National Institute of Technology Goa

The Internet of Things (IoT) has emerged as a key technology for scalable, real-time environmental monitoring, addressing the growing global challenge of air pollution. This review synthesizes findings from more than twenty major deployments across urban, industrial and smart-city environments, including implementations from Hindustan Institute of Technology [1], Sathyabama Institute [2], Taiwan’s integrated governance framework [3], the Bulgarian Academy of Sciences [4] and low-cost deployments in developing regions. Comparative analysis spans sensing hardware, communication technologies (WiFi, LTE/4G, LoRa), cloud architectures, machine learning integration and field performance. Results indicate significant improvements in spatial coverage, real-time responsiveness and data-driven governance, while highlighting persistent challenges in sensor calibration, cybersecurity and long-term reliability. Emerging advances in edge intelligence, spectral sensing, federated learning and hybrid communication architectures are examined as pathways toward next-generation environmental monitoring. This consolidated review provides a structured framework for scalable, resilient and policy-integrated IoT air-quality monitoring systems.

Carbon fiber-reinforced aluminum laminates (CARALL), a hybrid composite, hold significant promise for use in aircraft structures. However, the distinct behaviors of the materials in this multi-layered configuration pose considerable challenges during drilling. High drilling forces and elevated temperatures can cause various forms of damage, potentially compromising the performance of CARALL structures in service. This research involved performing drilling experiments on CARALL samples produced inhouse, examining various combinations of cutting speeds (30, 60, and 90 m/min) and axial feed rates (0.05, 0.15, and 0.25 mm/rev) to assess their impact on thrust force, drilling torque, machining temperature, and the quality of the hole surface. Increasing the cutting speed from 30 m/min to 90 m/min while reducing the axial feed from 0.25 mm/rev to 0.05 mm/rev resulted in a 40.4% reduction in thrust force and a 59.8% decrease in drilling torque, but also caused a 70.4% rise in machining temperature. At elevated cutting speeds, the rise in temperature leads to a decrease in cutting force, demonstrating the relationship between force and temperature. An increase in axial feed led to rougher bore surfaces. Although the surface finish initially improved with increasing cutting speed, a further increase caused degradation once the machining temperature exceeded the glass transition temperature of the epoxy resin. An acceptable surface finish ( R a ) of 1.93 µm was achieved when an intermediate cutting speed of 60 m/min and a lower feed rate of 0.05 mm/rev were employed. Typical defects observed in the drilled holes included grooves, fiber bending, fiber exposure, matrix degradation, material smearing, adhesion, and delamination. The observed delamination was mainly due to matrix degradation at temperatures beyond the glass transition temperatures.

IntroductionFor neural prosthetic devices, accurate classification of high dimensional electroencephalography (EEG) signals is significantly impaired by the existence of redundant and irrelevant features that deteriorate the classifier generalization and computation efficiency. This work presents a new and unified optimal-driven framework to challenge these issues and improve EEG-based MI signal decoding.MethodsThe proposed method combines a modified feature selection model of coati optimization algorithm (COA) and different machine/deep learning classifiers. The novelty of the COA is its dynamic and parameter-free adaptation mechanism, in association with opposition-based learning a better exploration exploitation balance can be maintained in high-dimensional feature space. The generated optimized feature subsets are then employed to train a battery of classifiers such as support vector machines (SVM), random forests (RF), convolutional neural networks (CNN) and recurrent neural networks (RNN) for motor imagery task recognition. In experiments, we verify SSRC on commonly used benchmark EEG datasets such as the PhysioNet Motor Movement/Imagery dataset.ResultsThe experimental results showed that the COA + CNN model had the best performance of classification. The model demonstrated a classification accuracy of 96.8% of prediction, with precision at moderate AH hour and predicted as either being more likely to discharge or remain in care = 96.4%, recall = 96.9% and F1-score = 96.6%. This presents a remarkable 6.5% gain in classification accuracy over the best rival feature selection technique and significantly outperformed conventional metaheuristic algorithms such as PSO (90.3% accuracy) and GA (89.7% accuracy) as well as filter-type techniques such as mRMR (86.8%) and ReliefF (84.3%).Discussion/ConclusionThe combined evolved metaheatistic for feature subset selection with deep learning architectures is a powerful approach for an accurate classification EEG signals. The findings confirm that the COA-based approach provides a robust, computationally-efficient, and scalable method for achieving high-accuracy classification—essential for promoting the reliability and real-time operation of future neural prosthetic control systems.

In cognitive radio, wideband spectrum sensing is challenging in detecting low signal-to-noise ratio signals dominated by non-Gaussian noises. Traditional methods outlined in current literature often fail to provide satisfactory detection performance in the aforesaid scenario. To address this issue, we propose the use of a simultaneous likelihood ratio test (SLRT) as a statistical method to detect the primary signal in a non-Gaussian noise environment. The SLRT-based detector (SLRTD) detects the changes in the received signal statistics with respect to the additive noise. In the SLRT detector, the test statistics of the received signal are estimated using the robust standard deviation to counteract the impact of non-Gaussian noise. Furthermore, detection performance is enhanced through cooperative sensing in a multi-node environment. Our simulations, conducted in both Gaussian and non-Gaussian noise environments using a Universal-Filtered Multi-Carrier (UFMC) signal, demonstrate the superior accuracy of our robust detectors in the presence of non-Gaussian such as Bernoulli, Laplacian, symmetric α-stable distributions noise. This improved accuracy provides reassurance of the effectiveness of our proposed method.

This article reports on developing a beam-forming array of miniature printed radiators for mm -wave applications, particularly 6 G communication. The sole radiator is miniaturized by $10.52 \%$ and offers a unidirectional radiation pattern at 27 GHz. Further, an array with four identical individually excited antennas has been constructed to perform beam-forming operations. The main beam of the array can be controlled by providing an accurate progressive phase to each element during excitation. The inter-element space of the array has been optimized in such a way as to achieve a lower side-lobe level at the desired direction of radiation. Additionally, the array is capable of scanning its main beam in a dual direction simultaneously without any mechanical involvement. The proposed array is compact, miniaturized, and suitable for 6 G applications, like IoT, V2X, and satellite networks.