Deploying AI-based emotion recognition at the edge enables real-world applications but is constrained by data scarcity, heavy models, hardware limits, and privacy issues. To overcome these, we propose CHEER (Cloud-HEavy to Edge-Ready), a self-supervised, transfer-efficient framework where the cloud pre-trains lightweight graph-based encoders using unlabeled data, stores them as frozen models, and deploys only the needed encoder. New users are locally matched via centroids of clusters, and a small on-device classifier is trained with minimal labeled data, reducing computation, memory, and energy use while preserving privacy. Experimental results in the WEMAC and WESAD datasets show an accuracy of 78.19% and 80.08% at the edge on a NVIDIA Jetson Orin Nano. Moreover, CHEER achieves more than 60% reduction in model size, and lowers both peak RAM usage and energy consumption by more than 50% compared to the state-of-the-art.

The rapidly increasing number of Internet of Things devices and the forthcoming arrival of quantum computing is imposing the transition to new quantum-resistant cryptography in constrained-resources devices. However, the energy impact of post-quantum cryptography on wireless sensor networks, critical to battery life and network up-time, remains largely unquantified. In this paper, this gap is addressed by evaluating the energy impact of standardized post-quantum cryptography CRYSTALS-Kyber and FALCON on extreme edge Internet of Things devices. In this work a new energy aware multilevel security framework over Contiki-NG has been developed and tested on extreme edge nodes based on EFR32MG12 Cortex-M4 System-on-Chip. This work provides energy consumption data measured in four different operational scenarios during one-hour experiments, allowing to estimate the impact on battery lifetime for four different battery capacities: 125, 225, 500 and 1000 mAh. The results presented in this work demonstrate that while the highest security level offers post-quantum confidentiality and integrity incurs in measurable costs, leading to a maximum lifetime reduction lower than 1% in the experiments with FALCON being identified as the dominant contributor to energy consumption related to post-quantum cryptography. At intermediate security levels, only post-quantum confidentiality using CRYSTALS-Kyber can be achieved at considerably lower energy costs. This study provides quantitative evidence supporting the feasibility of deploying post-quantum cryptography in resource-constrained wireless sensor networks through a novel energy-aware approach, offering crucial data for designing secure and sustainable quantum-resistant Internet of Things systems.

Intelligent transportation systems are critical for enhancing safety and efficiency in railway applications. However, their benefits must extend to rural and remote areas, requiring the deployment of autonomous extreme edge devices capable of handling complex computational tasks. Multi-object tracking plays a crucial role in these types of situational awareness systems, as it ensures temporal coherence across time and improves interpretability. Conventional approaches often assume that only lightweight tracking algorithms, such as probabilistic point-based, are viable for deployment on embedded devices due to their computational constraints. In this work, alternative existing trackers are explored and implemented demonstrating their feasibility and effectiveness, through a comprehensive experimental evaluation, as candidates to be deployed in constrained environments and resource-limited scenarios. Moreover, state-of-the-art tracking algorithms are optimised to reduce computational demands while maintaining high performance, particularly for processing point-cloud data. Using Light Detection and Ranging sensors as a case study, the feasibility of real-time execution is demonstrated. The proposed approach is adaptable to a different range of complex sensor technologies and data, making it suitable for other real-time tracking applications.

Designing AI-based applications personalized to each user's behavior presents significant challenges due to the cold start problem and the impracticality of extensive individual data labeling. These challenges are further compounded when deploying such applications at the edge, where limited computing resources constrain the design space. This paper introduces a novel approach to AI-driven personalized solutions in biosensing applications by combining deep learning with clustering-based separation techniques. The proposed Clustering and Learning for Emotion Adaptive Recognition (CLEAR) methodology strikes a balance between population-wide models and fully personalized systems by leveraging data-driven clustering. CLEAR demonstrates its effectiveness in emotion recognition tasks, and its integration with fine-tuning enables efficient deployment on edge devices, ensuring data privacy and real-time detection when new users are introduced to the system. We conducted experiments for model personalization on two edge computing platforms: the Coral Edge TPU Dev Board and the Raspberry Pi with an Intel Movidius Neural Compute Stick 2. The results show that initial cluster assignment for new users can be achieved without labeled data, directly addressing the cold-start problem. Compared to baseline validation without clustering, this proposal improves accuracy metric from 75% to 81.9%. Furthermore, fine-tuning with minimal labeled data significantly improves accuracy, achieving up to 86.34% for the fear detection task in the WEMAC dataset while remaining suitable for deployment on resource-constrained edge devices.

Securing sensitive medical data in IoT-based healthcare systems is increasingly critical due to growing cybersecurity threats and the emergence of quantum computing. This paper addresses these challenges by proposing a decentralized architecture that integrates blockchain technology and postquantum cryptography (PQC) to ensure data integrity, privacy, and resilience. The solution employs the Practical Byzantine Fault Tolerance (PBFT) consensus algorithm to maintain a tamper-resistant ledger of data locations, while encrypted medical records are distributed across Raspberry Pi nodes. Communications between sensor nodes and storage servers are protected using PQC primitives such as CRYSTALS-Kyber and Falcon. The methodology includes a lightweight protocol designed for resource-constrained environments and a storage scheme that optimizes scalability and energy efficiency. Experimental results validate the scalability and effectiveness of this architecture in real-world healthcare settings, demonstrating its potential as a secure and sustainable framework for electronic medical data management.

Negative emotions can lead to a variety of physiological and psychological problems. Identifying and interpreting negative emotions can help people and specialists to deal with their effects on the human body. This paper introduces a deep learning-based method for recognizing negative emotions utilizing physiological signals. It is based on extracting a set of 123 features from raw signals and organizing them into 2D feature maps. A hybrid CNN-LSTM model is then employed to classify these maps, simultaneously learning and integrating integral and sequential features to output emotional feedback. Feature fusion is incorporated during training to improve the method’s accuracy and reduce the data complexity. The approach has been validated using WEMAC and WESAD datasets, achieving F1-scores of 85.15% and 91.44% and accuracies of 85.03% and 90.98%, respectively. Furthermore, the feasibility of this method for real-world applications is demonstrated through deployment on the Coral Edge TPU, an embedded device, indicating its potential for run-time decision-making on the computing edge.

The rapid growth of the Internet of Things has expanded the research and implementation of wireless sensor networks in various application domains. However, the challenges associated with resource-constrained sensor nodes and the need for ultra-low power consumption pose significant problems. One fundamental strategy to address these challenges is transmission power control, which adjusts the transmission power of nodes to optimize network performance and lifetime. While traditional methods have focused on static scenarios, this work presents a novel approach for mobile wireless sensor networks based on a grey-fuzzy-logic transmission power control. The proposed system integrates grey prediction techniques with fuzzy inference to dynamically adapt transmission power levels. Unlike previous simulations, this work focuses on real implementations, considering practical aspects of wireless sensor network deployments and the characteristics of embedded sensor platforms. The objectives of this work are twofold: first, to implement a grey-fuzzy-logic transmission power control on an Internet of Things embedded hardware platform, ensuring compatibility with IEEE 802.15.4 networks, and second, to propose a runtime adaptive link recovery mechanism to enhance the robustness of the adaptive transmission power control in mobile and unstable contexts. Additionally, a multi-hop mobile grey-fuzzy-logic transmission power control strategy is introduced, enabling collaborative transmission power adaptation among sensor nodes. Experimental tests demonstrate the high prediction accuracy of the proposal even in multi-hop scenarios, confirming the system’s scalability. Results also show that the proposed system outperforms other strategies in terms or energy consumption, achieving up to 43 % of gains depending on the scenario.

Wireless sensor networks aim to collect environmental data for monitoring and decision-making purposes, often relying on low-power sensor nodes with limited computational resources, which makes it challenging to secure these networks using costly cryptographic primitives. Moreover, the emergence of quantum computers threatens traditional cryptographic schemes, and postquantum cryptographic schemes have been proposed as a solution. This work focuses on studying the behavior and performance of different combinations of postquantum digital signatures and key exchange mechanisms in wireless sensor networks where the number of nodes is large, including CRYSTALS-Dilithium, Falcon, SPHINCS+, CRYSTALS-Kyber, NTRU, and Saber, with a focus on their interaction and impact on network scalability. Simulation models are employed to generate metrics related to network functionality, application quality, and scalability with dynamic node behavior. The findings provide insights into the behavior of different combinations of postquantum schemes in wireless sensor networks and contribute to understanding their suitability and potential challenges in real-world deployments. In particular, the combination of Falcon and CRYSTALS-Kyber seems to be the most promising candidate for deploying secure sensor networks in the future. However, other combinations can present a better performance depending on their interactions with the parameters of the final application.

Applying affective computing techniques to recognize fear and combining them with portable signal monitors makes it possible to create real-time detection systems that could act as bodyguards when users are in danger. With this aim, this paper presents a fear recognition method based on physiological signals obtained from wearable devices. The procedure involves creating two-dimensional feature maps from the raw signals, using data augmentation and feature selection algorithms, followed by deep learning-based classification models, taking inspiration from those used in image processing. This proposal has been validated with two different datasets, achieving, in WEMAC, WESAD 3-classes, and WESAD 2-classes, F1-score results of 78.13%, 88.07%, and 99.60%, respectively, and 79.90%, 89.12%, and 99.60% in accuracy. Furthermore, the paper demonstrates the feasibility of implementing the proposed method on the Coral Edge TPU device, prepared to make inferences on the edge.

Smart cities and Industry 5.0 have the potential to improve efficiency and sustainability in industrial environments, logistics and intelligent transportation, through the use of advanced technologies such as European Rail Traffic Management System (ERTMS) level 3 in case of the railway domain. This type of technology allows for the localisation of trains and other rail assets, leading to improved safety, reliability, and capacity on railway lines. In this study, we propose a Wireless Sensor Network (WSN) support system for railway composition and integrity management using Internet of Things (IoT) technology, particularly applicable for freight transport. Our hardware solution is designed to function under extreme conditions and features a dynamic discovery system that combines range-based and range-free approaches using Received Signal Strength Indicator (RSSI) as a coverage indicator for train positioning. The system was tested in both static and dynamic scenarios and has shown promising results in a real-world use case.

The application of electronics, communication, and telemetry technologies in the field of sport training is a major trend today, specially with the explosion of the Internet of Things (IoT) paradigm. In particular, the rowing sport is one of the disciplines that could benefit from efficient, easy to integrate, and low cost IoT technologies to help analyzing and optimizing the performance of the rowers. However, traditionally monitoring tools in rowing are expensive and very limited. In this work, a rowing propulsion monitoring system for onboard analysis of rower’s key parameters is proposed, based on the development of a low-cost IoT device that combines sensing, global positioning, wireless communication, and data fusion for runtime information processing and monitoring on the boat. A power calculation model during the stroke is also proposed in order to estimate the energetic efficiency of the strokes in relation with the boat velocity. An acceleration sensor on the oar is used to decompose the movement; a torsion sensor to calculate the rower’s strength and energetic values; and navigation data for speed and distance values of the boat. A comprehensive comparative experimentation in a real rowing training scenario is performed, applying the proposed system and considering various rowers. The results show that the device provides a good vision in time of the movement and also very valuable technical information that can be used by rowers and trainers to define specific optimization profiles and improvement targets.

Research on post-quantum cryptography aims to solve the problematic of modern public-key cryptography being broken by attacks coming from quantum computers in the future and, moreover, by using classical electronics. This task is so critical that the National Institute of Standards and Technology (NIST) is in the final process of standardizing post-quantum schemes for the future protection of embedded applications. Though there are some research work done on embedded systems, it is important to study the impact of these proposals in realistic environments for the Internet of Things (IoT), where the limited computational resources and the strict requirements for power consumption can become incompatible with the usage of cryptographic schemes. In this work, the performance of one of the finalists of the standardization process called NTRU is studied and implemented in a custom wireless sensor node designed for applications in the extreme edge of the IoT. The cryptosystem is implemented and evaluated within the processes of the Contiki-NG operating system. Furthermore, additional experiments are performed to check if commonly integrated hardware peripherals for cryptography inside modern microcontrollers can be used to achieve better performance with NTRU, not only at the single node level but also at the network level, where the NTRU key encapsulation mechanism is tested in a real communication process. The results derived from these experiments show that NTRU is suitable for modern microcontrollers targeting wireless sensor networks design, while old devices present in popular platforms might not afford the cost of its implementation.

Light detection and ranging (LiDAR) sensors are increasing in popularity due to the advantages they provide over 2-D sensors in IoT object detection and classification applications, because of their ability to provide very precise distances to objects. Deep learning algorithms need a huge amount of data during training to obtain high accuracy results. When using 2-D images, a vast quantity of data sets are publicly available, but this is not the case for LiDAR point clouds. Each LiDAR model generates a point cloud with unique properties, which causes the data sets not to be compatible between different LiDAR models. As a result, when using deep learning with LiDARs, it is necessary to generate the data sets manually. For this purpose, the data must be captured and then labeled one by one, which is a very time and cost-consuming process. To overcome this issue and to reduce the development time when using LiDAR sensors with deep learning algorithms, a methodology is proposed in this article to automatically generate point cloud data sets using a 3-D simulator for autonomous cars. In this regard, a data set can be generated for any LiDAR model by adding the specific LiDAR parameters to the simulator. Besides, custom scenarios can be designed and generated, based on the final deployment location, to provide a simulated solution very close to the final implementation. With the proposed methodology, a simulation can be performed to select the LiDAR that best fits certain application requirements, in contrast to the traditional approach where the LiDAR must first be purchased.

In this article, a methodology based on machine learning for fault detection in continuous processes is presented. It aims to monitor fully distributed scenarios, such as the Tennessee Eastman process (TEP), selected as the use case of this work, where sensors are distributed throughout an industrial plant. A hybrid feature selection approach based on filters and wrappers, called hybrid Fisher wrapper method, is proposed to select the most representative sensors to get the highest detection quality for fault identification. The proposed methodology provides a complete design space of solutions differing in the sensing effort, the processing complexity, and the obtained detection quality. It constitutes an alternative to the typical scheme in Industry 4.0, where multiple distributed sensor systems collect and send data to a centralized cloud. Differently, the proposed technique follows a distributed approach, in which processing can be done eventually close to the sensors where data is generated, i.e., at the edge of the Internet of Things. This approach overcomes the bandwidth, privacy, and latency limitations that centralized approaches may suffer. The experimental results show that the proposed methodology provides TEP fault-detection solutions with state-of-the-art detection quality figures. In terms of latency, solutions obtained outperform in 37.5 times the implementation with the highest detection quality, using 1.99 times fewer features, on average. Also, the scalability of the framework provides a design space where the optimal implementation can be chosen according to the application needs.

Internet of Things based technologies are enabling the digital transformation in many sectors. However, in order to use this type of solutions, such as wireless sensor networks, in scenarios like transport, industry or smart cities, the deployed networks must meet sensible safety and security requirements. This article describes a Wireless Sensor Network design that applies multi-layered mechanisms and tools to ensure security, safety and reliability while maintaining usability in Rail and Industrial IoT scenarios. The proposed solution provides guidelines for choosing the best implementations given usual restrictions, offering a modular stack so it can be combined with other solutions.

In this work a novel on-site toolset-based architecture for tackling the main challenges of deploying and commissioning large scale WSN-based systems is proposed. This is one of the first implementations that addresses a complete set of runtime algorithms to efficiently deploy sensor platforms in the target scenarios based on the inclusion of the real behavior of the nodes within the in-situ simulation chain, combined with the integration of runtime diagnosis and reprogramming strategies to analyze the performance of the deployment in-field.

This 1-Page Demonstration paper is included in the track “Multimedia Systems and Applications”. The work has been already published in [1] and [2]. The main idea of the demonstration is to show how the Virtual Architecture ARTICo3works within a high performance wireless sensor node called HiReCookie. The selected demo includes an image processing application with several filters running as different kernels within the architecture ARTICo3. The virtual architecture works in a Spartan-6 FPGA included in the HiReCookie Node, [3] and [4]. During the demonstration, an image taken from a video camera attached to the node will be processed in real time by several dynamically reconfigurable kernels (median filters and edge detectors) under different working conditions. The solution scope includes solutions trading off among Low Power, Dependability and High Performance Computing.

Cyber-Physical Systems need to handle increasingly complex tasks, which additionally, may have variable operating conditions over time. Therefore, dynamic resource management to adapt the system to different needs is required. In this paper, a new bus-based architecture, called ARTICo3, which by means of Dynamic Partial Reconfiguration, allows the replication of hardware tasks to support module redundancy, multi-thread operation or dual-rail solutions for enhanced side-channel attack protection is presented. A configuration-aware data transaction unit permits data dispatching to more than one module in parallel, or provide coalesced data dispatching among different units to maximize the advantages of burst transactions. The selection of a given configuration is application independent but context-aware, which may be achieved by the combination of a multi-thread model similar to the CUDA kernel model specification, combined with a dynamic thread/task/kernel scheduler. A multi-kernel application for face recognition is used as an application example to show one scenario of the ARTICo3architecture.

The complexity of planning a wireless sensor network is dependent on the aspects of optimization and on the application requirements. Even though Murphy's Law is applied everywhere in reality, a good planning algorithm will assist the designers to be aware of the short plates of their design and to improve them before the problems being exposed at the real deployment. A 3D multi-objective planning algorithm is proposed in this paper to provide solutions on the locations of nodes and their properties. It employs a developed ray-tracing scheme for sensing signal and radio propagation modelling. Therefore it is sensitive to the obstacles and makes the models of sensing coverage and link quality more practical compared with other heuristics that use ideal unit-disk models. The proposed algorithm aims at reaching an overall optimization on hardware cost, coverage, link quality and lifetime. Thus each of those metrics are modelled and normalized to compose a desirability function. Evolutionary algorithm is designed to efficiently tackle this NP-hard multi-objective optimization problem. The proposed algorithm is applicable for both indoor and outdoor 3D scenarios. Different parameters that affect the performance are analyzed through extensive experiments; two state-of-the-art algorithms are rebuilt and tested with the same configuration as that of the proposed algorithm. The results indicate that the proposed algorithm converges efficiently within 600 iterations and performs better than the compared heuristics.

Remote reprogramming capabilities are one of the major concerns in WSN platforms due to the limitations and constraints that low power wireless nodes poses, especially when energy efficiency during the reprogramming process is a critical factor for extending the battery life of the devices. Moreover, WSNs are based on low-rate protocols in which as greater the amount of data is sent, the more the possibility to lose packets during the transmitting process is. In order to overcome these limitations, in this work a novel on-the-fly reprogramming technique for modifying and updating the application running on the wireless sensor nodes is designed and implemented, based on a partial reprogramming mechanism that significantly reduces the size of the files to be downloaded to the nodes, therefore diminishing their power/time consumption. This powerful mechanism also addresses multi-experimental capabilities because it provides the possibility to download, manage, test and debug multiple applications into the wireless nodes, based on a memory map segmentation of the core. Being an on-the-fly reprogramming process, no additional resources to store and download the configuration file are needed.

In this paper, a simulation tool for assisting the deployment of wireless sensor network is introduced and simulation results are verified under a specific indoor environment. The simulation tool supports two modes: deterministic mode and stochastic mode. The deterministic mode is environment dependent in which the information of environment should be provided beforehand. Ray tracing method and deterministic propagation model are employed in order to increase the accuracy of the estimated coverage, connectivity and routing; the stochastic mode is useful for large scale random deployment without previous knowledge on geographic information. Dynamic Source Routing protocol (DSR) and Ad hoc On-Demand Distance Vector Routing protocol (AODV) are implemented in order to calculate the topology of WSN. Hence this tool gives direct view on the performance of WSN and assists users in finding the potential problems of wireless sensor network before real deployment. At the end, a case study is realized in Centro de Electronica Industrial (CEI), the simulation results on coverage, connectivity and routing are verified by the measurement.

In this work a complete hardware-software support platform for a WSN testbed focused on developing wireless sensor applications in a simple and intuitive way is presented, as an alternative of commercial-motes-based testbeds that can be found in the state of the art. The main target of this hardware-software platform is to provide the highest abstraction level on the management of WSNs but in the simplest way in order to achieve a fast profiling mechanism for reliable prototyping based on the Cookies platform as well as helping users to develop, test and validate Cookie-Based WSN applications.

In this paper, we present the design and implementation of a prototype system of Smart Parking Services based on Wireless Sensor Networks (WSNs) that allows vehicle drivers to effectively find the free parking places. The proposed scheme consists of wireless sensor networks, embedded web-server, central web-server and mobile phone application. In the system, low-cost wireless sensors networks modules are deployed into each parking slot equipped with one sensor node. The state of the parking slot is detected by sensor node and is reported periodically to embedded web-server via the deployed wireless sensor networks. This information is sent to central web-server using Wi-Fi networks in real-time, and also the vehicle driver can find vacant parking lots using standard mobile devices.

In this paper, a solution to support the run-time read back, relocation and replication of cores in embedded systems with dynamic and partial reconfiguration capabilities is presented. The proposal shows a peripheral structure that allows an easy integration and communication with the rest of the system, including an API to make the reconfiguration details to be more transparent to software applications. Differently to other proposals, all functionality is implemented in hardware, achieving a higher reconfiguration speed. In addition, different design decisions have been taken in order to increase the portability of the solution to existing and, possibly, future FPGAs. Finally, a use case is provided, which shows the features of this module applied to the run-time scaling of a hardware coprocessor.