Research Article
Exploring the Future of Optical Fiber Communications Technologies and Applications
Abdullah Al Mamun Sadi, Mst Karima Khatun, Mondol Md Shayokh
Middle East Research Journal of Engineering and Technology; 142-149.
https://doi.org/10.36348/merjet.2025.v05i06.001
Abstract: Optical fiber communication (OFC) has entirely transformed our communication in that it enables data to be carried at lightning speeds over a long distance to the extent of excellent bandwidth and at low losses. The world is becoming more interdependent and therefore there is the need to have quick and more dependable communication systems. OFC is currently an important component of fulfilling these growing needs that enable businesses, consumers and emerging technologies to remain connected. The paper will examine the future of optical fiber communications which includes the innovations such as the Dense Wavelength Division Multiplexing (DWDM) and the Quantum Key Distribution (QKD) that will enhance the capacity and security of the network. The study is holistically written considering the integration of theory, mathematical modeling, and simulations to evaluate the performance of optical fiber systems in various circumstances. The study considers issues such as signal loss, fiber behavior and network capacity by applying techniques such as the Shannon-Hartley law and Beer-Lambert law. The findings indicate how DWDM can aid efficiency by enabling several streams of data to run through the same fiber, to help in reducing the increasing volume of data across the world. The paper also delves into the prospects of the Quantum Key Distribution (QKD) which is a game-changer in the field of data security. QKD provides privacy to data being transmitted using optical fiber networks by applying quantum mechanics. The research indicates that QKD can provide defense against rising cyber threats, and therefore, it is a fundamental component of the optical network in the future, particularly as the importance of secure communication grows. In addition to security, the paper discusses the ways in which optical fiber communication is facilitating the technologies of the future such as 5G, 6G and Internet of Things (IoT), which demand high-speed, low-latency, and reliable communication infrastructure. With more need of data than ever, OFC is well placed to meet those needs, it is the backbone of the innovation in the telecommunications, healthcare, smart cities and industrial automation industries. The importance of Fiber-to-the-Home (FTTH) networks is highlighted in the study, which is necessary to provide the high-speed internet to the households to facilitate the development of entertainment, education, telemedicine, and remote work. There are still challenges. Such problems as the degradation of the signal at long distances, the high cost of the system implementation of the fiber network (in rural areas in particular), the environmental consequences of fiber production, and disposal, and the necessity to provide more effective security need to be considered. Integration of optical fiber networks with 5G, 6G and IoT has also been mentioned in the paper as it presents difficulties with compatibility, scalability and integration of disparate technologies. The study also highlights how the optical fiber industry should be more environmentally friendly, in terms of material usage, recycling, and more healthy production processes. Also, the lack of skilled labor in the optical fiber sector will act as a bottleneck to the growth and sustenance of these networks and it is therefore important to make investments in human capital.
Research Article
Automated Liver Segmentation from CT Images Using Deep Residual Networks
Rokshana Akter Jhilik, Md Rakibul Islam, Sokoleshar Chandra Sarker
Middle East Research Journal of Engineering and Technology; 150-157.
https://doi.org/10.36348/merjet.2025.v05i06.002
Computer-aided diagnosis, surgical planning, and volumetric assessment involve accurate and automated liver segmentation of the computed tomography (CT) images as a crucial step. Radiologist annotation is time-intensive and subject to inter-observer variability, which encourages the creation of high-quality deep learning-based prediction methods. In this paper, we introduce a U-Net model that has been developed with ResNet-50 encoder to segment liver automatically. It has been shown that the model utilizes transfer learning to capture fine hierarchical features without compromising the spatial detail using encoder-decoder skip connections. The trained and tested method was provided using the LiTS17 dataset containing 131 volumetric CT scans of varying anatomies and intensities. The experimental findings show that the Dice Similarity Coefficient is 0.951, the Intersection over Union is 0.917 and the accuracy of pixels is 0.973 on the test set which confirms good generalization and high boundary accuracy. When compared to the current frameworks like PADLLS and GCHA-Net, the suggested ResNet-50 U-Net has competitive accuracy with less computational complexity. These results show that the model is an effective and clinically relevant approach to liver segmentation, which can be implemented in real-time radiological processes and preoperative planning systems.
This work is licensed under a Creative Commons
Attribution-NonCommercial 4.0 International License.
© Copyright Kuwait Scholars Publisher. All Rights Reserved.