This paper aims to evaluate detection algorithms for perimeter security systems based on phase-sensitive optical time-domain reflectometry ($\Phi$-OTDR). Our own designed and developed sensor system was used for the measurement. The main application of the system is in the area the perimeter fencing intrusion detection. The system is unique thanks to the developed motherboard, which contains a field-programmable gate array (FPGA) that takes care of signal processing. This allows the entire system to be integrated into a 1U rack chassis. A polygon containing two different fence types and also cable laid underground in a plastic tube was used for testing. Edge detection algorithms using the Sobel and Prewitt operators are considered for post-processing. The comparison is made based on the signal-to-noise ratio (SNR) values calculated for each event. Results of algorithms based on edge detection methods are compared with the conventional differential method commonly used in $\Phi$-OTDR systems.

Fiber optic infrastructure is essential in the transmission of data of all kinds, both for the long haul and shorter distances in cities. Optical fibers are also preferred for data infrastructures inside buildings, especially in highly secured organizations and government facilities. This paper focuses on a reference measurement and analysis of optical fiber cables sensitivity to acoustic waves. Measurement was carried out in an anechoic chamber to ensure stable conditions of acoustic pressure in the range from 20 Hz to 20 kHz. The frequency response, the signal-to-noise ratio per frequency, and the Speech Transmission Index are evaluated for various types of optical fiber cables and different ceiling tiles, followed by their comparison. The influence of the means of fixing the cable is also studied. The results prove that optical fiber-based infrastructure in buildings can be exploited as a sensitive microphone.

Distributed Acoustic Sensing (DAS) technology leverages optical fibers to detect acoustic signals over long distances, offering high-resolution data critical for applications such as seismic monitoring, structural health monitoring, and security. A significant challenge in DAS systems is the accurate classification of detected events, which is crucial for their reliability. Traditional signal processing methods often struggle with the high-dimensional, noisy data produced by DAS systems, making advanced machine learning techniques essential for improved event classification. However, the lack of large, high-quality datasets has hindered progress. In this study, we present a comprehensive labeled dataset of DAS measurements collected around a university campus, featuring events such as walking, running, and vehicular movement, as well as potential security threats. This dataset provides a valuable resource for developing and validating machine learning models, enabling more accurate and automated event classification. The quality of the dataset is demonstrated through the successful training of a Convolutional Neural Network (CNN).

Fiber-optic network infrastructures are crucial for the transmission of data over long and short distances. Fiber optics are also preferred for the infrastructure of in-building data communications. In this study, we use polarization analysis to ensure the security of the optical fiber/cables of the physical layer. This method exploits the changes induced by mechanical vibrations to polarization states, which can be easily detected using a polarization beam splitter and a balancing photodetector. We use machine learning to classify selected events that violate the safety of the physical layer, such as manipulation or temporary disconnection of connectors. The results show the resting state can be accurately distinguished from selected security breaches for a fiber route subjected to environmental disturbances, where individual events can be classified with nearly 99% accuracy.

This paper presents an advanced perimeter protection system that integrates phase-sensitive Optical Time-Domain Reflectometry (Φ-OTDR) with Convolutional Neural Networks (CNNs) for real-time event classification near optical fibers. The proposed approach enhances traditional security methods by providing robust monitoring in challenging environments, such as low visibility and large-scale areas. We evaluated multiple signal preprocessing techniques, including Fast Fourier Transform (FFT), Redundant Discrete Fourier Transform (RDFT), Discrete Wavelet Transform (DWT), and Mel-Frequency Cepstral Coefficients (MFCC), to optimize classification accuracy and computational efficiency. While MFCC achieved the highest accuracy (85.61 %), RDFT provided the best balance between performance (85.47 %) and real-time feasibility, making it the preferred method for deployment. The system successfully differentiates events such as vehicle movement, fence manipulation, and construction work, while anomaly detection capabilities further enhance security by identifying irregular activities with minimal error. These findings demonstrate the potential of integrating fiber-optic sensing with deep learning to develop scalable, real-time perimeter protection solutions for critical infrastructure, border surveillance, and urban security.