Resource
2026 EN
Kavin Kumar Thangadorai · Krishna M. Sivalingam · Kumar Murugesan
+1 more
First responder and mission-critical IoT applications require infrastructure-less, energy-efficient ad-hoc networks. We present MRAM , a custom handheld platform featuring a multi-radio mesh system that integrates long-range Wi-Fi and HaLow technologies. MRAM leverages the B.A.T.M.A.N. advanced protocol with multi-radio extensions for intelligent link selection, radio aggregation, and full-duplex operation. Real-world urban evaluations demonstrate reliable voice and video performance, achieving up to 1.2 km line-of-sight and 2 km non-line-of-sight coverage over single and multi-hop links, consistently outperforming single-radio systems. The system dynamically adapts to varying link conditions, node density, and mobility patterns, ensuring robust and fault-tolerant mesh connectivity. Complementary Mininet-WiFi simulations validate the MRAM decision engine, yielding 2–3× throughput gains and 20–30 ms latency reductions across diverse mesh topologies. Experimental results with up to 10 nodes confirm the scalability of the MRAM platform under increasing mesh size and asymmetric link conditions. Additionally, the platform’s radio silence intelligence extends per-node battery life by up to 75% and doubles overall mesh lifetime, as confirmed by energy modeling. Dynamic activation of radios and relay nodes ensures energy-aware operation at scale, further enhancing resilience in power-constrained scenarios. These results establish MRAM as a resilient, energy-efficient solution for next-generation field-critical IoT deployments.
Resource
2026 EN
Antoine Siebert · Bertrand Le Gal · Guillaume Ferre
+1 more
In radiocommunications, multipath propagation often induces Inter-Symbol Interference (ISI), which requires accurate channel estimation and equalization. Neural networks have recently shown promising performance for this task but remain limited by their high computational cost and lack of interpretability, hindering their adoption in mission-critical systems. To overcome these limitations, this study proposes a hybrid estimation architecture that combines a linear Kalman filter with two lightweight neural networks that dynamically and deterministically adjust the noise covariance matrices of the filter, enabling efficient and explainable channel tracking. Experimental results show performance gains of up to 5 dB at low SNR and 3 dB at high SNR compared with the Least Squares (LS) algorithm. This low-dimensional design enables real-time deployment on energy-constrained programmable platforms. Implementations on an ARM Cortex-A72 CPU and Artix-7/Virtex-7 FPGAs demonstrate throughputs of 53.7 Mbps and 8-20 Mbps, respectively, with FPGA energy consumption over 20× lower than on the CPU. These results confirm the effectiveness of the proposed NN-enhanced Kalman estimator for interpretable and energy-efficient channel estimation in embedded and defense applications.
Resource
2026 EN
Ritabrata Maiti · A S Madhukumar · Tan Zheng Hui Ernest
Industry 5.0 is transforming edge caching from a passive traffic offloading mechanism into an active enabler of distributed intelligence and real-time analytics. However, conventional caching strategies lack the collaborative adaptability, inference-awareness, and domain-specific customization required for dynamic Industry 5.0 workloads. This survey provides a comprehensive and systematic review of machine learning (ML)-driven edge caching organized around two fundamental pillars: collaborative caching (distributed, federated, and multi-agent reinforcement learning paradigms) and inference-enabled caching (content and model placement accelerating real-time edge analytics). We systematically examine how traditional ML, deep learning (DL), deep reinforcement learning (DRL), and multi-agent RL (MARL) techniques address domain-specific challenges across five Industry 5.0 applications: robotics, manufacturing, logistics, immersive training, and healthcare. For each domain, we provide comparative assessments based on latency requirements, scalability targets, and collaboration needs. Our analysis reveals fundamental tradeoffs: MARL excels for distributed coordination with promising efficiency gains but requires higher computational resources, while DRL balances adaptation and complexity, and federated learning enables privacy-preserving collaboration across multiple sites. Each application section concludes with Key Insights synthesizing technique suitability, domain requirements, performance gains, and unresolved challenges. We identify five critical research directions: Bayesian optimization for sampleefficient learning, federated workflow-aware caching with sub-second staleness, resilient caching under intermittent connectivity, scalable multi-modal caching with low latency, and mission-critical prioritization meeting industry standards. This work provides a roadmap for designing adaptive, safe, and scalable ML-driven edge caching systems for intelligent industrial operations.
Resource
2026 EN
Regina Ochonu · Josep Vidal · Gurjot Singh Bhatia
+1 more
Smart factories require wireless networks capable of supporting diverse applications with stringent and heterogeneous quality of service (QoS) demands. Network slicing enables this by virtualizing resources into isolated slices optimized for specific service types. To effectively manage scarce radio resources and ensure seamless support for mixed-traffic industrial users, this paper proposes a suite of slice-aware resource allocation (RA) frameworks encompassing power and subchannel (SC) allocation for both uplink (UL) and downlink (DL) in industrial time-division duplexing (TDD)-orthogonal frequency division multiple access (OFDMA) networks. We transform diverse QoS constraints—latency, reliability, jitter, and throughput—into unified rate expressions enabling tractable convex optimization across three slice types: capacity limited (CL) for best-effort traffic, ultra-reliable low-latency communication (URLLC) for mission-critical control, and time sensitive (TS) for deterministic periodic traffic. The framework incorporates dynamic TDD adaptation for asymmetric traffic patterns and admission control to manage congestion. The effectiveness of the proposed algorithms in reducing energy consumption while guaranteeing slice-specific QoS is validated using realistic industrial ray-tracing channels and extensive simulations.
Resource
2026 EN
Inam Ullah · Hesham El-Sayed · Alexis Dowhuszko
+2 more
The rollout of Fifth-Generation (5G) technologies has highlighted the need for robust, high-reliability communications to support mission-critical applications. This new kind of service requires that future mobile networks ensure seamless connectivity, high bandwidth, and stringent quality of service, particularly in emergency-response scenarios. To meet these stringent demands, the use of Integrated-Access-and-Backhaul (IAB) nodes mounted on unmanned aerial vehicles offers a promising solution by not only enhancing data rates via improved line-of-sight propagation conditions to the donor Base Station (BS) via backhaul link, but also enabling on-demand coverage in mission-critical scenarios requiring temporary cellular coverage. This paper investigates the performance gain that is feasible in a Millimeter Wave (mmWave) terrestrial network that uses non-terrestrial IAB nodes to enhance coverage in mission-critical applications, with a focus on optimizing the antenna down-tilt angles of both donor BS and IAB node to maximize the End-to-End (E2E) spectral efficiency (SE) for ground mobile users. The proposed framework aims to improve the backhaul link quality and ensure reliable, high E2E SE connectivity in urban environments characterized by non-line-of-sight conditions when donor BSs and mobile users are placed on different streets. Simulation results show that the joint optimization of donor BS and non-terrestrial IAB node antenna down-tilt angles, combined with best path selection, enhanced the E2E SE of mobile users by 105% at IAB down-tilt angles of −45° (3.9 bps/Hz), compared to the baseline −90° configuration (1.9 bps/Hz).
Resource
2026 EN
Hassan Jalil Hadi · Muhammad Khurram Khan · Naveed Ahmad
UAVs are widely used in mission-critical tasks but remain vulnerable due to open communication links. Especially, the Ground Control Station to UAV (G2U) communication channel is particularly vulnerable, often exploited for large-scale intrusions due to its multi-connectivity and openness. To address these risks, significant efforts have been devoted to developing Intrusion Detection Systems (IDS) based on machine learning for UAVs. However, most existing ML-based IDS solutions concentrate on UAV operations rather than G2U communication, lack real-time detection capability, and rely heavily on limited or synthetic datasets. To address this gap, we present GCS-NIDD, a real-time dataset that captures both normal and malicious traffic across nine attack types, including Replay, DoS/DDoS, Evil Twin, and Fake Landing attacks, among others. We build a physical testbed using actual UAV devices (PX4 Vision Dev Kit V1.5) and diverse GCS platforms (laptops, Tablets) to emulate real G2U communication scenarios. Furthermore, we propose G2UIDS, a multi-tier IDS framework that leverages complementary strengths of ML models across three layers, Tier 1 (LightGBM) performs anomaly detection to separate normal and malicious traffic, Tier 2 (TabNet) conducts fine-grained multi-class attack classification, and Tier 3 (BLS) focuses on detecting zero-day attacks. These outputs are combined through decision-level fusion, ensuring both accuracy and robustness. Unlike prior simulation-based solutions, G2UIDS is deployed and evaluated in a real operational environment. Experimental results demonstrate that G2UIDS achieves 93.16% accuracy and a 94.80% detection rate, significantly outperforming existing methods while maintaining low computational overhead.
Resource
2026 EN
Carlos S. Alvarez-Merino · Alejandro Ramirez-Arroyo · Rasmus Suhr Mogensen
+6 more
Reliable low-latency communication is a key requirement for mission-critical and mobile autonomous systems, including teleoperation, autonomous navigation, and real-time uplink-dominant telemetry applications. While commercial 5G networks often provide adequate downlink performance, uplink performance in rural deployments may be constrained by radio-resource limitations and uplink power-control mechanisms. This paper presents a comprehensive experimental evaluation of multi-connectivity strategies over commercial 5G Non-Standalone networks, based on measurement campaigns conducted in urban, suburban, and rural environments. The study analyzes per-packet uplink and downlink latency, packet loss, and radio-layer KPIs across two mobile network operators. The measurements indicate that latency and reliability cannot be inferred solely from coverage indicators such as RSRP. In coverage-constrained scenarios, performance appears to be strongly influenced by uplink power-limited operation and partially correlated impairments across operators. Several multi-connectivity strategies are evaluated, including link aggregation, switching-based policies, and conditional packet duplication. A Primary-Anchored Adaptive Failover (PAAF) framework is introduced to selectively activate redundancy based on radio, latency and service cost considerations. The results suggest that Partial Duplication (PD) approaches can approach the latency and reliability performance of full duplication while substantially reducing duplication overhead in the evaluated rural scenario. The proposed PAAF mechanisms are evaluated through offline replay of synchronized dual-operator traces, allowing switching, aggregation, full duplication, and PD to be compared under matched radio conditions within the evaluated commercial 5G NSA rural deployment.
Resource
2026 EN
Primatar Kuswiradyo · Shan-Hsiang Shen
Post-disaster logistics require rapid delivery under disrupted infrastructure and stringent operational constraints. Unmanned aerial vehicles (UAVs) can reach affected areas efficiently, yet their missions are limited by battery endurance, payload capacity, wind disturbances, and multi-UAV coordination requirements. This paper proposes Enhanced Adaptive Path Optimization (EAPO), a lightweight and deterministic heuristic for coordinated multi-UAV logistics. EAPO integrates (i) return-aware energy feasibility, (ii) non-partial single-visit service, and (iii) priority-weighted node selection into an online decision loop that operates without offline training or instance-specific parameter tuning. Simulation-based evaluations under stochastic wind and heterogeneous demand, involving scenarios with several hundred service nodes and multiple UAVs, demonstrate that EAPO achieves systematic reductions in total operation time (up to 15.3%) and energy consumption (approximately 5.7–12.6%) relative to baseline algorithms, while maintaining competitive average delivery time compared with established routing strategies. Runtime scaling experiments indicate smooth growth with respect to problem size, with collision-avoidance overhead remaining negligible within the evaluated operating regimes. These results position EAPO as an interpretable baseline for mission-level UAV logistics under explicit feasibility and service constraints.
Resource
2026 EN
Giulio Poggiana · Mohammed Terrah · Riccardo Torchio
+4 more
Wireless power transfer (WPT) systems offer a promising solution for automating the charging process and extending the mission duration of unmanned aerial vehicles (UAVs). However, the weight of the WPT system's receiving part mounted on the UAV can significantly impact its energy consumption and overall performance. In this regard, the ferrite plate is a critical component. While it is necessary for improving magnetic coupling and reducing the stray magnetic field, it also increases weight. Topology optimization (TO) is an effective tool for addressing this challenge and identifying the most suitable shape for the ferrite plate. In this study, a multi-objective topology optimization (TO) approach based on the solid isotropic material with penalization (SIMP) method is employed to design an optimized ferrite plate for the receiver side of a WPT system for UAV charging. The multi-objective optimization, implemented through a weighted-sum formulation, identifies ferrite configurations that maximize mutual coupling while minimizing material usage. Three representative Pareto-optimal designs are fabricated and experimentally validated against a full-ferrite reference plate. The results show that the reference configuration is a dominated solution in the Pareto sense, as equivalent mutual coupling can be achieved with approximately 40% less ferrite material. The trade-off between the resulting reduction in charging efficiency and the decrease in the required flight power due to ferrite reduction is analyzed, further demonstrating the potential of TO to substantially reduce system weight without compromising overall system performance.
Resource
2026 EN
Oskar Wintercorn · Jan van Deventer
Modern industrial applications are increasingly built from independently developed systems that discover each other at run time and compose into mission-specific behavior. The challenge is not only to collect data, but to preserve enough context to understand and diagnose the behavior of the resulting system-of-systems. This paper presents the Arrowhead Framework Ontology, an explicit, queryable semantic representation of Arrowhead framework deployments. Using a climate-control demonstrator, we export timestamped RDF snapshots from live systems, ingest them into GraphDB, validate them with SHACL, and apply OWL reasoning to derive additional facts about structure and dependencies. We show that this enables practical diagnosis and impact analysis via SPARQL, even though none of the individual systems were designed to answer such questions. The same representation also supports digital threads for traceability and lifecycle insight. Together, these results illustrate how explicit semantics can provide a non-intrusive path to more transparent and evolvable industrial systems.