ACTA TECHNICA NAPOCENSIS ELECTRONICS AND TELECOMMUNICATIONS
URI permanent pentru această colecțiehttps://oasis.utcluj.app/handle/123456789/447
Navigare
16 rezultate
Rezultatele Căutării
Articol DECISION TREES-BASED ALGORITHM FOR INTELLIGENT ALLOCATION OF PROCESSES IN CLOUD(Technical University of Cluj Napoca, 2025-06-25) DOLCESCU Constantin-Valentin; BOTEZ Robert; ZINCA Daniel; DOBROTA VirgilThe paper presents a decision tree–based scheduler for intelligent cloud process allocation that evaluates features such as source area, instruction count, payload size, priority, throughput, and delay to guide real-time placement decisions. The model was trained and validated on a diverse, scenario-driven synthetic dataset covering four controlled workload conditions plus randomized fallback cases. For the training dataset, the classifier achieved 93% accuracy, while for the validation and test set, an accuracy of 92% was obtained. A Kubernetes-inspired simulation framework further visualizes and confirms the scheduler’s allocation logic under dynamic conditions. These results underscore the approach’s effectiveness, interpretability, and suitability for production-grade cloud orchestration.Articol AUDIO DENOISING USING U-NET ARCHITECTURE(Technical University of Cluj-Napoca, 2025-06-27) JIMON, L.-Daniel; VAIDA Mircea-F.Audio denoising is a pivotal task in audio signal processing. This paper presents a machine learning approach using a U-Net architecture to denoise musical audio signals affected by four distinct noise types: white noise, urban noise, reverberation, and noise cancellation artifacts. The model was evaluated on datasets derived from IRMAS and UrbanSound8K. Objective and subjective evaluation metrics were used, which show the model's effectiveness in filtering white and urban noise. However, performance on reverberation and noise cancellation artifacts is limited, indicating areas for future architectural and methodological improvements.Articol TEXTURE BASED POLARIMETRIC SYNTHETIC APERTURE RADAR IMAGE CLASSIFICATION USING COVARIANCE MATRICES(Technical University of Cluj-Napoca, 2025-06-26) ILEA Ioana; MICLEA-CECALACA Andreia Valentina; CISLARIU Mihaela; MALUTAN Raul; GROSU GeorgeThis paper proposes a workflow for polarimetric SAR (PolSAR) image classification based on statistical texture descriptors. The methodology presented in this paper involves spatial interdependence between neighboring pixels as well as multiscale texture representation using wavelet decomposition. The collected features are modeled by zero-mean Multivariate Gaussian Distributions (MGDs). Then, their estimated covariance matrix acts as the texture descriptor and is employed in a k-nearest neighbors (kNN) classifier. Experiments using real PolSAR data validate the proposed approaches' accuracy in land cover categorization, showing their potential for reliable class identification.Articol MOBILE APPLICATION FOR BLE INDOOR POSITIONING(Technical University of Cluj-Napoca, 2023-12-12) MĂLUȚAN Raul; BRISC BiancaDistance estimation in indoor localization is mainly solved by Received Signal Strength Indicator (RSSI). In our approach, the RSSI is determined empirically, and it’s used to estimate the distance between the user and beacons in a mobile app solution. Trilateration and Kalman filter help obtain better accuracy. The paper proposes a classical solution for indoor localization, but with features like blueprint implementation, Google Maps integration, and manual beacon positioning.Articol SECURITY ISSUES IN INTERNET OF THINGS BOTNETS: A HIGH INTERACTION HONEYPOT APPROACH(Technical University of Cluj-Napoca, 2023-12-11) LAZAR Alexandru; BLAGA Tudor-Mihai; ZINCA Daniel; DOBROTA VirgilFor a better awareness of the tactics employed by the malicious entities in Internet of Things, a system called Honeypot tricks the attackers into exploiting its “sweet” fake resources. While implementing nine types of devices only six of those were attacked. Despite this, the honeypot managed to capture attacks destined for three devices that were not implemented. Overall, several real-world attacks were captured and analyzed providing different indicators of compromise. No new threats were identified but the server only ran for a short period of time and with limited resources. This approach looks like a promising avenue for creating attacker profiles and collecting data on botnet behavior.Articol UHF-RFID ANTENNA FOR SEMI-ACTIVE ASSISTED TAGS(Technical University of Cluj-Napoca – Romania, 2023-12-09) CRIȘAN NicolaeThis paper proposes a new concept of using RFID tag antennas more actively in the identification process. This innovative concept is based on antenna native properties that allow RF harvesting for DC biasing, either from the reader or from other proximity RF sources at the same time. Nevertheless, for now, RFID identification tags without a self-DC power supply have become more popular and are referred to as passive. Eventually, both RF and light harvesting processes could be addressed as a foreseeable solution for identification range extension. Soon this trend will lead to the cheapest (printable) RFID tags at competitive prices closely shifted to those of the already mass-produced barcode tags, mainly in use today. The proposed semi-passive tag antenna is working by replacing the IMPINJ-MONZA chip with the EM4325 chip which can be DC supplied by a PV panel. This tag can work as RF energy harvester and can be DC biased as well, being the most capable for the proposed concept. The antenna is printed on paper using a metal powder ink printer and the tag is interrogated using a UHF-RFID reader. The harvesting for tag DC biasing is extended through a very small solar panel, which acts also as a substrate for the RFID antenna. Thus, the extended concept of the semi-active tag could increase the coverage and the sensing range of the reader significantly. This happens especially in daylight or when other nearby RF sources are available, otherwise, the tag will work as an ordinary one. This approach will allow the antenna to be shrunk even further near the Chu-Harrington limit.Articol SPATIAL-SPECTRAL CLASSIFICATION OF HYPERSPECTRAL DATA WITH CONTROLLED DATA SEPARATION(Technical University of Cluj-Napoca, 2023-06-22) MICLEA Andreia Valentina; ABRUDAN MihaelaExploring spatial-spectral data frequently involves classifying hyperspectral images using convolutional neural networks. Due to the high complexity of the data and the scarcity of available training samples, hyperspectral image classification presents significant difficulties. In the context of supervised classification, we find that traditional experimental designs are frequently misused in the spectral-spatial processing context, resulting in unfair or biased performance evaluation, particularly when training and testing samples are selected at random from the same dataset. Under these circumstances, the dependence caused by the overlap between training and testing samples may be artificially increased, in breach of the data independence assumption upheld by supervised learning theory. In order to prevent an unbiased classification result, we present in this paper a controlled strategy designed to minimize the overlap between the samples present in the training and the testing data sets. The proposed controlled sampling strategy ensures a more trustworthy generalization of the CNN model by minimizing the issues present in the random sampling approach, such as the inability to determine whether or not an increase in classification accuracy is due to the spatial information incorporated into a classifier or to an increase in the overlap between training and testing data sets. Experiments performed with a wavelet CNN on different HSIs, namely Indian Pines, Pavia University, and Salinas, ensure the generalization of the data under the assumption that the training and data sets are independent from one another, based on a controlled strategy. Considering the high dimension of the HSI image, as a pre-processing step, the evaluation of the proposed framework is done by PCA and FA methods.Articol SECURE ACCESS WITH TELTONIKA GPS TRACKING DEVICES FOR INTELLIGENT TRANSPORTATION SYSTEMS(Technical University of Cluj-Napoca, 2023-03-27) ANDREICA Gheorghe-Romeo; STANGU Ciprian; IVANCIU Iustin-Alexandru; ZINCA Daniel; DOBROTA VirgilGPS tracking devices are widely used in industries like logistics, transportation, and security. However, they are susceptible to cyber-attacks, including Man-in-the-Middle (MITM). This study focuses on Teltonika GPS tracking devices and examines the impact of MITM attacks on their operation. We propose implementing encryption protocols and other measures to enhance the security and resilience of Teltonika GPS tracking devices.Articol DDoS ATTACK DETECTION USING SUPERVISED MACHINE LEARNING ALGORITHMS OVER THE CIDDOS2019 DATASET(Technical University of Cluj-Napoca, 2023-06-22) ZINCA Daniel; DOBROTA VirgilDistributed Denial-of-Service (DDoS) attacks are one of the most common types of cyber-attacks that can cause severe damage to networks and systems. Traditional methods to detect them rely on signature-based Intrusion Detection Systems (IDS), which are limited by the need of prior knowledge of specific patterns and by the usual ineffectiveness against zero-day attacks. However machine learning (ML) algorithms have the potential to support the detection of new and unknown attacks. This article compares the DDoS detection performance of three Machine Learning techniques: Gaussian Naïve Bayes, Logistic Regression and Random Forest, based on validation metrics such as precision, recall and F1 score. The system was trained using three datasets extracted from CICDDoS2019 database. The results proved the detection of attacks at Layer 4 (TCP SYN/ UDP flood), and at reflective Layer 7 (MSSQL, NetBIOS). The Random Forests and Logistic Regression methods achieved a precision between 93.7% and 99.4 % over these three datasets.Articol A NUMERICAL MODEL FOR I/Q MODULATION/DEMODULATION FOR AN FPGA-ENABLED SDR PLATFORM(Technical University of Cluj-Napoca, 2023-06-21) BOTA Bianca L.M. ; BUTA Rareș ; KIREI Botond S. ; FĂRCAȘ Calin A. ; HINTEA Sorin A.This paper presents a numerical model for I/Q modulation/demodulation implemented in Matlab. This numerical model serves as a starting point for the HDL/FPGA implementation of a baseband processor for a previously developed SDR platform. The numerical model verification is carried out by performing 4, 8, 64, and 256-point QAM transmission over an additive White Gaussian Noise Channel and by comparing numerical and theoretical bit error rates. A partial HDL implementation of the baseband processor is given: the baseband processor generates a 1MHz tone signal that is mixed by the RF fronted, generating a double sideband modulated signal.