ABSTRACT. Part A: Machine Learning Enhanced Microwave Biosensor for Glucose Detection in Fruit Juices: Rapid assessment of glucose in physiologically and commercially consumed liquids is important to health and food-quality monitoring. However, conventional enzyme-based methods are invasive and inefficient for continuous or real-time glucose quantification. This work proposes a miniature U-shaped microwave biosensor integrated with an interdigital capacitor (IDC) and fabricated on a Rogers RO4003C substrate for in vitro dielectric characterization of glucose based liquids. To realize bandpass response and enhance selectivity, a 1.5 pF series capacitor is incorporated into the feedline. The sensor demonstrates distinct and sharp resonance dips in both reflection coefficient (S11) and transmission coefficient (S21), validating its strong resonant response suitable for dielectric sensing. Glucose-water solutions with concentrations of 5%, 10%, 15%, and 25% were measured, demonstrating minimal resonant frequency drift and a repeatable 0.8 dB variation in the magnitude of S11, corresponding to an average sensitivity of approximately 0.05 dB/% glucose. Comparative measurements of fresh and processed mango, guava, orange, and grape juices indicated that different S-parameter responses are related to the concentration of sugar and the effect of processing. For automated data interpretation, a hybrid machine learning (ML) framework combining Random Forest (RF)–based feature selection, XGBoost refinement, and artificial neural network (ANN) classification was employed for juice-type discrimination. Glucose concentration is predicted using a deep neural network (128–64–32–1), achieving MSE of 0.2297, MAE of 0.21%, and R2 of 0.979, with over 95% of predictions within ± 2% error, demonstrating reliable in vitro glucose sensing and food-quality assessment. Part B: Sensing Antenna System for Termites Detection and Control : Termites infestation is considered one of the main challenges affecting the wood integrity, particularly in damp environments, causing extensive damage to wooden structures and resulting in heavy economic losses worldwide A Novel sensing antenna system for termites detection and control is proposed for combating termites. Considering the effect of the moisture content in the presence of termites, a non-invasive technique that is based on detecting the moist spots of wood is introduced. The proposed sensing antenna system consists of two identical microstrip-fed H-slot antennas and a reflector at a distance of 1 cm. An artificial intelligence-based solution is proposed to improve design efficiency and alleviate the burden on human efforts. By leveraging artificial intelligence (AI), this approach aims to streamline H-slot antenna design process. Based on the transmission coefficient level between the antennas, the moisture within the wood is detected. By utilizing the reflector, the system can achieve multiband operation and enhance the simulated gain to approximately 8.76 dBi, thereby improving its detection capability. The proposed single antenna was built on a commercial low cost FR4 substrate with compact dimensions of 20 mm × 20 mm × 1.6 mm. Through optimization of the stub dimensions, the proposed antenna resonates effectively at 4.2 GHz. Simulation results demonstrate that the system can achieve excellent moisture sensing capabilities not only across the wood surface but also to a depth of 0.95 inch using a perfect electric conductor (PEC) reflector. Regardless of the wood type, the proposed sensor also can identify moisture levels of 30%. The complete sensing system has been manufactured, and simulation results have been validated through measurements, confirming the efficiency of the proposed system as a moisture sensor for wooden structures.

ABSTRACT. Accurate liver segmentation in abdominal CT remains challenging because of low contrast boundaries and anatomical variability. Our work proposes UGRFormer, a Boundary-Aware and Uncertainty-Guided Refinement Transformer of 2.5D liver CT segmentation. The model consists of hierarchical Transformer encoder which models the global context, image-derived Sobel-based boundary priors which are injected to the multi-scale features. It also includes Uncertainty-Guided Refinement which rectifies ambiguous predictions through residual learning. The inter-slice continuity is represented in 2.5D input formulation at lower computational cost. Structural consistency is improved by using multi-scale decoding, deep supervision and auxiliary boundary prediction. Using patient-wise 5-fold cross-validation on 3D-IRCADb-01, UGRFormer achieved a mean Dice of 0.9355 and IoU of 0.8788, indicating that it has high volumetric overlap and better boundary delineation.

ABSTRACT. This paper presents a rapid and non-destructive approach to distinguishing mineral, purified, and tap water using near-infrared (NIR) transmission spectroscopy and chemometrics in the 900-1700 nm spectral region. Following z-score standardisation, Principal Component Analysis (PCA) was implemented to investigate spectral variance. The first principal component accounted for 67.78% of the overall variance, showing a significant discriminative structure in the dataset. Fisher discriminant analysis identified crucial wavelengths at 1155 nm, 1159 nm, and 1244 nm as the most informative spectral characteristics corresponding to second-overtone C-H and combination-band O-H absorption. Stratified 10-fold cross-validation was used to evaluate three supervised classification algorithms: k-Nearest Neighbour (KNN), Support Vector Machine with radial basis function kernel (SVM-RBF), and Linear Discriminant Analysis (LDA). With only one misclassification out of 120 samples, the confusion matrix revealed that the SVM-RBF and LDA classifiers had a typical classification accuracy of 99.2%. These results reveal that NIR spectroscopy offers an immediate, reagent-free, and highly sensitive technique for water source verification and quality monitoring when combined with proper pre-processing and chemometrics.

ABSTRACT. Bone fractures require accurate and timely diagnosis, while conventional imaging modalities such as X-ray, CT, and MRI suffer from limitations related to radiation exposure, cost, and accessibility. Microwave imaging has emerged as a non-ionizing alternative based on dielectric property contrast. In this work, a wearable slotted monopole antenna operating over 2.1–6 GHz is designed for near-field microwave imaging. The sensor has a compact size of 28.8mm × 48 mm and is implemented on a flexible Rogers 4350B substrate. It achieves good impedance matching while maintaining compliance with SAR limits (≤ 1.6 W/kg) The results demonstrate the feasibility of microwave sensing as a low-cost, portable, and safe complementary approach for bone fracture detection.

ABSTRACT. Underwater acoustics is the main sensing and communication approach used in underwater and deep sea environments, due to the severe limitations of electromagnetic wave propagation in water. With the growing interest in deep ocean exploration and increasingly complex underwater missions, acoustic systems have become essential tools for detection, localization, classification, and communication in both research and industrial applications. This talk provides an overview of underwater acoustics from a research perspective, with emphasis on acoustic signal behavior in different marine environments, including shallow, deep, and ultra deep waters. It discusses how depth related factors such as pressure, temperature variations, sound absorption, ambient noise, and multipath propagation affect signal transmission and system performance. The main challenges associated with deep sea operations are addressed, considering both physical propagation effects and signal processing limitations. Recent developments in underwater acoustic signal processing are also presented, including feature extraction methods, statistical models, and machine learning techniques for signal classification. The focus is on practical research approaches that aim to improve detection and recognition performance under noisy and highly variable underwater conditions. In addition, the seminar highlights the role of autonomous underwater vehicles (AUVs) and unmanned marine systems in modern underwater operations. Issues related to the integration of acoustic sensing with autonomous navigation, target classification, and decision making systems are discussed, with particular attention to the use of intelligent algorithms to enhance autonomy and reliability. Finally, selected examples from the recent researches on underwater acoustic signal classification and autonomous marine systems are introduced as case studies, demonstrating experimental results and real world implementations. The seminar concludes with a brief discussion of current challenges and future research directions in underwater acoustics and autonomous underwater technologies, emphasizing their importance for deep sea exploration, environmental monitoring, and maritime security.

ABSTRACT. Underwater Acoustic Sensor Networks (UWASNs) enable ocean monitoring but face severe acoustic channel challenges. This paper proposes integrating Long Range (LoRa) one of Chirp Spread Spectrum (CSS) techniques with Non-Orthogonal Multiple Access (NOMA) for UWASNs. The framework combines CSS modulation's interference resilience with power-domain multiplexing to enhance spectral efficiency in challenging underwater channels. We present uplink and downlink system models incorporating realistic path loss, ambient noise, multipath, and Doppler effects. MATLAB simulations demonstrate that LoRa-NOMA achieves aggregate spectral efficiency of 0.458 bits/s/Hz compared to 0.153 bits/s/Hz for equivalent orthogonal multiple access. Also attains enhanced improvement in total system throughput while maintaining acceptable user fairness index. LoRa with lower spreading factor (SF)=8 accomplishes reliable communication with bit error rate (BER) < 10⁻³ at signal to noise ratio (SNR) as low as -15 dB. Successive interference cancellation (SIC) feasibility is validated, with uplink showing superior performance through centralized interference management. The best performance user achieves BER approximately 10⁻⁴ at 15dB SNR. These results establish LoRa-NOMA as a promising solution for energy-efficient underwater networks.

ABSTRACT. Underwater Optical Wireless Communication (UOWC) has emerged as a compelling high-bandwidth alternative to conventional acoustic and radiofrequency (RF) communication systems for subsea applications. However, severe signal degradation due to absorption, scattering, turbulence-induced scintillation, and multipath temporal dispersion continue to challenge reliable high-speed data transmission. This paper presents a comprehensive technical design of an advanced UOWC system that integrates a Deep Belief Network (DBN)-based receiver architecture for robust signal reconstruction and denoising. The proposed system employs a Monte Carlo Simulation (MCS) framework for accurate modeling of the inhomogeneous underwater channel, governed by the Radiative Transfer Equation (RTE) with realistic scattering phase functions. A novel signal-to-image pixelization algorithm transforms the received one-dimensional time-series signals into two-dimensional grayscale feature maps, enabling DBN to exploit spatial pattern recognition for superior denoising. Simulation results demonstrate that the proposed DBN-enhanced receiver achieves a coding gain of 7.8–12 dB over conventional Maximum Likelihood Estimation (MLE) detectors at a Bit Error Rate (BER) of 10⁻³ under turbid harbor water conditions. The proposed architecture represents a significant advancement toward resilient, multi-gigabit underwater optical links for autonomous underwater vehicle (AUV) networks and marine sensor systems. Keywords: Underwater Optical Wireless Communication (UOWC); Deep Belief Networks (DBN); Monte Carlo Simulation; Signal Reconstruction; Scattering Phase Function; Turbulence Modeling.

ABSTRACT. Visible Light Communication (VLC) has emerged as a promising candidate for IOT applications and mobile beyond 5G/6G wireless systems due to its wide unlicensed spectrum and inherent security. However, to get a high data rate in VLC systems, achieving high data rates in VLC requires multicarrier modulation techniques such as optical OFDM, which introduce critical challenges, particularly power inefficiency caused by LED nonlinearity. Overcoming the system's high peak-to-average power ratio (PAPR) is one of the key solutions to get a power-efficient system. In this paper, a hybrid approach is proposed for a conventional FLIP-OFDM system by combining precoding and nonlinear companding techniques. Specifically, discrete sine transform precoding is integrated with A-law or μ-law companding schemes to further compress signal peaks. Simulation results demonstrate that the proposed hybrid FLIP-OFDM scheme achieves significant PAPR reduction compared to conventional FLIP-OFDM by 7.5 dB, while maintaining comparable BER performance.

ABSTRACT. CRC (cyclic redundancy checks) are bits appended to data to detect errors that may corrupt data while being transmitted or stored. The CRC bits are obtained by a polynomial division involving XOR operations which can be easily implemented in hardware or software. CRC, which were proposed 60 years ago, are widely used in practice, for example in 5G. Typically, sequences protected by CRC are encoded into codewords in an error-correcting code before transmission. The receiver first decodes the error-correcting code to correct errors caused by the channel and then CRC are used to detect miscorrections by the decoder. In this talk, we consider how the choice of the CRC and the codewords encoding CRC protected sequences affect performance.

ABSTRACT. The global movement towards resilient and sustainable smart cities is generating a wealth of knowledge and the emergence of de facto best practices. This presentation will explore international best practices and standards including successful approaches that are shaping the future of urban development. This talk will analyze how leading cities are benchmarking their progress against established sustainability targets and resilience metrics. Through a comparative study of diverse case studies, it will extract valuable lessons learned and highlight replicable best practices in areas like smart energy management, circular economy initiatives, climate adaptation strategies, and citizen engagement. It will also emphasize the role of standardization in facilitating knowledge sharing, promoting interoperability, and accelerating the adoption of effective solutions for building smart cities and communities that are both environmentally sound and robust in the face of evolving risks.

ABSTRACT. RAKICT representative will discuss Why this session topic matters. Then introducing the Technology Training Areas Offered by RAKICT, especially the Artificial Intelligence Training with AI CERTs. The main objectives of AI CERTs Tracks are to be explained for Computer Engineers, Developers and all related specialists. The Impact of Using AI Tools on the Real Work Scenarios is studied from different perspectives. Finally, the talk will be concluded with tips on Career Guidance for engineers interested in smart cities applications RAKICT will distribute a free online course as a gift to all IEEE members and attendees entitled: “AI+ Foundation Course” (self-based learning) provided by AICerts.

ABSTRACT. Nowadays smart cities are in bad need for sustainable and green communications solutions. Many challenges will be discussed to apply some advanced RF circuits and systems for sustainable development and green communications. Starting from energy harvesting techniques, an Outdoor RF spectral study available from cell-phone towers in sub-urban areas for ambient RF energy harvesting is investigated. Based on this measured data, a designer can decide the maximum distance away from a cell-phone tower that meets certain detection sensitivity. Consequently, two prototype designs of dual band radio frequency energy harvester (RF-EH) rectifier circuits are introduced to harvest RF energy from four different local RF ambient sources simultaneously. Both of the demonstrated prototype rectifier circuits are successfully tested in lab environment and shows improved results at low levels of incident RF power. Accordingly, energy sustainable IOT and RFID could be developed. In addition, an improved self‐interference canceller for X‐band transceivers are introduced which is applicable for radars of autonomous vehicles. Finally, an optimized technique is introduced for maximizing the RF power amplifier efficiency without compromising other performance parameters such as linearity and output power, helping network operators to have environment friendly infrastructure.