Telfor Journal Vol.14 No.2 (2022)

Content

On Guaranteed Correction of Error Patterns with Artificial Neural Networks

S. Brkić, P. Ivaniš, and B. Vasić
Topic:
Communications Systems
Abstract
In this paper, we analyze applicability of single- and two-hidden-layer feed-forward artificial neural networks, SLFNs and TLFNs, respectively, in decoding linear block codes. Based on the provable capability of SLFNs and TLFNs to approximate discrete functions, we discuss sizes of the network capable to perform maximum likelihood decoding. Furthermore, we propose a decoding scheme, which use artificial neural networks (ANNs) to lower the error-floors of low-density parity-check (LDPC) codes. By learning a small number of error patterns, uncorrectable with typical decoders of LDPC codes, ANN can lower the error-floor by an order of magnitude, with only marginal average complexity incense.
Keywords
error-floors, neural networks, linear block codes, low-density parity-check codes, ML decoding.
Full Text
HyperLink On Guaranteed Correction of Error Patterns with Artificial Neural Networks
Page(s)
51-55
Doi
10.5937/telfor2202051B

Software Optimization for Fast Encoding and Decoding of Reed-Solomon Codes

S. Skorokhod and A. Barlit
Topic:
Communications Systems
Abstract
In this work, we propose a software library written in C for encoding and decoding Reed-Solomon codes. Library consists of one scalar codec and two vectorized codecs for x86 architecture. Vectorized codecs use the benefits of SSSE3 or AVX2 instruction sets. We compare the performance of our three codecs with the JPWL RS codec from the Open JPEG library. The performance comparison methodology is described, and it is based on the measuring of the encoding and decoding speed. The results demonstrate a 4.1x speed gain with the scalar codec and a 19.6x gain with the vectorized codec. Based on testing results and supported instruction sets, a dynamic selection of codec version is proposed.
Keywords
FEC (Forward Error Correction), RS (Reed-Solomon) codes, SIMD (Single Instruction Multiple Data).
Full Text
HyperLink Software Optimization for Fast Encoding and Decoding of Reed-Solomon Codes
Page(s)
56-60
Doi
10.5937/telfor2202056S

A Comparative Study of Deep Learning and Decision Tree Based Ensemble Learning Algorithms for Network Traffic Identification

N. Nikolić, S. Tomović, and I. Radusinović
Topic:
Telecommunications Networks
Abstract
In this paper, we apply Deep Learning (DL) and decision-tree-based ensemble learning algorithms to classify network traffic by application. Various Deep Learning (DL) models for network traffic identification have been presented, implemented and compared, including 1D convolutional, stacked autoencoder, multi-layer perceptron, and combination of the aforementioned. Then the results of DL models have been compared to those obtained with two popular ensemble learning models based on decision trees - Random Forest and XGBoost. To train and test the classification models, a dataset containing both encrypted and unencrypted traffic has been collected in a real network, under normal operating conditions, and pre-processed in a way that ensures non-biased results. The classification uncertainties of the models have been also quantified on publicly available ISCX VPN-nonVPN dataset. The models have been compared in terms of precision, recall, F1 score and accuracy, for different levels of complexity and training dataset sizes. The evaluation results indicate that the decision-tree ensemble learning algorithms provide more accurate results and outperform the DL algorithms. The performance gap reduces with the dataset complexity.
Keywords
computer networks, decision-tree algorithms, deep learning, ensemble machine learning, traffic identification.
Full Text
HyperLink A Comparative Study of Deep Learning and Decision Tree Based Ensemble Learning Algorithms for Network Traffic Identification
Page(s)
61-66
Doi
10.5937/telfor2202061N

System for 3D Mapping using Affordable LIDAR

V. Krsmanović, M. Barjaktarović, and A. Gavrovska
Topic:
Multimedia
Abstract
In this paper a new system for 3D (three-dimensional) mapping using affordable LIDAR (light detection and ranging) is presented. The implementation of LIDAR technology-based approach enables obtaining a point cloud as a representation of indoor surrounding. In recent years with the help of LIDAR this kind of sensing has found numerous applications across various industries. Here, a cloud of points is generated during room scanning using Arduino platform based rotating system. The obtained results are promising, and the proposed solution can find its practical application in different fields. Moreover, it can provide many possibilities for future experiments with surrounding mappings, image matching, autonomous driving, obstacle observation, collision avoidance, material type detection such as transparent ones.
Keywords
LIDAR, sensor, affordable system, 3D, point cloud, computer vision.
Full Text
HyperLink System for 3D Mapping using Affordable LIDAR
Page(s)
67-72
Doi
10.5937/telfor2202067K

Comparison of Machine Learning Approaches to Emotion Recognition Based on DEAP Database Physiological Signals

T. Stajić, J. Jovanović, N. Jovanović, and M. Janković
Topic:
Signal Processing
Abstract
Recognizing and accurately classifying human emotion is a complex and challenging task. Recently, great attention has been paid to the emotion recognition methods using three different approaches: based on non-physiological signals (like speech and facial expression), based on physiological signals, or based on hybrid approaches. Non-physiological signals are easily controlled by the individual, so these approaches have downsides in real world applications. In this paper, an approach based on physiological signals which cannot be willingly influenced (electroencephalogram, heartrate, respiration, galvanic skin response, electromyography, body temperature) is presented. A publicly available DEAP database was used for the binary classification (high vs low for various threshold values) considering four frequently used emotional parameters (arousal, valence, liking and dominance). We have extracted 1490 features from the dataset, analyzed their predictive value for each emotion parameter and compared three different classification approaches – Support Vector Machine, Boosting algorithms and Artificial Neural Networks.
Keywords
DEAP database, emotion recognition, machine learning, physiological signals.
Full Text
HyperLink Comparison of Machine Learning Approaches to Emotion Recognition Based on DEAP Database Physiological Signals
Page(s)
73-78
Doi
10.5937/telfor2202073S

Serious Game Development of COVID-19 Social Distancing Simulator using Agent-based Modelling

W. P. A. Setiawan, A. B. Gumelar, A. T. Wibowo, MY T. Sulistyono, A. Julian, G. AK Laksono, and Rr D. N. Setyowati
Topic:
Software Tools and Applications
Abstract
A social distancing protocol has been acknowledged and encouraged practically everywhere, since the global COVID-19 epidemic in 2020. This research resulted in the Serious game to simulate social distance using agent-based modeling so that it can be used as a medium of entertainment and educate the public during the Pandemic COVID-19 conditions by developing simulation games using UNITY 3D to educate in the middle of the COVID-19 pandemic and reduce transmission rates on an individual scale.
Keywords
COVID 19, Serious Game, Social Distancing, Unity 3D
Full Text
HyperLink Serious Game Development of COVID-19 Social Distancing Simulator using Agent-based Modelling
Page(s)
79-84
Doi
10.5937/telfor2202079S

Simulation Environment for Scalability and Performance Analysis in Hierarchically Organized IoT Systems

H Turkmanović, I. Popović, Z. Čiča, and D. Drajić
Topic:
IoT
Abstract
The accelerated development of technologies, especially in the field of telecommunications, ease the integration of embedded devices within various IoT applications. Modern IoT applications assume heterogenous embedded platforms capable of collecting, processing, and exchanging data between the tiers of the IoT system architecture. Designing a multi-tier IoT system, even in the case of architecture that involves a small number of intelligent embedded devices, can be a very demanding process, especially when dealing with the strict requirements of IoT application concerning application performance, scalability, and energy consumption. In this paper, an open-source simulation framework for the performance analysis of an arbitrary multi-tiered IoT system is presented. Framework supports insight into the data availability within the tiers of IoT system enabling designers to evaluate the performance of IoT application and to engineer the system operation and deployment. Besides the performance analysis, proposed framework enables the analysis of energy consumption, architecture scalability utilizing different communication patterns and technologies. The case study of a large-scale IoT application for demonstrating the framework potential regarding the scalability and data availability analysis is also given.
Keywords
IoT, Simulation framework, Large Scale, Scalability.
Full Text
HyperLink Simulation Environment for Scalability and Performance Analysis in Hierarchically Organized IoT Systems
Page(s)
85-90
Doi
10.5937/telfor2202085T

Role of Sensors in the Paradigm of Industry 4.0 and IIoT

A. A. Porokhnya and I. U. Yakimenko
Topic:
IoT
Abstract
The purpose of this article is to review new trends in monitoring the condition of oil on all factory area processes. New solutions are being introduced into this industry with new advantages in the development of artificial intelligence, as well as machine learning and sensor technologies, which are applicable for data-based maintenance. They are called predictive maintenance. This paradigm is going to replace the old one. It changes the traditional routine preventive maintenance scheme and provides a deep understanding of the equipment performance [1]. Monitoring and checkout of conditions are necessary to maintain in a real-time environment because on-line control of equipment status can put down an operating cost, by eliminating the need for equipment outage for everyday diagnostics. The analysis based on oil samples is an effective tribotechnical systems approach for early diagnosis of failures, as it contains valuable information about the process of degradation of oil and the state of tribotechnical pairs [2]. But there are some problems with this method. The first is the way of oil sampling. There are lots of mistakes that may be made during the oil sampling process, and they can affect the results. The second is a delivery to laboratory which complicates the diagnostic process. That’s why we cannot say this approach is an on-line method of diagnostics. For the better prognosis of pending machinery failure one needs to know a real-time correlation between size, shapes, and concentration of wear debris parts [3].
Keywords
hydraulic fluids, lubricating oils, maintenance, operating conditions, sensors.
Full Text
HyperLink Role of Sensors in the Paradigm of Industry 4.0 and IIoT
Page(s)
91-97
Doi
10.5937/telfor2202085P