2025. 1^st Issue

Volume XVII, Number 1

Table of contents 

Full issue  

 

MESSAGE FROM THE EDITOR-IN-CHIEF

Pal Varga
Fr  
Let's 

Download

 

PAPERS FROM OPEN CALL

Zoltán Belső, and László Pap
Effect of the Imperfect Channel Estimation on Achievable NOMA Rate 
In recent years, Non-orthogonal Multiple Access (NOMA) has been proposed as an alternative to the more traditional Orthogonal Multiple Access (OMA) schemes for mobile communication. In the NOMA method, the resource domains (like power and bandwidth) are not split but shared between the users of the network. The non-orthogonality means that there is cross-talk between the signals of different users, and the interference is either cancelled by a method called successive interference cancellation (SIC) or treated as part of the noise. Comparing the achievable capacity region of OMA and NOMA schemes show that NOMA has advantage over OMA. The SIC method requires knowledge of the channel characteristic between the base station and the user. In the ideal case where all the channel conditions are precisely known, NOMA always performs better than or equal to OMA. In real application, the channel characteristic can only be estimated, which can be nonperfect. In this paper, we will examine the effect of non-perfect channel estimation on the performance of NOMA and will find that in some cases, NOMA still perform better than OMA, but in other cases OMA would perform better.

DOI: 10.36244/ICJ.2025.1.1
Download 

 

Arun Kumar, and Aziz Nanthaamornphon
Reducing the Peak to Average Power Ratio in Optical NOMA Waveform Using Airy-Special Function based PTS Algorithm 
This paper introduces a novel Peak-to-Average Power Ratio (PAPR) reduction technique for Non-Orthogonal Multiple Access (NOMA) waveforms, leveraging an Airy function-based Partial Transmit Sequence (PTS) method. The proposed technique is evaluated on NOMA waveforms with subcarrier configurations of 64, 256, and 512, and its performance is benchmarked against conventional PTS, Selective Mapping (SLM), and Clipping and Filtering methods. Comprehensive analysis is conducted on key metrics, including PAPR, Bit Error Rate (BER), and Power Spectral Density (PSD). Results demonstrate that the Airy-based PTS method achieves substantial PAPR reduction across all subcarrier scenarios, consistently surpassing traditional approaches. Furthermore, the proposed method maintains competitive BER performance, particularly in high subcarrier scenarios, where conventional methods typically face limitations. PSD analysis further highlights the spectral efficiency of the Airy-based PTS method, exhibiting minimal out-of-band emissions. These findings position the Airybased PTS technique as a promising solution for improving NOMA waveform performance in 5G and beyond, achieving an optimal balance between PAPR reduction, BER, and spectral efficiency.

DOI: 10.36244/ICJ.2025.1.2
Download 

 

Amalia Eka Rakhmania, Hudiono, Umi Anis Ro’isatin, and Nurul Hidayati 
Channel Estimation Methods in Massive MIMO: A Comparative Review of Machine Learning and Traditional Techniques 
Massive Multiple Input Multiple Output (MIMO) has emerged as a crucial technology in 5G and future 6G networks, offering unprecedented improvements in capacity, energy efficiency, and spectral efficiency. A key challenge for Massive MIMO systems is accurate and efficient channel estimation, which significantly impacts system performance. Traditional channel estimation methods such as Least Squares (LS) and Minimum Mean Square Error (MMSE) have been widely employed, but their limitations, particularly in complex and dynamic environments, have led to the exploration of more sophisticated approaches, including machine learning (ML)-based techniques. This review aims to compare traditional channel estimation methods with modern machine learning-based techniques in Massive MIMO systems, providing insights into their performance, computational complexity, and scalability. Furthermore, this paper outlines potential future research directions, emphasizing the integration of machine learning, optimization techniques, and hardware-friendly design for enhanced performance. 

DOI: 10.36244/ICJ.2025.1.3
Download 

 

 

Lajos Nagy
Horn Antenna Development at 80 GHz for Tank  Level Probing Radar Applications 
One important application and research area of radar technology is tank-level measurement and detection. Radar contactless level measurement is a safe solution even in extreme process conditions, such as significant overpressure, high temperatures, and the presence of corrosive vapors. The main categories of these principles are ultrasonic, and electromagnetic wave radars. We will now consider only radars using electromagnetic waves. The use of millimeter radio waves, which we use, is nowadays becoming more and more common also for automotive radars, human presence detection and human vital signs. To meet electromagnetic requirements such as high gain, low spurious levels, and high bandwidth, special antennas are required. The low sidelobe level and narrow main beam mainly reduce reflections from the side of the tank while the bandwidth determines the distance resolution of the measurement system. A further requirement is the small size and reliable manufacturability of the antenna. In the presence of corrosive vapors, antennas must be resistant to corrosion. The article briefly summarizes the material parameter measurements required for the design of the radar, and the design of the main components. We analyzed the possible dielectric materials that can be used as random or dielectric lenses for such antennas. In the next part of the paper, we present a conical horn antenna design for the 80 GHz band and compare the parameters of an open horn antenna with those of a horn antenna with a dielectric lens. Finally, a tank-level radar designed with the Texas Instruments IWR1443 radar chip is presented.

DOI: 10.36244/ICJ.2025.1.4
Download 

 

 

Botond Tamás Csathó, and Bálint Péter Horváth
Physically Tenable Analysis and Control of Scattering from Reconfigurable Intelligent Surfaces 
Reconfigurable intelligent surface is a promising concept within the scope of smart radio environment, which is a key enabler of the future wireless networks. Efficient numerical modeling of such devices constitutes a fundamental and actively pursued research challenge. This study numerically analyzes the reflection properties of a particular reconfigurable intelligent surface with the aid of computational electromagnetics. A key advantage of utilizing full-wave simulations is that they capture all the physical phenomena within the structure, thus providing a physically stenable analysis method. An essential aspect of RIS modeling is the configuration pattern design of the surfaces. A standard objective function of the pattern design is the amount of energy reradiated toward the target direction. The utilization of full-wave simulations limits the applicable optimization methods. In this article, an intuition-based pattern search method is presented to design RIS configurations, with the radiation pattern of the RIS structure in free space as the objective function. The suggested method first identifies a set of configuration values, then exhaustively searches through their combinations, seeking for the highest anomalously reflected power. The first presented result is the demonstration of creating anomalous reflections, with the dominant reflection being electrically tunable. The second contribution is the aforementioned pattern search method, which enables the reradiation of the incident energy for numerous anomalous directions. The average scattering parameter amplitude for the scenario is 0.78. Finally, we also demonstrate the effect of the structure being finite in size. We conclude that the dominant radiation directions coincide with the modes of the infinite periodic counterpart.

DOI: 10.36244/ICJ.2025.1.5
Download 

 

 

Shahad A. Hussein, Suadad S. Mahdi and Alharith A. Abdullah
Quantum Network Security: A Quantum Firewall Approach  
The increasing prominence of quantum networks has necessitated the exploration of their vulnerabilities and the development of effective countermeasures. This paper investigates the potential threats faced by quantum networks, particularly focusing on the exploitation of quantum TCP-threeway handshake connections. To mitigate these attacks, a novel approach involving the implementation of a quantum firewall is proposed. The paper emphasizes that the security of quantum networks is primarily reliant on pre-established agreements for creating quantum entanglement among devices, which inherently limits external attacks. However, it highlights the adverse impact of quantum assaults on network availability due to the consumption of quantum bits required for establishing connections. By leveraging unique node identification and coherence time of quantum memory, the proposed quantum firewall effectively mitigates the effects of attacks while ensuring network availability. Through this security strategy, the paper demonstrates the robustness of the quantum firewall in safeguarding the integrity and operation of quantum networks against potential threats. 

DOI: 10.36244/ICJ.2025.1.6
Download 

 

 

Ákos Leiter, Döme Matusovits, and László Bokor
Evaluation of traditional and eBPF-based packet processing in Kubernetes for network slicing 
In recent years, the proliferation of cloud-native technology enablers, such as microservice deployment and management with Kubernetes, have presented new challenges for telecommunications service providers. Strict data transmission requirements have emerged in various areas, such as immediate interventions in intelligent transportation, video conferencing, etc. With the advent of 5G networks, this demand can also be fulfilled thanks to an innovative technology called Network Slicing. In terms of its operation, we can separate networks into individual segments to continuously satisfy the desired service requirements. However, packet processing on top of Kubernetes may need to be changed to support the emerging number of microservices during slicing. This is where the Extended Berkeley Packet Filter (eBPF) comes into the picture to boost the capacity of data centers and keep service guarantees. This paper presents how eBPF can support network slicing through its performance evaluation in a Kubernetes environment.

DOI: 10.36244/ICJ.2025.1.7
Download 

 

 

Mohammed Jabardi 
Support Vector Machines: Theory, Algorithms, and Applications  
Support Vector Machines, or SVMs, are a strong group of supervised learning models that are commonly used for tasks like regression and classification. SVMs are based on the theory of statistical learning and try to find the best hyperplane that maximizes the gap between different classes. This makes it easier to apply to new data. Since kernel functions are used with SVMs, they are more flexible and can handle both linear and nonlinear situations well. Even though they have a strong theoretical base, they still face problems in the real world, like being hard to code and difficult to tune parameters, especially for big datasets. Recent improvements, like scalable solvers and estimated kernel methods, have made them a lot more useful. This essay talks about SVM theory, its main algorithms, and how it is used in the real world. It shows how it is used in bioinformatics, banking, and image processing, among other areas.

DOI: 10.36244/ICJ.2025.1.8
Download 

 

 

Attila Zoltán Jenei, Réka Ágoston, and István Valálik
A Siamese-based Approach to Improve Parkinson’s Disease Detection and Severity Prediction from  Speech Using X-Vector Embedding 
Parkinson’s disease is incurable and is considered one of the most common neurological diseases. It is a progressive disease, which highlights the importance of early detection. Machine learning-based diagnostic support is desirable since the diagnosis is based on history, visual inspection, and drug tests. Speech is presumed to be one of the promising biomarkers that can predict the state of the disease. Combining speech data with deep learning feature extraction in Siamese-based architecture may improve the detection compared with direct regression with acoustic and prosodic features. Read text-based speech samples were acquired from 98 patients with Parkinson’s disease and 107 healthy participants. Feature vectors were extracted with pre- trained x-vector embedding and were used directly with a support vector regressor in a nested cross-validation setup (baseline approach). Furthermore, pairs were allocated, and difference vectors were calculated. These difference vectors were then used to train support vector regressor models in nested crossvalidation (Siamese-based approach). Severity predictions and classification were performed with the outcomes. The Siamesebased setup outperformed the baseline approach both in regression and classification metrics. The relative improvement in root mean square error is 14.4%, and the Pearson correlation is 12.5% at best. After the classification, the relative improvement is 6.0% in sensitivity, 3.0% in specificity, and 4.5% in accuracy. Furthermore, comparing the test sample to not only one but multiple others decreases the average standard deviation of the predicted severity by 16.5% in relative value. Changing only the architecture of the traditional examination setup to a Siamesebased approach may increase the performance of the models.

DOI: 10.36244/ICJ.2025.1.9
Download