Early detection of earthquakes is essential for minimizing potential damage and ensuring public safety. Recent advancements in deep learning, particularly convolutional neural networks (CNNs), provide a promising alternative for analyzing seismic waves. In contrast, traditional methods such as the short-term average/long-term average (STA/LTA) algorithm and the Akaike information criterion (AIC) have limitations in detecting primary (P) waves in high-noise conditions, caused by industrial and anthropogenic disturbances. This study presents a CNN-based automatic P-wave detection model tailored for the Almaty city region. The seismic dataset used in this research was obtained from the IRIS database and includes data collected from seven stations within a 333 km radius of Almaty, Kazakhstan. The proposed model achieves a recall rate of 89.1% and an accuracy of 94.1% compared to other deep learning-based models. Experimental results demonstrate that this method enhances the reliability of automatic early earthquake warning systems and improves the accuracy of P-wave detection. The research outputs presented for the local region are unique. Applying CNNs in seismic monitoring facilitates the development of efficient automated systems that minimize risks and improve response measures for natural disasters.
26 June 2024 in Eurasian Physical Technical Journal
Wireless Mesh Networks are gaining traction as a solution for delivering reliable connectivity without centralized infrastructure. They operate through wireless node interconnections, forming self-configuring networks ideal for scenarios where wired networks are impractical. Routing is crucial in Wireless Mesh Networks to ensure efficient communication among nodes. However, the suitability of routing algorithms for Wireless Mesh Network’s topology requires further investigation. This paper proposes an investigation into the effectiveness of routing algorithms like AODV, DSDV, and OLSR across various Wireless Mesh Networks topologies using NS-3 simulation. It also aims to determine the optimal number of nodes and protocols to maximize throughput and minimize packet loss within a limited area. Through rigorous NS-3 simulations, the study demonstrates that AODV, DSDV, and OLSR exhibit differing effectiveness across random, mesh grid, and Fruchterman-Reingold topologies. These results emphasize the importance of considering topology-specific factors when selecting and optimizing routing protocols for Wireless Mesh Networks. In summary, Wireless Mesh Networks offer decentralized connectivity, but the effectiveness of routing algorithms in different topologies remains understudied. This investigation addresses this gap by evaluating routing algorithms across various topologies, shedding light on their suitability and performance in Wireless Mesh Networks.
This work aims to develop approaches to processing and interpreting spectral entropy outcomes in the context of seismic data, as well as to establish a methodological foundation for subsequent integration into practical monitoring solutions. The objective of this study is to evaluate the effectiveness of the Shannon spectral entropy method in detecting and assessing short-term seismic events through a seismogram analysis. This method has demonstrated sensitivity to variations in the spectral characteristics of the registered signals. A threshold value for the increase in spectral entropy information has been pinpointed for reliable P-wave detection. The results could be applied in real-time automated seismic monitoring systems. In addition to the conventional spectral analysis techniques, the proposed methodology may serve as the input to the neural network models used in seismological applications.
This paper describes a cryptographic protection system hardware device designed to improve data encryption and decryption performance and preserve data integrity. The cryptosystem is implemented by hardware-software method, where the encryption and decryption of data are carried out in a stand-alone FPGA device based on non-positional polynomial number system (NPN). For data encryption the next block of text to be encrypted is divided into sub-blocks and represented as separate binary polynomials and binary polynomials-keys are assigned to them, as well as irreducible polynomials (modules). Then, split blocks are calculated in parallel and a ciphertext is formed. For this purpose, the special algorithm where calculation of NPN parameters and check on irreducibility of polynomials (modules) and the program of generation of direct and inverse keys are developed and application functional is developed that implements operations in the ring of polynomials with coefficients GF(2) using an object-oriented approach. We have developed polynomial multipliers modulo sequential and parallel action (matrix multiplier) on the basis of which data encryption and decryption are performed.
This study presents an efficient method for the automatic detection of primary (P) seismic waves, based on time-frequency analysis and spectrogram classification using the YOLO11x-cls deep learning model. Unlike traditional detection algorithms that rely on handcrafted feature extraction and domain-specific heuristics, the proposed approach directly processes visualized seismic signals generated via the continuous Morlet wavelet transform. The main output of the method is the precise identification of P-wave arrival times, which are critical for earthquake early warning systems.To train the model, a balanced dataset comprising 1700 labeled spectrogram images was prepared using seismic data from a single station, TLG, located near the city of Almaty. To evaluate the generalization capability of the model, testing was conducted on datasets acquired from three independent seismic stations situated within approximately 333 km of Almaty (two in Kyrgyzstan and one in Kazakhstan), as well as on recordings of global high-magnitude earthquakes (M ≥ 7.0) that occurred worldwide during the period from January 1, 2023 to December 31, 2024. The experimental results demonstrated the high effectiveness of the proposed method. On the independent seismic stations TARG and JNKS (Kyrgyzstan), as well as SHLS (Kazakhstan), the model achieved F1-scores ranging from 91% to 95%. In the combined evaluation across all three stations, the model yielded a total of 564 true positives, 40 false positives, and 35 false negatives. The aggregated performance metrics were as follows: precision – 93.4%, recall – 94.2%, and F1-score – 93.8%, confirming the model’s high accuracy and robustness in detecting P-waves on datasets that differ from the training distribution. Additionally, on recordings of global high-magnitude earthquakes, the model achieved an overall accuracy of 93.5%, further highlighting its reliability and applicability under conditions of worldwide seismic activity.
Unmanned Aerial Vehicles (UAVs) are increasingly employed for real-time object detection in critical appli cations such as security surveillance, disaster response, and environmental monitoring. However, accurate detection in UAV imagery remains challenging due to small target sizes, cluttered backgrounds, and varying environmental conditions. This study evaluates the performance of YOLOv8n/v8s and YOLOv11n/11s models for human detection in UAV-captured imagery across diverse natural landscapes. To ensure practi cal deployment in resource-constrained environments, the models were implemented on a Raspberry Pi 5 using the OpenVINO framework. Experimental results show that both YOLO series achieve comparable detection accuracy in the range of 80–82%, with YOLOv8n and YOLOv11n demonstrating the highest processing speeds of 10.50 and 11.04 frames per second (FPS), respectively. These findings confirm the feasibility of using lightweight YOLO models for real-time human detection on embedded systems. The re sults highlight the potential of integrating edge AI and UAVs for autonomous, on-site monitoring in remote or complex terrains, offering scalable solutions for intelligent aerial surveillance. Key words: UAV, object detection, accuracy, YOLO models.