TELKOMNIKA (Telecommunication Computing Electronics and Control)
Not a member yet
3120 research outputs found
Sort by
Neuromarketing case study: recognition of sweet and sour taste in beverage products based on EEG signal features
Full text linkConsumers’ acceptance of a food product hinges on its taste. Culinary practitioners typically conduct organoleptic tests to evaluate a food/beverage’s taste. Organoleptic tests have a subjective nature, making a clear description difficult. In this study, we suggest implementing a brain signal-based electroencephalogram (EEG) taste assessment system to evaluate consumer responses to the tastes of a drink, specifically sour and sweet. The system distinguishes flavors based on EEG data. These classifiers, including recurrent neural network (RNN), long-short term memory (LSTM), and gated recurrent unit (GRU), are utilized for the classification process. Total 35 participants’ EEG data were recorded for this study. Temporal (T3 and T4) and centro parietal (CP1 and CP2) channels are used for recording. EEG signal processing involves filtering, artefact elimination, and band decomposition into delta, theta, alpha, beta, and gamma frequencies. In the time domain of clean EEG data, mean absolute value, standard deviation, and variance are used for signal feature extraction. Several classifiers (RNN, LSTM, and GRU) will be fed with the signal feature values as input. An accuracy of 88.62% was achieved using LSTM in the classification. The RNN and GRU models achieved classification accuracies of 88.56% and 87.15% respectively
Jacobian approximation of the Sum-Alpha stopping criterion
Full text linkThis article will report the development of new application of the SumAlpha stopping criterion to the case of log – maximum a posterioru LogMAP turbo decoding. It shows how to adapt Sum-Alphas quantities when using the Log-MAP algorithm and how to deduce a good decision threshold. We apply a logarithm to the quantity Sum-Alpha which is evaluated by the same Jacobian logarithm of the Log-MAP algorithm. We call this new adaptation Jacobian Approximation of Sum-Alpha (JASA) criterion. The simulation results demonstrate that the JASA criterion achieves comparable performance (in terms of bit error rate (BER) and frame error rate (FER)) to the Sum-Alpha and cross-entropy (CE) criteria, with the same average number of iterations
Adulterated beef detection with redundant gas sensor using optimized convolutional neural network
Full text linkVarious types of research have been developed to detect beef adulteration, but the accuracy and reliability of these results still require improvement. This study proposes designing a highly precise redundant electronic nose system using an optimized convolutional neural network (CNN) method to detect adulterated beef mixed with pork. As baselines, other classifiers are also utilized, namely the decision tree (DT), K-nearest neighbor (KNN), artificial neural network (ANN), and support vector machine (SVM). Several data preprocessing methods are employed to increase prediction accuracy, namely feature selection, principal component analysis (PCA), and time series smoothing. The weight of each data sample was 100 g with 15 classes of pork and beef mixing ratios of 0%, 0.1%, 0.5%, 1%, 5%, 10%, 20%, 30%, 40%, 50%, 60%, 70%, 80%, 90%, and 100% pork. With the single-layer sensor configuration, the average CNN classification success rates were 97.15%, 96.29%, and 99.64% for layers 1, 2, and 3, respectively. In addition, from the combination of the three layers, a prediction results of 99.72% was obtained. Thus, a redundant gas sensor array configuration can improve the classification results. In addition, the relatively high accuracy of the optimized CNN provides a convincing alternative for identifying possible beef adulteration
Elevating cultural understanding: interactive museum exploration using 3D AR and MDLC framework
Full text linkLimited access to information and interaction with artifacts in museums often hinders visitors from gaining a deeper understanding of the culture and historical context presented. This study addresses this challenge by developing a three-dimensional (3D) augmented reality (AR)-based interactive museum that enhances the museum visitor experience through an intuitive user interface (UI) and enriched content related to the exhibited artifacts. This study explores the potential of 3D AR technology in enhancing visitor engagement and interaction with museum exhibits, providing a more immersive and informative experience. This study uses the multimedia development life cycle (MDLC) as a framework to develop a 3D AR-based interactive museum. By applying the MDLC approach, this study integrates advanced AR technology with comprehensive and detailed content, resulting in a structured and user-centered interactive platform. Key benefits of this approach include enhanced interactivity, enriched artifact information, and an intuitive interface that facilitates easier access to museum content. The findings indicate that the developed interactive museum successfully overcomes the barriers of limited accessibility of information and interaction with artifacts. Through the application of advanced AR technology, the museum visitor experience is significantly enhanced, making the museum more inclusive, interactive, and educative for visitors
The comparison of underwater source localization between Riemannian MFP and blind channel equalizer
Full text linkBlind channel equalization (BCE) has been widely used in underwater communications due to its strong robustness against multipath propagation and its suitability for rapidly varying environments. However, there has been little research on the application of BCE for underwater source localization. On the other hand, conventional matched field processing (MFP), and particularly Riemannian MFP (RMFP), have been regarded as highly effective for this problem. In this paper, based on the statistical characterization of the signal-to-noise ratio (SNR) in underwater acoustic channels, we propose a method for estimating the channel transfer function, which is then used to construct a blind channel equalizer. A source localization approach using the proposed BCE is also presented. The localization performance using BCE is comparable to that of RMFP, achieving a depth error of 10 meters and a range error of 100 meters, while requiring significantly lower computational complexity
Grid search vs Bayesian optimization for intensity scoring classification and channel recommendation prediction
Full text linkTechnological advancement has spurred financial technology growth, transforming traditional financial operations into digital. Peer-to-peer (P2P) lending is a key fintech solution offering online loans, though it struggles with repayment issues due to customer financial instability. To overcome these challenges, XYZ is a startup that focuses on enhancing the efficiency of collections and communication with customers. XYZ necessitates the implementation of a collection intensity scoring (CIS) model and a prediction model for interaction on recommended communication channels in order to optimize the collection process. This study evaluates the performance of grid search and Bayesian optimization on random forest (RF) classification models and K-nearest neighbors (KNN) regressor prediction models. RF and KNN regressor algorithms optimization are necessary to enhance their performance in CIS classification and channel recommendation prediction. The research stages follow the cross industry standard process-data mining (CRISP-DM) framework, which consists of business understanding, data understanding, data preparation, modeling, and evaluation. The model performance is assessed by accuracy and mean absolute error (MAE). The results of this study show that Bayesian optimization surpasses grid search, enhancing the accuracy of the RF model to 98.34% and reducing the MAE of the KNN regressor model to 0.24530
Modified electro-optical modulator based bipolar optical code division multiple access for free-space optics
Full text linkThis study introduces a modified electro-optical modulator (EOM) to enhance bipolar optical code division multiple access (OCDMA) in free-space optical (FSO) communication. The proposed system improves signal quality, spectral efficiency, and resilience to atmospheric turbulence. Unlike conventional dual EOM techniques, the modified EOM enables simultaneous transmission of ‘0’ and ‘1’ chip values, reducing multiple access interference (MAI) and enhancing system robustness. This approach optimizes bandwidth utilization and ensures stable performance in varying environmental conditions. Simulations were conducted in an additive white Gaussian noise (AWGN) channel using three spectral amplitude coding (SAC) schemes: modified M sequence, Walsh-Hadamard, and random diagonal (RD) codes. Results indicate that the modified EOM scheme significantly improves FSO system performance, achieving an average 47.1% lower minimum log of bit error rate (BER) in normal weather and 43.3% lower in extreme weather, ensuring superior noise suppression and signal integrity across varying environmental conditions. Additionally, the system maintains superior performance over longer distances, demonstrating its suitability for high-speed, long-range FSO applications. These findings highlight the potential of the modified EOM based bipolar OCDMA system in advancing next-generation optical wireless networks, offering a more efficient, interference-resistant, and high-capacity communication solution for future technologies such as 6G and beyond
DDoS attack detection using optimal scrutiny boosted graph convolutional and bidirectional long short-term memory
Full text linkThe distributed denial of service (DDoS) attack occurs when massive traffic from numerous computers is directed to a server or network, causing crashes and disrupting functionality. Such attacks often shut down websites or applications temporarily and remain among the most critical cybersecurity challenges. Detecting DDoS is difficult and must occur before mitigation. Recently, machine learning and deep learning (ML/DL) have been employed for detection; however, architectural limitations restrict their effectiveness against evolving attack methods. This paper presents a novel framework, scrutiny boosted graph convolutional–bidirectional long short-term memory and vision transformer (SBGC-BiLSTM-ViT), which integrates graph convolutional, BiLSTM, and ViT models with machine learning classifiers such as support vector machine (SVM), Naïve Bayes (NB), random forest (RF), and K-nearest neighbors (KNN). The integration enables autonomous extraction of critical features, enhancing precision in detecting and classifying DDoS attacks. To further boost performance, a Bayesian optimization algorithm (BOA) is applied for hyperparameter tuning of SBGC and ML methods. Evaluation on benchmark datasets UNSW-NB15 and CICDDoS2019 demonstrates that the proposed approach achieves higher accuracy and effectively identifies new DDoS variants, outperforming conventional methods
Design of a soft circular patch antenna operating in the 60 GHz band for 5G/6G applications
Full text link5G/6G technologies require higher-performance antennas in terms of bandwidth, gain, radiation, compact size, efficiency, and low cost. At the same time, fewer natural disturbances, such as rain and snow, and fewer non-natural disturbances. This is the challenge facing scientific research into antenna design and manufacture. In addition, in this paper we study and design a flexible circular microstrip patch antenna operating in the 60 GHz band for 5G/6G applications. This antenna is based on a biosourced substrate for industrial, scientific, and medical applications. For this study, we will use two techniques: one concerns the deformation of the ground plane deformation of the ground plane and substrate to improve the electrical performance of a proposed antenna. At the same time, the other is the parametric study of the appropriate position of a coaxial feed probe. This technique has the advantage of requiring no radiation contrition on the part of the coaxial probe. Next, specialized high-frequency structure simulator (HFSS) simulation software is used to design this antenna; it has a wide bandwidth above 3 GHz, a gain of 7.41 dB, a directivity of 7.53 dB, a radiated power of 13.55 dBm, an accepted power of 13.67 dBm, an incident power of 15.08 dBm, a radiation efficiency of 97.29 % and an efficiency of 98.4 %
Graphene-based high-gain MIMO antenna for enhanced 6G wireless communication systems
Full text linkThis paper presents a novel design and analysis of a high-performance multiple-input multiple-output (MIMO) terahertz (THz) antenna intended for next-generation sixth-generation (6G) wireless communication systems. The proposed antenna operates over a wide frequency range of 1 THz to 4.9 THz, achieving a broad bandwidth of 3.9 THz with three distinct resonant frequencies at 2.05 THz, 3.9 THz, and 4.52 THz, each exhibiting excellent return loss characteristics. The antenna features a graphene-based patch with a copper ground plane, etched on a polyimide substrate with a dielectric constant (εr) of 3.5 and a thickness of 10 micrometers (μm). Key performance metrics, including a high gain of 15.9 decibels (dB), an efficiency of 95.95%, an envelope correlation coefficient (ECC) of 0.0005, and a diversity gain (DG) of 9.997 dB, indicate outstanding performance. The measured isolation between the two antenna elements is -31.91 dB, signifying excellent isolation. An equivalent resistor-inductor-capacitor (RLC) circuit model is developed using advanced design system (ADS), validated by comparing S11 results from both computer simulation technology (CST) and ADS simulations. The proposed MIMO antenna’s wide operating range and robust performance demonstrates great potential for high-speed THz wireless communication, imaging, spectroscopy, sensing, and offers valuable contributions to industry and innovation