agreement with capnography (mean bias 0.44 breaths/min; 95% limits of agreement -1.86 to 2.75), with minor under-detection at the highest rate (50 breaths/min). The device operated reliably across oxygen flow rates of 4-15 L/min, with reduced performance at 0-2 L/min due to saturation of the sensor. These results demonstrate that BLE-enabled MEMS thermohygrometric sensors can accurately monitor respiratory rate under controlled conditions while highlighting operational limits due to sensor saturation. This study extends prior work on humidity-based respiratory monitoring and illustrates the potential for low-cost, portable respiratory rate devices suitable for low-acuity or resource-limited clinical settings.

Growing demand for safer and energy efficient vehicles is driving the development of advanced sensors, such as smart tires capable of sensing multiple parameters and processing data locally. However, this leads to increased energy requirements, incompatible with batteries of conventional tire pressure monitoring systems (TPMS). Strain energy harvesting offers a viable approach to harness the forces acting on tires. This article presents a complete energy harvesting solution for sustainable battery-less operation of automotive Bluetooth smart tire sensors. High-voltage synchronous electric charge extraction (HV-SECE) from flexible PVDF transducers is proposed, modeled in overdamped regimes, and validated for the first time in a realistic automotive scenario. The proposed HV-SECE circuits can manage input voltages up to 300 V and provide a 3 V regulated voltage in a minimum footprint area down to 22 × 20 × 7.3 mm. At 40 km/h, in 205/55-R16 tires, the maximum harvested power from 10 × 5 cm bimorphs made with 45 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">µ</i>m-thick PVDF foils amounts to 925 <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">µ</i>W. Even at low speeds, HV-SECE interfaces can cold-start the wireless tire sensor in less than 60 s, faster than regulatory TPMS requirements, and ensure energy surplus.

The rising average age in advanced economies is driving an increasingly urgent demand for technological solutions to monitor population health. Given that cardiovascular diseases remain the leading cause of mortality worldwide, the continuous assessment of cardiovascular function has become critically important. In this context, Pulse Wave Velocity (PWV) analysis represents a valuable tool for cardiovascular evaluation. However, the widespread adoption of PWV monitoring remains limited by the absence of wearable devices suitable for continuous use. This work addresses this gap by developing a wearable PWV monitoring device based on Force Sensing Resistor (FSR) technology designed to ensure high wearability. This objective was achieved by creating a miniaturized Printed Circuit Board (PCB) integrating signal acquisition and conditioning circuitry and a Bluetooth Low Energy (BLE) module for wireless data transmission. The data were analyzed by extracting the fiducial point, known as intersect tangent point (ITP), for the estimation of Pulse Transit Time (PTT) and the subsequent calculation of PWV. The system was validated on a cohort of 101 voluntary participants by comparing the PWV values obtained with the proposed device against those measured using the clinical gold standard, SphygmoCor. The results demonstrate a strong correlation between the two measurement systems, with a mean error close to zero 0.03 m/s) and a standard deviation of 1.03 m/s). These findings confirm the reliability and accuracy of the proposed solution in a clinical context.

Falls represent a major health risk for older adults living independently, motivating the development of unobtrusive and privacy-preserving monitoring solutions. This study investigates whether Bluetooth Low Energy (BLE) 6.0 Channel Sounding (CS) can support device-free fall detection using low-complexity signal representations suitable for residential deployment. The proposed system employs two BLE nodes performing periodic channel sounding, from which only scalar distance estimates are extracted. Time-domain and temporal-dynamic features are computed from sliding windows of the distance signal and used for supervised classification. Three widely used classifiers-Support Vector Machine with radial basis function kernel, Random Forest, and gradient-boosted decision trees (XGBoost)-are evaluated under both a default operating point and a sensitivity-first regime achieved through validation-based decision threshold adjustment, reflecting the higher cost of missed fall detections in assisted living scenarios. Experiments conducted in a furnished indoor environment with six participants performing realistic fall and non-fall scenarios demonstrate strong window-level sensitivity under subject-independent evaluation, with XGBoost providing the most favorable sensitivity-specificity balance. Under sensitivity-first operation, very high recall is achieved at the expense of increased false alarms. Given the limited dataset and single-environment setting, the reported results should be interpreted as a proof-of-concept demonstration of feasibility rather than definitive large-scale performance. The findings suggest that BLE CS captures motion-relevant signal variations that may support practical fall detection while maintaining low deployment complexity and privacy preservation.

The “Remote Keyless Entry” system stands out as one of the most prevalent wireless interface features found in nearly all commercial vehicles. However, the integration of the Internet of Things (IoT) into automotive applications has significantly impacted the design and functionalities of this system. The infusion of IoT technology is not only redefining the Remote Keyless Entry system but also enhancing safety and the overall user experience in the realm of automotive technology. This paper presents the design and implementation of a novel platform called the “Connectivity-Control-Unit” (CCU), which serves as a central IoT hub for a given vehicle and is extended to incorporate the Keyless Entry system. The paper highlights the features of a designed CCU, including testing and communication protocol specifics. The benefits and efficacy of multiple, independent communication channels, such as Bluetooth low energy and cellular networks, have been utilized to enhance the security aspects of the presented design of CCU. Finally, the paper presents real-world performance test results obtained from deploying the CCU in an actual vehicle, validating the system's practicality and effectiveness. The overall findings indicate that the presented IoT platforms for automobiles can significantly influence the design and functionality of keyless entry systems, leading to enhanced user experience and improved safety. The CCU, with its design features highlighted in this paper, is an ideal vehicle monitoring and access control device for the rental car segment.

Inhaling particulate matter (PM2.5) and ozone during sports exercise dramatically disrupts performance. We describe a unique wearable electrochemical sensor device to assess athletes’ real-time pollutant exposure during training and competition. Personal exposure evaluation is achieved using electrochemical sensors, wireless data transfer, and machine learning algorithms. We created a Bluetooth Low Energy body sensor network with wearable electrochemical sensing nodes and physiological monitors that communicate with a mobile app and cloud backend. College track and field athletes’ 21-day training data was used to validate the framework by comparing sensor values to reference monitoring stations. Results indicated high correlation (r = 0.725, p < 0.001) between wearable electrochemical measurements and standard devices for both PM2.5 and O3. The algorithm assessed inhaled doses 18.4% more accurately than permanent monitoring stations by detecting spatiotemporal pollution exposure fluctuations. We found that personalized wearable electrochemical sensors can measure sports exposure in real time, enabling evidence-based training program optimization and performance prediction.

Indoor positioning in hospitals is challenging because global navigation satellite systems signals are unavailable and existing solutions struggle with complex indoor propagation and high maintenance requirements. Fingerprinting-based methods using Wi-Fi, Bluetooth Low Energy (BLE), or magnetic field depend on extensive site surveys, while time or angle-based systems such as ultra-wide band, angle of arrival, and Wi-Fi round trip time require additional infrastructure. Recent machine learning approaches improve performance but remain limited by Pedestrian Dead Reckoning (PDR) drift and unstable spatial representations. This study proposes an AI-generated spatial pattern matching framework that integrates an AI-based PDR model with BLE Received Signal Strength Indicator (RSSI) to construct a user RSSI surface. Spatial similarity between user-generated patterns and the pre-built radio map is evaluated using Surface Correlation (SC), and a bi-directional candidate generation strategy with SC-based heading correction is employed to mitigate inertial drift. Experiments in a real hospital setting show that the proposed method achieves robust and accurate localization even in complex indoor environments where conventional fingerprinting and PDR techniques often fail. The results indicate that combining AI-driven inertial modeling with SC-based spatial pattern matching offers a practical and infrastructure-friendly solution for hospital indoor positioning.