Department Publications

This paper proposes a secure and cost-effective communication and management approach for distributed disaster early warning networks to ensure reliable, low-latency alert dissemination. While decentralized architectures reduce latency by enabling direct device communication, they are highly vulnerable to attacks due to their exposure in public community settings and the challenges posed by Network Address Translation (NAT) and firewalls. Traditional NAT traversal techniques often lack the built-in security required for such sensitive infrastructure. We identify Nebula, an open-source overlay networking tool, as a viable solution because it is lightweight enough for low-resource devices such as the Raspberry Pi typically found in decentralized early warning systems, supports NAT traversal, and provides endto-end encryption. Preliminary experiments conducted in an AWS environment—designed to simulate real-world Raspberry Shake sensor deployments—confirm Nebula’s ability to establish direct peer-to-peer (P2P) connectivity with a single hop and effectively encrypt traffic

Voice communication is essential in post-disaster scenarios where users may be unable to type due to injury, stress, or environmental constraints. While Meshtastic provides a decentralized LoRa-based mesh network for long-range messaging, it does not natively support voice and is limited by low data rates and strict regulatory duty-cycle requirements. This paper investigates two methods for enabling voice over such networks: waveform-based audio packetization and a semantic speech-to-text (STT)/text-to-speech (TTS) approach. Preliminary experimental evaluations demonstrate that waveform-based methods, even when using low-bitrate codecs like Codec2, are impractical for urgent communication; delivering a 3-second voice message requires hundreds of packets and several minutes due to ransmission latency and duty-cycle constraints. In contrast, the proposed offline STT/TTS method transcribes speech locally, transmits it as a standard text payload, typically in a single packet and reconstructs the audio at the receiver. This semantic approach reduces delivery time from minutes to seconds, while maintaining high reliability and regulatory compliance. The study concludes that semantic communication is the most viable solution for voice interaction in bandwidth-constrained LoRa mesh environments.

This work presents the development of an opensource standalone (SA) 5G testbed with end-to-end slicing, using open-source software on commercial off-the-shelf (COTS) hardware and software-defined radio (SDR). Key contributions include customized slice configurations in the 5G Core Network (5GC) and dedicated static resource allocation in the radio access network (RAN). The testbed is validated through slice isolation tests. Its utility is then demonstrated via a warehouse application featuring Automated Guided Vehicles (AGVs) on one slice and environmental sensors & actuators on another. Critical AGV response times are evaluated under mixed traffic conditions, showing robust slice isolation and superior performance over baseline networks. The testbed offers a scalable, cost-effective platform for 5G education, research, and application testing. It also highlights the transformative potential of such open …

The deployment of 5G networks using open-source software platforms on commodity hardware and software defined radios (SDRs) has become increasingly important for both research and prototyping. Most work is constrained to a single platform such as OpenAirInterface (OAI), inhibiting the potential for cross-platform trade-offs. This paper presents Open5GFusion, a heterogeneous 5G Standalone (SA) testbed that combines OAI, srsRAN, and Open5GS to enable flexible experimentation with different core network (CN) and radio access network (RAN) alternatives and to exploit the strengths of each. A graphical user interface (GUI) simplifies deployment management and monitoring across different CN/RAN combinations. The testbed leverages commercial off-the-shelf (COTS) hardware with USRP B210 and N310 SDRs, allowing real-time evaluation of throughput, latency, modulation schemes, network slicing …

Unmanned aerial vehicles (UAVs)-assisted communication systems are considered as a promising technology in diverse verticals. This paper studies the deployment of UAVs in wireless sensor network (WSN) systems. Considering the energy-constrained nature of the wireless sensors, we have proposed a multi-UAV deployment algorithm that minimizes the maximum power transmitted among the sensor nodes (SN) for given minimum data collection rate, minimum data-transferring rate, maximum power and height constraints. The problem is divided into three subproblems in order to reduce the complexity involved in solving them as a single problem. The subproblems are UAV-SN association, 2D positioning of the UAVs and the altitude optimization of the UAVs. Each subproblem is optimized by fixing other parameters as constant. First, the UAV-SN association is addressed using a customized Gale–Shapley …

It is envisioned that 6G networks will be supported by key architectural principles, including intelligence, decentralization, interoperability, and digitalization. With the advances in artificial intelligence (AI) and machine learning (ML), embedding intelligence into the foundation of wireless communication systems is recognized as essential for 6G and beyond. Existing radio access network (RAN) architectures struggle to meet the ever growing demands for flexibility, automation, and adaptability required to build self-evolving and autonomous wireless networks. In this context, this paper explores the transition towards AI-driven RAN (AI-RAN) by developing a novel AI-RAN framework whose performance is evaluated through a practical scenario focused on intelligent orchestration and resource optimization. Besides, the paper reviews the evolution of RAN architectures and sheds light on key enablers of AI-RAN including digital twins (DTs), intelligent reflecting surfaces (IRSs), large generative AI (GenAI) models, and blockchain (BC). Furthermore, it discusses the deployment challenges of AI-RAN, including technical and regulatory perspectives, and outlines future research directions incorporating technologies such as integrated sensing and communication (ISAC) and agentic AI.

With growing applications of drones, there is a demand for drone audition functionality that allows microphone arrays to be embedded into or attached to the body of drones for sound source localization and signal enhancement. In practical acoustical environments, the performance of drone audition is degraded due to the free-field sound propagation assumption. In this paper, we propose (i) a method for modeling the scattering and diffraction of a drone body beyond free-field propagation as a Drone-Related Transfer Function (DRTF), and (ii) a practical method to estimate the DRTF coefficients using circular loudspeaker arrays. The method incorporates the acoustic reciprocity principle and a finite order spherical harmonic expansion by utilizing a discrete set of measurements by multiple loudspeaker placements on circular arrays. We validate our DRTF model and parameterization method on a known rigid sphere …

Test-time prompt tuning (TPT) has emerged as a promising technique for adapting large vision-language models (VLMs) to unseen tasks without relying on labeled data. However, the lack of dispersion between textual features can hurt calibration performance, which raises concerns about VLMs' reliability, trustworthiness, and safety. Current TPT approaches primarily focus on improving prompt calibration by either maximizing average textual feature dispersion or enforcing orthogonality constraints to encourage angular separation. However, these methods may not always have optimal angular separation between class-wise textual features, which implies overlooking the critical role of angular diversity. To address this, we propose A-TPT, a novel TPT framework that introduces angular diversity to encourage uniformity in the distribution of normalized textual features induced by corresponding learnable prompts …

Solar photovoltaic (PV) systems have become increasingly popular as a sustainable energy source due to the growing need for renewable energy on a global scale. However, the buildup of dust and dirt on solar panels frequently compromises the effectiveness of these systems. Existing cleaning methods often require manual labor, exhibit low efficiency, and are restricted to either a single cleaning method or operational mode. To address these issues, we designed and implemented a semi-autonomous solar panel cleaning robot capable of both dry and wet cleaning. The robot can autonomously navigate on solar panels with a tilt of up to 20 or be manually controlled using a mobile app, and it operates for approximately 75 minutes on a single charge. This study enhances key aspects of the robotic system, integrating advanced edge detection and transition mechanisms using both ultrasonic sensors and image processing techniques. We implemented a robust localization system based on bus-bar detection for precise navigation, upgraded the water supply mechanism and refined the orientation estimation algorithm.

The emergence of 5G technology has transformed network architecture, with network slicing technology standing out as a crucial feature for managing diverse and demanding applications. This innovative technology allows the creation of multiple virtual networks within a single physical infrastructure, each dedicated to specific needs such as latency, bandwidth, and reliability. These capabilities are vital for supporting services like enhanced mobile broadband, ultra-reliable low-latency communications, and massive machine-type communications. Predicting the optimal configuration of network slices is challenging due to the complex and dynamic nature of 5G environments. While traditional methods often lack the necessary accuracy, machine learning models have shown promise in addressing this challenge. However, the adoption of these models introduces the challenge of interpretability, as understanding how …

Material that constitute aggregate depend on their packing efficiency that is defined by the shape distribution of aggregates. Geometrical shape descriptors would not capture morphological aspects of different dimensional scales and frequencies, failing to wholistically numerically represent shape in material prediction models. Fourier shape descriptors are used to analyse single particle morphology, but efficiency to characterize shape of lump samples has not been assessed. This study analyses aggregates milled for different number of revolutions (0, 200, 500 and 1000) in a Los Angeles Abrasion Value instrument, that contrives morphological alterations of different scales. Means of zonal frequency components amplitudes (form, angularity and texture) are statistically different for all classes with more than 99 % confidence. Support Vector Machine and K Nearest Neighbour algorithms classified pairwise …

This manuscript introduces NeuroDetect, a model-free deep learning-based signal detection framework tailored for phase-modulated wireless systems with low-resolution analog-to-digital converters (ADCs). The proposed framework eliminates the need for explicit channel state information, which is typically difficult to acquire under coarse quantization. NeuroDetect utilizes a neural network architecture to learn the nonlinear relationship between quantized received signals and transmitted symbols directly from data. It achieves near-optimum performance, within a worst-case 12% margin of the maximum likelihood detector that assumes perfect channel knowledge. We rigorously investigate the interplay between ADC resolution and detection accuracy, introducing novel penalty metrics that quantify the effects of both quantization and learning errors. Our results shed light on the design trade-offs between ADC resolution and detection accuracy, providing future directions for developing energy-efficient high-speed and wideband wireless systems.

The rapid evolution of network technologies and frequent network intrusions have amplified the demand for intrusion detection systems (IDSs) powered by artificial intelligence (AI). However, the performance of IDSs is limited by labeled intrusion data and class imbalances in existing datasets. Collecting such data through live network simulations is costly, insecure, and poses regulatory challenges. This paper presents a multi-model framework that uses a conditional tabular generative adversarial network (CTGAN) to synthesize realistic network intrusion data. The framework addresses challenges such as unstable training, mode collapse, and handling mixed data types, which are common in most generative AI models when applied to non-Gaussian tabular data generation while validating the synthesized data through qualitative and quantitative analyses. Overlaid density histograms and cumulative distribution …

A four-dimensional light field (LF) captures both textural and geometrical information of a scene in contrast to a two-dimensional image that captures only the textural information of a scene. Post-capture refocusing is an exciting application of LFs enabled by the geometric information captured. Previously proposed LF refocusing methods are mostly limited to the refocusing of single planar or volumetric region of a scene corresponding to a depth range and cannot simultaneously generate in-focus and out-of-focus regions having the same depth range. In this paper, we propose an end-to-end pipeline to simultaneously refocus multiple arbitrary planar or volumetric regions of a dense or a sparse LF. We employ pixel-dependent shifts with the typical shift-and-sum method to refocus an LF. The pixel-dependent shifts enables to refocus each pixel of an LF independently. For sparse LFs, the shift-and-sum method introduces ghosting artifacts due to the spatial undersampling. We employ a deep learning model based on U-Net architecture to almost completely eliminate the ghosting artifacts. The experimental results obtained with several LF datasets confirm the effectiveness of the proposed method. In particular, sparse LFs refocused with the proposed method archive structural similarity index higher than 0.9 despite having only 20% of data compared to dense LFs.

Image-to-point cross-modal learning has emerged to address the scarcity of large-scale 3D datasets in 3D representation learning. However, current methods that leverage 2D data often result in large, slow-to-train models, making them computationally expensive and difficult to deploy in resource-constrained environments. The architecture design of such models is therefore critical, determining their performance, memory footprint, and compute efficiency. The Joint-embedding Predictive Architecture (JEPA) has gained wide popularity in self-supervised learning for its simplicity and efficiency, but has been under-explored in cross-modal settings, partly due to the misconception that masking is intrinsic to JEPA. In this light, we propose CrossJEPA, a simple Cross-modal Joint Embedding Predictive Architecture that harnesses the knowledge of an image foundation model and trains a predictor to infer embeddings of specific rendered 2D views from corresponding 3D point clouds, thereby introducing a JEPA-style pretraining strategy beyond masking. By conditioning the predictor on cross-domain projection information, CrossJEPA purifies the supervision signal from semantics exclusive to the target domain. We further exploit the frozen teacher design with a one-time target embedding caching mechanism, yielding amortized efficiency. CrossJEPA achieves a new state-of-the-art in linear probing on the synthetic ModelNet40 (94.2%) and the real-world ScanObjectNN (88.3%) benchmarks, using only 14.1M pretraining parameters (8.5M in the point encoder), and about 6 pretraining hours on a standard single GPU. These results position CrossJEPA …

Cardiovascular diseases rank among the leading causes of mortality worldwide and the early identification of diseases is of paramount importance. This work focuses on developing a novel machine learning-based framework for early detection and classification of heart murmurs by analysing phonocardiogram signals. Our heart murmur detection and classification pipeline encompasses three classification settings. We first develop a set of methods based on transfer learning to determine the existence of heart murmurs and categorize them as present, absent, or unknown. If a murmur is present it will be classified as normal or abnormal based on its clinical outcome by using 1D convolution and audio spectrogram transformers. Finally, we use Wav2Vec encoder with raw audio data and AdaBoost abstain classifier for heart murmur quality identification. Heart murmurs are categorized based on their specific attributes …

The Relative Transfer Matrix (ReTM), recently introduced as a generalization of the relative transfer function for multiple receivers and sources, shows promising performance when applied to speech enhancement and speaker separation in noisy environments. Blindly estimating the ReTM of sound sources by exploiting the covariance matrices of multichannel recordings is highly beneficial for practical applications. In this paper, we use covariance subtraction to present a flexible and practically viable method for estimating the ReTM for a select set of independent sound sources. To show the versatility of the method, we validated it through a speaker separation application under reverberant conditions. Separation performance is evaluated at low signal-to-noise ratio levels in comparison with existing ReTM-based and relative transfer function-based estimators, in both simulated and real-life environments.

The combined use of algorithms for approximate discrete cosine transform (DCT) with approximate computing is examined toward area-efficient field programmable gate array (FPGA) hardware architectures. The performance of previously reported, multiplierless approximate DCT transforms are explored in the presence of non-exact hardware adders which provide better utilisation of FPGA resources in terms of look-up-tables (LUTs). In the context of 2D image compression, provided examples are demonstrating a 20% LUT reduction in FPGA utilization compared to accurate adders, with a corresponding image quality degradation of 0.6 dB in terms of PSNR. The potential benefit of such area-efficient FPGA hardware designs is explored on 3D (video compression), 4D (static light field compression), and 5D (light field video compression) cases where significant FPGA resource savings are obtained at the cost of …

This concept paper proposes an architecture that integrates agentic AI models with digital twin representations for each subsystem component in an RF transceiver. The digital twins, constructed using machine learning techniques such as deep learning and reinforcement learning, are paired with specialized, autonomous AI agent(s) capable of independent reasoning, planning, and adaptive action. Agentic AI models go beyond traditional rule-based/symbolic AI to autonomously select optimal strategies, synthesize data from diverse sources, and interact with both virtual and real-world environments. Ideas for simulations and experimental validations are proposed for future work.

Future wireless systems are expected to rely on large antenna arrays at the base station (BS) to meet growing demands for data rates. However, the power consumption and hardware complexity of high-resolution analog-to-digital converters (ADCs) pose significant challenges to the practical deployment of such arrays. Low-resolution ADCs offer a power- and cost-effective alternative, motivating a deeper understanding of information-theoretic limits with low-resolution quantization. This letter establishes capacity scaling laws for both single-input single-output (SISO) and multi-antenna systems with low-resolution quantization. We show that, unlike the classical capacity growth in unquantized systems, the capacity of low-resolution systems approaches n at a rate for SISO and multiple-input single-output (MISO) systems, where n is the number of quantization bits and is the signal-to-noise ratio. For …

The energy consumption of telecommunication networks has been rising steadily due to the expansion of networks, power-hungry equipment and ever-increasing consumer demand. This contributes significantly to the world’s carbon footprint and greenhouse emissions. In particular, radio access networks (RANs) have been identified to have a larger contribution to the energy consumption of telecommunication networks, and a thorough evaluation is required to understand the avenues for energy reduction. In light of that, this study discusses various strategies to lower RAN power usage to minimize telecommunication’s environmental effect. Furthermore, a two-pronged strategy that combines dynamic operation management of base stations (BSs) and renewable energy integration is introduced to reduce the environmental impacts of RANs. With this hybrid strategy energy consumption and carbon footprint may be …

Real-time sensing and perception of the radio spectrum based on artificial intelligence (AI) is crucial for emerging intelligent wireless and electronic warfare systems. However, sensing can be greatly impacted by harmful radio frequency interference (RFI). Emerging drone warfare allows many RFI sources/jammers to be distributed across a wide field-of-view which necessitates real-time measurement, adaptation and aperture nulling to remove the RFI before AI-based sensing and perception of sources of interest can occur. This work explores algorithmic innovations that improve the computational complexity of classical Howells-Applebaum adaptive nulling algorithm to enable fast, real-time adaptive operation at significantly lower arithmetic complexity. Design examples for AI-enabled sensing and perception across a 32-element antenna receiver with 32 independent channels and a Xilinx Virtex-6 Sx475 FPGA …

This paper investigates an intelligent reflective surface (IRS) assisted secure multi-user two-way communication system. The aim of this paper is to enhance the physical layer security by optimizing the minimum secrecy-rate among all user-pairs in the presence of a malicious user. The optimization problem is converted into an alternating optimization problem consisting of two sub-problems. Transmit power optimization is handled using a fractional programming method, whereas IRS phase shift optimization is handled with semi-definite programming. The convergence of the proposed algorithm is investigated numerically. The performance gain in minimum secrecy-rate is quantified for four different user configurations in comparison to the baseline scheme. Results indicate a 3.6-fold gain in minimum secrecy rate over the baseline scheme when the IRS is positioned near a legitimate user, even when the malicious …

The demand for spectrum access continues to grow exponentially and dynamic spectrum access (DSA) is a candidate to alleviate issues with spectrum scarcity by adapting spectrum resources in real-time. Despite the benefits of intelligent DSA protocols, many algorithms are complex, and thus prohibitively difficult to implement in a real-time system. This paper proposes a contention-based access method named Guided-Aloha (or G-Aloha), that stems from the simple and well-known slotted-Aloha (S-Aloha) access method, for DSA in the context of primary users (PU) and secondary users (SU) of the network. The G-Aloha method is proposed for SU access to provide extended protection to the PU compared to the S-Aloha method by being guided by the knowledge from spectrum prediction (or spectrum forecasting), considering any potential regulatory requirements for protecting the PU. Using Markov modeling, an …

Drone swarms pose a threat to national security. There is growing interest in an “Iron Dome for America” to protect against weaponized drones, missiles and other threats. The sensing, classification and characterization of a swarm of drones-typically with hundreds of devices flying in close proximity to each other-is a challenging radar and AI/ML problem. In this paper, we report on early work on using state-of-the-art AI/ML algorithms with advanced digital signal processing towards demonstrated capabilities for UAS swarm detection and classification at the laboratory scale. An automotive grade 77 GHz FMCW radar evaluation board, and commercial (toy) micro drones are used for experimentation. The FMCW pulses use 3 GHz of bandwidth. All experiments are conducted within a short range whereas long range experiments reserved for future work.