Engineering

Abstract: Trainable wave networks process radio-frequency (RF) signals in analog form before digitization, and have been proposed to reduce the cost of converting every antenna signal in a large array. They combine four physical mechanisms: memory, coupling, resonance, and nonlinearity. We performed a systematic simulation-driven ablation study across these mechanisms on a representative wave network for multi-antenna Direction of Arrival (DOA) estimation, and paired each removal with a variant designed to fail if our explanation were correct. On a task with a coherent reflection, removing memory raised the normalized score from at most 0.518 to 0.873, since the order of array snapshots carries no information. Coupling and resonance could be removed without loss of accuracy once a trained linear stage combined the antenna signals, whereas removing the nonlinearity reduced the scores to 0.15 and 0.03. The front end that remains, a trained linear combination of the antennas followed by square-law detection averaged over 16 snapshots, achieves scores of 0.985 and 0.940 on the two main tasks, at least as high as every resonator variant. At matched accuracy, it converts 50% to 75% fewer bits than the strongest digital estimators we tested. Based on our findings, we propose it as an effective new baseline for wave networks in direction finding.

Abstract: This paper presents an end-to-end, CPU/GPU-based real-time framework for code-independent static positioning using signals of opportunity (SoOP) from IridiumNext and Orbcomm. The proposed architecture combines low-cost SDR signal acquisition with Doppler feature extraction, satellite identification, and fast orbital dynamics prediction, followed by parallel robust/nonlinear estimation and fusion. To mitigate the problems of outliers and non-ideal Doppler conditions, a Huber nonlinear least-squares stage is used for robust initialization, while EKF and UKF provide sequential refinement. To accommodate variations in measurement uncertainty and potentially multimodal state distributions, an interacting multiple-model unscented Kalman filter (IMM-UKF) and a Gaussian sum filter (GSF) are implemented, respectively. These five estimators operate concurrently, and their outputs are fused using a heatmap-based accumulation with anisotropic rotated Gaussian kernels and adaptive search-range control, enabling stable real-time operation under intermittent satellite availability. In a 30-minute outdoor static experiment (15° elevation mask, cloudy conditions, −8°C), five IridiumNext and four Orbcomm satellites were detected (of 13 and seven that were predicted to be visible, respectively). Post-processing analysis revealed approximately Gaussian but non-zero-mean Doppler residuals, indicating effective frequency offsets potentially associated with receiver and transmitter oscillator uncertainties. An investigation of zero-Doppler crossing (ZDC) events highlighted their potential for coarse frequency-bias calibration. With a processing window of δt=0.01s, the system achieved a minimum 2D positioning error of ~1m and a final 2D error of ~9.91m at the end of the test, thus demonstrating the feasibility of real-time, code-independent SoOP static positioning with meter-level performance under real-world conditions.

Abstract: While Low Earth Orbit (LEO) satellite constellations face increasing vulnerabilities in contested radio frequency environments, tropospheric scatter (troposcatter) communications provide a resilient, space-independent Beyond-Line-of-Sight (BLOS) alternative for tactical networks. However, escalating climate volatility increasingly induces non-stationary atmospheric ducting layers—such as marine evaporation ducts and thermal inversions—that severely alter the vertical modified refractivity profile, causing catastrophic multipath trapping, deep fading, and link outages. Classical numerical wave solvers suffer from prohibitive computational complexity that prevents real-time deployment on tactical edge modems, whereas pure data-driven deep learning models lack physical inductive biases and fail during out-of-distribution weather anomalies. To address these fundamental limitations, we propose a hybrid Quantum Physics-Informed Neural Network (QPINN) integrated within a hierarchical Space-Ground Environmental Digital Twin (SG-EDT) architecture for real-time troposcatter link adaptation. The QPINN encodes classical atmospheric telemetry into a variational quantum circuit with entangling gates, leveraging Hilbert-space representation learning to evaluate concurrent multi-variable cross-correlations. Crucially, the model embeds exact 2D narrow-angle Parabolic Wave Equation constraints alongside a Huberized thermodynamic gradient penalty bounded by physical atmospheric trapping boundaries directly into the autograd optimization loop. Evaluated on tactical edge hardware, our framework eliminates loss penalty explosions, achieves high-fidelity refractivity profile reconstruction, and executes low-latency forward inference, enabling preemptive Adaptive Transmit Power Control (ATPC) and spatial nulling before multipath link degradation occurs.

Abstract: Conventional single-mode coupling-of-mode (COM) theory fails to describe multimode coupling between acoustic modes in surface acoustic wave (SAW) resonators. Although multimode COM theory has been reported, explicit analytical formulas for perturbed wavenumbers and rule discovery of mutual coupling effects are still insufficient. In this work, a multimode COM model considering complete coupling terms is established for periodic grating structures. The analytical expressions of perturbed wavenumbers for two-mode coupling are derived, which directly quantifies how mutual-coupling coefficients influence dispersion relations and eigen-modes near Bragg condition. Numerical investigations reveal that both coupling strength and frequency separation between modes jointly distort Bragg stop-band features. Further, finite-length IDT admittance simulations demonstrate that even weak coupling can induce observable deviations on both main-mode resonance and remote spurious responses, indicating that resonator response cannot be simply superposed by independent single-mode results. Finally, the proposed multimode COM model is applied to fit finite-element simulation results of LN/SiO₂/SiC heterostructured SAW resonator. Compared with conventional single-mode COM, this model successfully reproduces spurious resonances far away from main resonance. This work provides an efficient theoretical tool for broadband modeling and spurious-mode analysis of multimode thin-film SAW resonators.

Abstract: This paper presents the design, fabrication, and experimental validation of a compact wearable UHF RFID tag integrated onto a superhydrophobic textile substrate for industrial worker identification and tracking. The proposed antenna combines a meander-line structure with a rectangular loop to achieve a small form factor while maintaining robust electromagnetic performance. Electromagnetic simulations demonstrate that the antenna exhibits an input impedance of 34 + j 205 Ω at 868 MHz, closely matching the complex conjugate of the Alien Higgs-3 microchip, and achieves a gain of 1.75 dBi with a directivity of 1.9 dBi. Fabrication was performed using inkjet printing with conductive inks, and the microchip was mounted with a conductive adhesive. Experimental measurements confirm a read range of approximately 4 m under a standard test setup, in good agreement with theoretical predictions. The proposed design demonstrates high reliability, flexibility, and washability, making it suitable for integration into personal protective equipment (PPE) for real-time worker monitoring in industrial environments.

Abstract: In this letter, we propose an integrated sensing and communication (ISAC) system based on an electronically steerable parasitic array radiator (ESPAR) antenna that employs a single-radio frequency (RF) chain. By jointly designing the baseband transmit waveform and the dynamic reactance profile of the parasitic elements over time, the proposed system exploits mutual electromagnetic coupling as a spatio-temporal synthesis degree-of-freedom (DoF). We formulate a non-convex co-design optimization problem that minimizes the mean square error (MSE) between the achieved and desired radar transmit beam patterns while satisfying the signal-to-interference-plus-noise ratio (SINR) requirements of communication users (CUs), varactor reactance tuning limits, slew-rate hardware constraints, and transmit power constraints. To solve the problem, we develop an alternating optimization (AO) framework that combines majorization-minimization (MM) and convex relaxation techniques. Numerical results confirm that the single-RF-chain ESPAR-based ISAC architecture achieves promising dual-functional sensing and communication performance while significantly reducing hardware complexity and power consumption compared with conventional multi-RF-chain digital array system.

Abstract: The growing demand for intelligent and adaptive services is accelerating the transition from conventional communication systems to semantic communication. Task-Oriented Semantic Communication (ToSC) has emerged as an efficient framework that reduces communication cost by transmitting only task-relevant information for downstream processing. This survey reviews the theoretical foundations, system architectures, and enabling technologies of ToSC, providing a taxonomy of design strategies spanning collaborative edge inference, end-to-end optimization, and multi-agent cooperation. Key technologies, including deep learning-based semantic representation and adaptive joint source-channel coding, are discussed. Device heterogeneity and partial observability are examined alongside recent tasks' performance evaluations, such as classification and recognition. The survey also discusses the integration of large-scale multimodal foundation models into semantic encoders for zero-shot task adaptation, continual learning frameworks for on-device semantic processing, cooperative multi-agent reasoning with interpretable information exchange, and cross-layer co-design within AI-native 6G infrastructures, leveraging reconfigurable intelligent surfaces and digital twins. By consolidating existing knowledge and identifying key design trade-offs in latency, accuracy, and bandwidth efficiency, this work aims to serve as both a reference and a roadmap for advancing ToSC toward reliable and resource-efficient communication in 6G and beyond.

Abstract: The classical Extended Euclidean Algorithm (EEA)-based decoder for binary Goppa codes is designed to correct up to \( \lfloor w/2 \rfloor \) errors, where \( w \) is the degree of the Goppa polynomial. When the number of errors exceeds this correction capability, the EEA decoding output may contain non-binary symbols, resulting in an invalid estimate of the transmitted binary codeword. This paper proposes an EEA-based decoding framework enhanced by a localized maximum-likelihood (LML) refinement step, referred to as the EEA-LML decoder. In the proposed method, the received vector \( r\in\mathbb{F}_2^n \) is first decoded using the EEA decoder to obtain a tentative estimate \( \hat{\mathbf{c}} \). If \( \hat{\mathbf{c}} \) contains non-binary components, these positions are treated as ambiguous and resolved through a localized likelihood-based search subject to the syndrome constraint, yielding a final binary estimate \( \bar{\mathbf{c}}\in\mathbb{F}_2^n \). Monte Carlo simulations are conducted over additive white Gaussian noise (AWGN) and flat Rayleigh fading channels using BPSK modulation. The bit error rate (BER) performance of the proposed EEA-LML decoder is compared with the locator-only EEA decoder, maximum-likelihood decoding (MLD), and Patterson decoding for selected binary Goppa codes. The simulation results show that the proposed EEA-LML decoder improves the BER performance relative to the locator-only EEA decoder, while maintaining a more practical decoding structure than exhaustive MLD. The comparison with Patterson decoding further illustrates the potential benefit of the localized refinement step in improving the reliability of EEA-based decoding beyond the conventional correction radius.

Abstract: This paper presents two related benchmark studies of AI-enabled wireless power transfer (WPT) for FR3 (7–24 GHz) ultra-massive MIMO (UM-MIMO) 6G base stations. Study 1 benchmarks time-slotted WPT scheduling across multiple IoT devices, comparing naive, sustainability-aware, round-robin and proportional-fair policies with a learned time-to-depletion predictor. Across 80 randomized 15-device populations, naive battery-priority scheduling saved none of the 238 physically sustainable devices, sustainability awareness raised the save rate to 23.5%, and a proportional-fair policy weighted by harvesting surplus performed best (51.7%), ahead of the learned predictor (42.4%). Study 2 asks whether an array should share its aperture between communication and WPT by partitioning its elements (power-splitting) or by alternating between the two functions in time (time-switching) under time-varying battery and queue demand. Power-splitting yielded shorter communication queues at every tested distance, and at 5 m, adaptive power-splitting avoided battery depletion in all 40 trials, whereas fixed allocation depleted the battery in every trial; adaptivity provided no such benefit under time-switching. Together, the studies indicate that the aperture-sharing architecture sets the achievable ceiling and the scheduling policy determines how close a deployment comes to it, so the architecture decision should be made first.

Abstract: This paper presents two related benchmark studies of AI-enabled security for FR3 (7-24 GHz) ultra-massive MIMO (UM-MIMO) 6G IoT networks, addressing two different layers of the security stack. Study 1 evaluates machine learning approaches for network-layer intrusion detection using a telecom-relevant benchmark dataset (UNSW-NB15), cross-validating multiple classifiers and reporting a systematic literature synthesis, data-quality audit, and hyperparameter-tuned test-set performance comparison. Study 2 evaluates physical-layer angle-of-arrival (AoA) spoofing detection for FR3 UM-MIMO IoT authentication, benchmarking equal and adaptive snapshot-allocation policies for multi-device authentication under a shared sensing budget, validated against closed-form Cramer-Rao bound theory. Together, the two studies show that AI-enabled security benchmarking must be evaluated separately at each layer, network-layer intrusion detection depends primarily on model choice, hyperparameter tuning, and data-quality awareness, while physical-layer spoofing detection depends primarily on how a shared sensing budget is allocated across devices, indicating that a defence-in-depth FR3 UM-MIMO IoT security posture needs distinct, layer-appropriate AI benchmarking rather than a single unified approach.

Abstract: Integrated sensing and communication (ISAC) is evolving from dual-functional waveform and transceiver design toward an end-to-end network capability in which sensing observations are acquired, coordinated, processed, protected, and exposed to applications. This paper presents an architecture-centric synthesis of that evolution, organized using six operational contracts: input/output, uncertainty, timescale, decision, evaluation, and admissibility–applied across seven functional layers. These contracts provide a consistent basis for examining whether information produced at one layer remains valid, timely, interpretable, and implementable at the next. The survey covers physical-layer signal models; waveform and beamforming design; sensing metrics and communication–sensing tradeoffs; cooperative and service-oriented ISAC; intelligence and edge computation; propagation and hardware constraints; security and trust; and standards, datasets, prototypes, and testbeds. We pay particular attention to metric compatibility, uncertainty propagation, information age, computation and deadline reliability, sensing-service exposure, and the distinction between achieved performance and feasible operating points. The strength of reported results depends on their supporting evidence, ranging from analytical formulations and simulations to hardware measurements, real-time prototypes, and networked testbeds. A representative literature audit and research-gap matrix identify recurring requirements for cross-layer operation, including the transfer of estimator uncertainty and calibration validity, synchronization, characterization of latency tails and deadline violations, and feasibility checks for learned actions. The resulting synthesis identifies end-to-end interface design, trustworthy decision making, and reproducible validation as central requirements for deployable ISAC systems.

Abstract: Cybersecurity education must explain how awareness is associated with willingness to enact protective practices while maintaining a clear distinction between intentions and verified behavior. This study examines a Cisco-oriented awareness–appraisal–intention framework using 385 completed primary-survey records reported as collected from technology users in Bangladesh. Protection motivation theory motivates two parallel appraisal pathways between cybersecurity awareness and secure-practice intention. The analysis combines measurement diagnostics, standardized composite-score regression with heteroskedasticity-consistent inference, 10,000 bootstrap resamples, and ten repetitions of ten-fold cross-validation. Awareness is positively associated with threat appraisal and coping appraisal. In the intention equation, awareness, threat appraisal, and coping appraisal have coefficients of 0.213, 0.197, and 0.273, respectively, with R-squared of 0.294. The two indirect associations are 0.077 and 0.124, but their difference is uncertain. The full model reduces held-out root mean squared error from 0.473 to 0.400 on the original five-point scale. Internal consistency is low across all four scales, and substantial discriminant-validity problems limit confirmatory interpretation. Accordingly, the reported associations are exploratory evidence from the analyzed survey records, not proof of causal mechanisms or actual secure behavior. The findings motivate improved instrument validation, task-specific security measurement, and longitudinal evaluation.

Abstract: Useful quantum-key supply from intermittent satellite contacts depends on finite-block security, authentication, relay inventories, and application demand, not optical transmittance alone. We simulate a trusted high-altitude-platform (HAP) relay between a ground station and a 500 km low-Earth-orbit satellite using three-intensity decoy-state BB84, finite-key estimation, authenticated key inventories, and a 200-pass service model. The comparator is the satellite-to-ground (S→G) downlink; a groundto-satellite uplink is not simulated, so the margins apply only to this downlink. With a 0.20 m HAP receive radius, the relay yields 13.175 Mbit/pass, 15.65% of the 84.185 Mbit/pass direct yield. With a 0.50 m HAP radius, the relay yield rises to 116.981 Mbit/pass, 38.77% above the 84.298 Mbit/pass direct yield. The unchanged direct-link estimates vary by under 0.2% from independent Monte Carlo sampling. The advantage remains 29.57% under a common contact window and 28.42% with equal elevation masks. Protocol tuning raises the 0.20 m relay yield by 36.74% without closing its gap to the direct link. For workloads derived from the 0.20 m relay supply, a 30 Mbit buffer produced no starved schedules for the 0.50 m relay in 10,000 simulations per demand. Receive-aperture collection is the main modeled condition for relay advantage under the stated assumptions.

Abstract: 5G Advanced and 6G V2X Environments require intelligence that is real-time, resource-light and robust across high-density vehicle mobility, non-stationary channels and hard ware constrained edge nodes. Existing AI-driven resource man agement approaches — including CNNs, LSTMs, GNNs, Deep Reinforcement Learning, Transformers, and Federated Learning — have limitations, such as high computational overhead, large data requirements, poor generalization under topology churn, and catastrophic forgetting under continuous retraining. These limitations render them ill-suited for deployment, especially at the level of radio units. To address this issue, this paper proposes NeuroChaos-V2X, a chaos-inspired edge intelligence framework that integrates vehicle mobility, wireless channel, and edge computing states into a unified multi-domain state vector. A Neu roChaos Feature Transformation module, employing GLS and Logistic-map neurons, converts raw network measurements into chaos-derived descriptors — firing time, firing rate, energy, and entropy — without requiring deep architectures or large training datasets. These features drive a unified NeuroChaos Decision Engine supporting edge server selection, task offloading, resource prioritization, and handover management simultaneously. The framework’s low-sample learning capability, CPU-only footprint, and sensitivity to small state perturbations make it inherently suited to resource-constrained V2X edge tiers. This position paper establishes the conceptual foundation and motivates experimental validation against GNN, DRL, and Transformer baselines in future 6G-V2X simulation environments.

Abstract: The increasing use of connected devices in medical laboratories generates diverse Internet of Medical Things (IoMT) settings whose real network architecture is not yet well understood. Therefore, in this multicentric in situ evaluation of IoMT topologies in eight fertility and andrology laboratories in Morocco (n=5 for assisted reproductive technology (ART); n=3 for andrology), our main objective is to classify the topology for each laboratory based on the biosensitive sample (gametes and embryos) analysis site, with topology reconstruction. The topology classification and centrality analyses included degree, density, betweenness, and closeness. Networks consisted of 14–34 nodes and 13–33 edges with densities of 0.06–0.14. Despite the large disparities in network scale, technology, and architecture, all eight networks fulfilled (E=N-1), suggesting a common sparse tree-like backbone at the core of star-centric and hybrid star-cellular architectures. Centrality profiles revealed significant site-specific organization, with maximum degree centrality from 37% to 82% and maximum betweenness from 71% to 99%, indicating highly centralized networks and structural bottlenecks. These results provide an empirical baseline for future optimization scenarios in ART and andrology laboratories with IoMT, considering the risk of electromagnetic interference linked to sensitive biological samples and the maintenance of equipment.

Abstract: This paper presents two related benchmark studies of artificial intelligence (AI)-enabled wake scheduling for ultra-low-power devices in Frequency Range 3 (FR3, 7–24 GHz) ultra-massive multiple-input multiple-output (UM-MIMO) Internet of Things (IoT) networks, evaluated as Pareto-frontier trade-offs between energy and access performance. Study 1 benchmarks wake scheduling for narrowband IoT (NB-IoT) devices under simulated bursty traffic, comparing a classical fixed-interval policy, a genie-aided policy with perfect intensity knowledge, and a learned traffic-intensity estimator, and characterizes the energy–latency frontier together with tail (95th-percentile) latency. Study 2 addresses beam-sweep-aware wake scheduling, in which an idle-mode UM-MIMO IoT device must also account for directional beam-sweep paging overhead, and evaluates listen-window policies against a learned arrival-time predictor on an energy–reliability frontier. In both studies, learned, time-varying state estimates did not uniformly improve on well-designed fixed-policy baselines. They helped in specific regimes, substantially reducing tail latency relative to the genie-aided reference in Study 1 and uniquely reaching near-perfect paging reliability in Study 2, whereas simple classical policies remained preferable elsewhere. The mechanism behind each gain differed between the two problems, indicating that gains from learned prediction are regime- and context-dependent rather than universal.

Abstract: Unmanned aerial vehicles increasingly exchange commands, telemetry, and mission data through interconnected ground, edge, and aerial networks. Their communication security depends on authentication, key management, implementation assurance, and constrained resources rather than on a cryptographic primitive alone. This narrative review examines lightweight authentication for UAV networks, with emphasis on the role of elliptic curve cryptography (ECC) and the trade-offs among software, FPGA, and hybrid implementations. Representative protocol studies are organized by communicating parties, trust assumptions, supporting mechanisms, and reported evidence. Standards for authenticated key establishment, authenticated encryption, and post-quantum cryptography provide a separate reference layer. The synthesis distinguishes mathematical security claims, symbolic protocol analysis, software timing, hardware realization, and operational deployment. ECC can support compact classical public-key mechanisms, but it does not independently provide message confidentiality, replay protection, capture resistance, or quantum resistance. PUF-assisted and infrastructure-assisted designs redistribute trust and implementation costs rather than eliminating them. A reporting framework identifies the information needed for meaningful comparisons, including credential provisioning, state recovery, communication overhead, and end-to-end energy. The review presents no new authentication protocol or experimental benchmark; its contribution is a critical, implementation-aware organization of the evidence and research gaps.

Abstract: Vehicle-to-Everything (V2X) networks may experience intermittent connectivity and frequent link disruption due to high node mobility, variable density, and heterogeneous mobility patterns. When a contemporaneous end-to-end path cannot be guaranteed, Delay-Tolerant Networking (DTN) enables message delivery through store-carry-forward communication. Among DTN routing protocols, PRoPHET exploits historical and transitive encounters to estimate destination-specific delivery predictability. However, its conventional Greater Than (GRTR) forwarding rule has limited contextual awareness because relay selection is based primarily on relative delivery predictability. This paper proposes ProphetRouterRL, a two-stage routing protocol that augments PRoPHET with Q-learning. First, PRoPHET identifies eligible relay candidates satisfying ΔP > 0. A Q-learning agent then determines whether to send or hold each message using a six-dimensional state comprising relay-buffer availability, encounter history, message load, remaining time-to-live, hop count, and node type. ProphetRouterRL was evaluated against conventional PRoPHET using the ONE Simulator in a map-based urban scenario representing the UKM Bangi area in Malaysia. The seven-day simulation comprised 129 heterogeneous nodes, including pedestrians, cyclists, cars, buses, metro trains, and stationary communication points. ProphetRouterRL increased the delivery probability from 46.47% to 93.12%, while reducing routing overhead and average latency by 39.4% and 41.4%, respectively. These results demonstrate the potential of combining PRoPHET’s delivery-predictability estimation with context-aware learned forwarding decisions in opportunistic V2X environments.

Abstract: Underwater multistatic sonar cooperative detection faces the issue that distant sonars fail to capture target echoes. This paper proposes a networkcodingaided multistatic multicast system with relaysonar assistance for underwater acoustic networks. Relay sonars forward activesonardetected target data to remote sonars and apply network coding prior to forwarding to cut retransmissions and boost transmission efficiency. To mitigate latency in traditional opportunisticlisteningbased network coding, a feedbackdriven proactive coding scheme is adopted: remote sonars feed back receivedpacket indexes, enabling relays to proactively encode complementary missing packets. A throughput analytical model is built to derive closedform expressions under maximum encodingnumber constraints for arbitrary remotesonar quantities. Simulations reveal prominent networkcoding gains at moderate packetloss rates. In highloss scenarios, the scheme improves throughput efficiency by 30%60%, outperforming schemes without network coding or relays and supporting reliable, bandwidthsaving underwater acoustic transmission.

Abstract: Millimeter-wave communications at 60 GHz offer access to large contiguous spectrum but require accurate propagation models to support reliable system design, as path losses at these frequencies are strongly influenced by buildings, vegetation, balconies, and intervening obstacles. This paper presents path gain models for outdoor-to-balcony and outdoor-to-indoor fixed wireless access (FWA) links at 60 GHz, derived from an extensive measurement campaign conducted in Athens, Greece. Measurements were collected with a street-mast transmitter and receiving nodes placed on balconies or indoors, covering a range of non-line-of-sight (NLOS) geometries representative of residential FWA deployments. Two existing channel models are adapted, tested and shown to predict measured path gain with lower RMS error than existing 3GPP models. Link-budget calculations using the proposed models show that, under the stated assumptions, 60 GHz FWA can deliver gigabit-class service to balcony-mounted receivers at short residential ranges with a 400 MHz deployment bandwidth.