TELKOMNIKA (Telecommunication Computing Electronics and Control)
Not a member yet
3120 research outputs found
Sort by
A compact five-band patch antenna covering WLAN, WiMAX, X, and Ku-bands
Full text linkThis paper introduces a compact, low-profile, five-band antenna for wireless communication systems operating across 2.4/5 GHz WLAN, 3.5 GHz WiMax, and 7.5 GHz X- and Ku-band frequencies. The proposed antenna utilizes a novel configuration with a dual-overlapping rectangular patch coupled to a wide circular slot and an inverted L-shaped strip. Fabricated on a single-layer FR4 substrate (εr=4.3, thickness=1.6 mm), the antenna employs a 50-ohm coplanar waveguide feed, resulting in a compact footprint of 40×40×1.6 mm³. Experimental measurements and simulations demonstrate a bidirectional radiation pattern covering five distinct frequency bands: 2.4-2.485 GHz, 3.4-3.6 GHz, 5.15-5.825 GHz, 7.25-8.4 GHz, and 13.4-17.7 GHz. The antenna exhibits a return loss better than 10 dB across all bands and provides average gains of 1.78 dBi, 3.04 dBi, 3.4 dBi, 4.27 dBi, and 4.46 dBi, respectively. These results confirm the successful development of a five-band antenna with excellent performance characteristics, making it a promising candidate for 2.4/5 GHz WLAN, WiMax, and X- and Ku-band satellite communications, with automotive vehicles being one example
High-speed dividing device with the formation of quotient and remainder
Full text linkConsidered the possibility of accelerating the time-critical operation of division for multi-bit integers. This problem is significant since, so multi-bit integers are widely used in specialized devices, including cryptographic transformations. A method for high-speed quotient and remainder determination with optimal hardware costs is proposed. A preliminary increase in the divisor and its subsequent decrease by shifting it to the right are used. A structural diagram and functional diagram of the hardware implementation have been developed using high-speed combinational logic circuits. The device’s principle of operation, its step-by-step process, and specific examples illustrating its correct operation and resource efficiency are addressed. On average, it takes (k/2+1) clock cycles to obtain the result, where (k+1) bit capacity of the quotient. In most division schemes with optimal hardware costs, the number of clock cycles required to obtain the quotient (without remainder) is (k+1). High-speed division with simultaneous determination of several quotient bits requires (m/p) clock cycles for the division operation, where p- the number of simultaneously determined quotient bits, m-bit capacity of dividend. However, this approach will require additional hardware. The research will continue by modeling the device in Vivado Design Suite computer aided design (CAD) based on Artix-7 field programmable gate array (FPGA) from Xilinx
Energy scavenging-aided NOMA uplink communications: performance analysis
Full text linkEnergy scavenging-aided nonorthogonal multiple access (NOMA) networks significantly ameliorate energy-and-spectral efficiencies thanks to superimposing a multitude of user signals for concurrent transmission and harvesting radio frequency energy. Practically, energy harvesters possess non-linear characteristic and their efficiency is enhanced considerably with deployment of multiple antennas. Moreover, communication reliability and harvested energy are directly influenced by wireless propagation which induces simultaneous effects of shadowing, path loss, and fading. Accordingly, the current paper assesses analytically outage probability and throughput of energy scavenging (ES)-aided NOMA uplink communications (eNOMAu) taking into account the above-addressed realistic factors (κ − µ shadowed fading, multi-antenna deployment, ES nonlinearity). The results reveal considerable performance degradation caused by ES non-linearity and wireless propagation. Additionally, desired system performance can be reached flexibly with appropriate specification selection. In addition, accreting a quantity of antennas drastically mitigates the outage probability of eNOMAu, which can be minimized with optimal ES time selection. Furthermore, the proposed eNOMAu is considerably superior to its eOMAu counterpart
Hybrid Kolmogorov-Arnold and convolutional neural network model for single-lead electrocardiogram classification
Full text linkThis study proposes a hybrid Kolmogorov-Arnold networks (KANs) and convolutional neural networks (CNN) to classify electrocardiogram (ECG) signal abnormalities in one lead ECG data of wearable telemedicine. The hybrid model combines CNN to extract hierarchical features from sequential data and KANs to model non-linear relationships with fewer parameters as an efficient classification. The study explores the model’s capacity to balance accuracy, computational efficiency, and memory usage as critical factors for real-time health monitoring in resource-constrained environments on the single-lead MIT-Beth Israel hospital (MIT-BIH) Supraventricular Arrhythmia database with five different class labels. For comparison, standalone CNN and KAN models were also trained on the same balanced dataset. The CNN model achieved an accuracy of 96.62%, precision of 96.81%, and recall of 96.53%. The KAN model, while computationally efficient, performed less effectively, with an accuracy of 94.15%, precision of 95.01%, and recall of 92.57%. In contrast, our hybrid KAN-CNN model outperformed both, attaining an accuracy of 97.53%, precision of 97.66%, recall of 97.40%, and a low loss of 0.0840. The study also explores the impact of quantization and compression on model performance, revealing that both CNN and Hybrid KAN-CNN models retained high accuracy post-quantization, whereas the KAN model exhibited a more significant drop in performance
An extensive framework for assessing the quality of websites
Full text linkThe quality of the website is quite important in generating customer satisfaction and loyalty. A website’s quality depends on several factors, features, and characteristics. Several computational methods are necessary to evaluate the quality of each factor and subsequently determine the overall quality of the entire website. Each factor does not contribute to the same level of quality required by the end users and thus requires a weighting system. Expert systems, which are either manually defined or learnt using artificial intelligence (AI), are to be modelled for assessing the quality of a factor/sub-factor or characteristics of a sub-factor. The quality of a website varies depending on the context. Context-based quality assessment of the websites is required. There is a need to generate example sets to assess the quality of websites and to establish relationships between web-related quality factors, subfactors, and characteristics. In this paper, a comprehensive framework is presented that caters to parametric structure building and mapping, parsers for computing characteristic values, context assessment, building expert systems, and learning models for assessing the quality of websites and weighing the factors that have specific significance on the quality of the website
Outlier detection and clustering of fifth-generation wireless channel model datasets
Full text linkThe fifth-generation (5G) wireless communications system offers faster data rates, lower latency, and more interconnecting devices. Various 5G channel models were developed to study its stochastic characteristics before implementation. These channel models generate multipath components that are grouped into clusters. The multipath clusters serve as datasets in multipath clustering. The clustering results are then used to examine the propagation properties of the 5G system. However, datasets are prone to outliers. They tend to affect clustering accuracy. Hence, this study clusters the datasets generated by the channel models using five clustering approaches, removes the outliers using mean-shift outlier detection, and clusters the datasets free of outliers again using the same clustering algorithms. Outlier detection shows that 5G channel model datasets contain noise, and outlier removal improves the modeling characteristics, as demonstrated by enhanced clustering accuracy. Results show that most of the outliers are detected in the 2×SD threshold. The removal of the outliers using the said threshold increased the clustering accuracy of K-means and AC-Single in Semi-Urban B1 LOS multiple links by 78.85% and 55%, respectively, and DBSCAN in Semi-Urban B2 LOS multiple links by 57.14%. Outlier detection and removal also work well with 5G channel model datasets
A neuro-game model for analyzing strategies in the dynamic interaction of participants of phishing attacks
Full text linkA dynamic model of countering phishing attacks is considered. Cryptocurrency exchanges (CCE) and/or their clients are considered as an example of a phishing victim. The model, unlike similar ones, is based on the assumption that the dynamics of the states of the player-victim of phishing attacks and the player-intruder (fisher) is set by means of a system of differential equations. The peculiarity of this model is that it represents a bilinear differential game of quality, for which methods for solving linear differential games are not applicable and, in addition, the absence of functional restrictions on the strategies of players (even immeasurable functions are allowed) does not allow the use of traditional approach. And their solution makes it possible to form payoff matrices, which are part of the training set for artificial neural networks (ANNs). Such a collaboration of models will make it possible to accurately build an anti-phishing strategy, minimizing the costs of both a potential victim of phishing attacks and the defense side when building a secure system of communication with CCE clients. The neuro-game approach makes it possible to predict the process of countering phishing in the context of costs for both parties using different strategies
Proposing a new method for calculating DC sources in an extended multilevel converter
Full text linkIn this paper, we propose a method for calculating the DC source amplitude in an extended multilevel inverter (MLI) structure so that the maximum number of levels and the output voltage waveform are as close as possible to the sinusoidal wave with minimum total harmonic distortion (THD). For the developed structure, three algorithms are suggested to determine the amounts of DC voltage resources. The first important point about choosing the right amounts for the DC resources is that the number of levels should be as large as possible, and the second important point is that the intervals between the levels should be the same throughout the waveform. By observing these two points, the output voltage waveform can be as near as possible to the sinusoidal wave that we want. In this study, we used iteration-based methods to find suitable values for DC sources. Simulation results are offered to confirm the capability of the extended multilevel converter. After we solved the problem through calculation and analysis, a code was written in MATLAB with the aim that this time the code will tell us for what values of DC sources we will have the largest number of levels and as we expected, the output of the MATLAB code confirmed the correctness of the calculations
Face recognition for smart door security access with convolutional neural network method
Full text linkThis study focuses on enhancing office security through a smart door system, designed to protect sensitive documents and critical data. Emphasizing exclusive access for authorized personnel, the system integrates advanced biometric authentication, predominantly facial recognition. The project's aim is to optimize face recognition using convolutional neural network (CNN) techniques, identifying the best preprocessing methods and hyperparameter settings. A significant aspect of the research involves developing a smart door system with remote authentication and control capabilities via internet connectivity. Employing transfer learning with MobileNet V2, the study presents a compact model tailored for the Raspberry Pi platform. The model utilizes a dataset with five facial recognition classes and an additional class for unknown faces, ensuring a diverse representation. The trained model achieved a high accuracy (0.9729) and low loss (0.09). System evaluation revealed an overall accuracy of 0.96, perfect recall (1.00), and a precision of 0.897. These results demonstrate the system's efficacy in secure access control, making it a viable solution for contemporary office environment
Moving-horizon estimation approach for nonlinear systems with measurement contaminated by outliers
Full text linkAn application of moving-horizon strategy for nonlinear systems with possible outliers in measurements is addressed. With the increased success of movinghorizon strategy in the state estimation for linear systems with outliers acting on the measurement, investigating the nonlinear approach is highly required. In this paper we applied the nonlinear version which has been presented in the literature in term of discrete-time linear time-invariant systems, where the applied strategy considers minimizing a least-squares functions in which each measure possibly contaminated by outlier is left out in turn and the lowest cost is propagated. The moving horizon filter effectiveness as compared with the extended Kalman filter is shown by means of simulation example and estimation error comparison. The moving horizon filter shows the feature of resisting outliers with robust estimatio