2018 International Conference on Artificial Intelligence and Data Processing (IDAP), 2018
In this study, a real-time system which counts the number of people with the help of a camera was... more In this study, a real-time system which counts the number of people with the help of a camera was demonstrated. The system can send the number of people to a mobile application via Internet of Things (IoT) and monitor simultaneously. This work is carried out in the main library of Inonu University. Background subtraction method was used to recognize moving humans on the visual field of the camera. According to motion information of humans, a counter was used to count the number of people in the saloon by determining whether going inside or outside. The counter will inform the users about what percentage of the saloon is empty. Matlab and Thingspeak combination help to send counter information to internet environment. A mobile application was used to track the counter information from Android and iOS smartphones. The results were presented in Matlab environment and mobile application simultaneously. Thanks to this work, students will not have to look for a place to work when the library is crowded and will not bother other working students. It is believed that this project will serve all students.
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
In this study a synthetic aperture radar-based reconstruction technique is applied to focus the u... more In this study a synthetic aperture radar-based reconstruction technique is applied to focus the unwanted disruptive effects in the two-dimensional ground penetrating radar data. The formulation of the applied Matched Filter Algorithm (MFA) which completely removes the hyperbolic effects is presented in detail. To test the capability of the proposed technique an experimental measurement setup is constructed in the laboratory and MFA is operated to gathered real GPR data. Thanks to the applied technique, high-resolution focused B-Scan GPR images are successfully obtained.
Journal of Intelligent Systems with Applications, 2018
In this study, an image processing techniques (IPTs) and a Sugeno-typed neuro-fuzzy system (NFS) ... more In this study, an image processing techniques (IPTs) and a Sugeno-typed neuro-fuzzy system (NFS) model is presented for classifying the wheat grains into bread and durum. Images of 200 wheat grains are taken by a high resolution camera in order to generate the data set for training and testing processes of the NFS model. The features of 5 dimensions which are length, width, area, perimeter and fullness are acquired through using IPT. Then NFS model input with the dimension parameters are trained through 180 wheat grain data and their accuracies are tested via 20 data. The proposed NFS model numerically calculate the outputs with mean absolute error (MAE) of 0.0312 and classify the grains with accuracy of 100% for the testing process. These results show that the IPT based NFS model can be successfully applied to classification of wheat grains.
2021 International Conference in Advances in Power, Signal, and Information Technology (APSIT), 2021
Novel chaotic map (CM) based on pi number socalled Pi-map is introduced for image encryption sche... more Novel chaotic map (CM) based on pi number socalled Pi-map is introduced for image encryption scheme (ImES). The chaotic behavior of the Pi-map is appreciated with respect to Lyapunov Exponent (LE) and bifurcation graphs. The LE which is stable and highest is also compared those reported elsewhere. The image encryption performance of Pi-map is corroborated on a simple ImES consisting of only two operations which are permutation and diffusion. The ImES with Pi-map is crypto-studied through various analysis and, they are even crosschecked with the others in the literature. It is shown that ImES is eminent and reliable thanks to the capability of the Pi-map regarding randomness and complexness.
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
In this study, a compact C-shaped printed antenna with reconfigurable band-notched characteristic... more In this study, a compact C-shaped printed antenna with reconfigurable band-notched characteristic is designed for ultra-wideband (UWB) operation. The basic structure is a C-shaped radiator with a very small size of 8×8 mm2 fed by a microstrip transmission line (MTL) with a matching line overall on 8×26 mm2 substrate under which a portion of 8×26 mm2 ground conductor is used. The capability of reconfigurable band-notched characteristic is gained by inserting a pin diode between the edges of C-shaped radiator. The antenna design is printed on a double sided copper with FR4 substrate. Based on measured result, the antenna provides an impedance bandwidth of 3.1-12.9 GHz (at VSWR=2) when it operates at ON state (full UWB). At OFF state, it passes the band of 3.0-3.5 GHz and 5.6-11.7 GHz, thus stop the band of 3.5-5.6 GHz corresponding to WIMAX and WLAN standard. Therefore, the proposed UWB antenna stands out over the literature in terms of geometry’s simplicity, compactness, near omnidirectional pattern, high efficiency as well as band-notched configurability.
Bu calismada, UHF bandinda calisan H sekilli kompakt mikroserit antenlerin (HKMA) rezonans frekan... more Bu calismada, UHF bandinda calisan H sekilli kompakt mikroserit antenlerin (HKMA) rezonans frekansi, cok katmanli algilayicilara (CKA) dayanan yapay sinir aglari (YSA) ile hesaplanmistir. Fiziksel ve elektriksel parametreleri farkli 216 adet HKMA’nin simulasyonu, FDTD (Finite Difference Time Domain) yontemini kullanan bir elektromanyetik simulasyon programi ile yapilmis ve her bir anten icin rezonans frekansi degeri belirlenmistir. Bu antenlerin 196’si, YSA modelinin egitilmesinde kullanilirken geriye kalan 20’si ise test edilmesinde kullanilmistir. Olusturulan YSA modelinde, egitim ve test icin ortalama yuzde hata degerleri, sirasiyla, % 0,68 ve % 0,84 olarak elde edilmistir. Ayrica, YSA yapisinin dogrulugunu ve gecerliligini gostermek icin, literaturde sunulan bazi calismalarda verilen simulasyon ve olcum sonuclari ile bu calismada tasarlanan HKMA’nin olcum sonuclari da test amaciyla kullanilmis ve ortalama yuzde hata degerleri, sirasiyla % 1,18 ile % 1,20 olarak elde edilmistir. ...
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
Because of increasing the use of smartphones, it has become easier to identify location of any us... more Because of increasing the use of smartphones, it has become easier to identify location of any user. The most popular technique for outdoor positioning is the GPS signal which is commonly used in smartphones and transport vehicles. However, position detection can not be achieved indoor with GPS. Therefore, in this study, a location determination based on WiFi signal strengths was performed indoor where user could not correctly receive the GPS signal. The data includes the strengths of seven WiFi signals that provide information about four different rooms. Based on the WiFi signal strength values coming from seven different sources to smartphone, the position of the user at which room can be determined. In this study, classification was achieved for the determination of the indoor room. Six different Machine Learning (ML) methods were applied to the classification. These methods are Artificial Neural Networks (ANN), K-Nearest Neighbors (k-NN), Decision Trees (DT), Naive Bayes (NB) Classifier, Extreme Learning Machine (ELM) and Support Vector Machines (SVM). Successful results were obtained from all the methods and these results were compared with each other.
In this paper, a color feature-based classification of the wheat grains into bread and durum usin... more In this paper, a color feature-based classification of the wheat grains into bread and durum using artificial neural network (ANN) model with bayesian regularization (BR) learning algorithm is presented. Images of 200 wheat grains are taken by a high resolution camera in order to generate the data set for training and testing processes of the ANN-BR model. Data of 3 main colour features (R, G and B) for 200 wheat grains (100 for durum and 100 for bread) are acquired for each grain using image processing techniques (IPTs). Features of R, G and B are separately determined by taking arithmetic average of the pixels within each grain. Several colour features of R/TRGB, G/TRGB, B/TRGB, R-G, G-B and R-B where TRGB is the total of R+G+B are reproduced. Then ANN-BR model input with the 9 colour parameters are trained through 180 wheat grain data and their accuracies are tested via 20 data. The ANN-BR model numerically calculate the outputs with mean absolute error (MAE) of 0.0060 and classi...
In this study, classification of two types of wheat grains into bread and durum was carried out. ... more In this study, classification of two types of wheat grains into bread and durum was carried out. The species of wheat grains in this dataset are bread and durum and these species have equal samples in the dataset as 100 instances. Seven features, including width, height, area, perimeter, roundness, width and perimeter/area were extracted from each wheat grains. Classification was separately conducted by Artificial Neural Network (ANN) and Extreme Learning Machine (ELM) artificial intelligence techniques. Then the performances of models are compared each other. The accuracy of testing was calculated 97.89% and 96.79% for ANN and ELM, respectively.
A microstrip meander line with a single H-shaped Defected Ground Structure (HDGS) is employed to ... more A microstrip meander line with a single H-shaped Defected Ground Structure (HDGS) is employed to design a novel compact lowpass filter with sharp roll off and wide stop band. Two parametric studies were conducted to analyze the sensitivity of the HDGS position and the size effects on the filter response. The design is fabricated to have a 6 GHz cutoff frequency using Rogers TMM 10i substrate, with an area of 7.6 x 6 mm. The measured results are in a good agreement with simulation prediction. Index Terms – Defected ground structures, compact filter, lowpass filter and microstrip meander line.
2020 IEEE XXVth International Seminar/Workshop Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2020
In this study, a dual-element inverted multiple input multiple output (MIMO) F antenna (IFA) oper... more In this study, a dual-element inverted multiple input multiple output (MIMO) F antenna (IFA) operating at 2.4 GHz band is designed by means of computational electromagnetic for WLAN wireless devices. The shape of IFA element resembles two superimposed E-shaped strips. IFA element with a small size of 10 x13mm2 operates at the frequency range of 2.28-2.62 GHz. They are orthogonally placed at the corners of a 50 x100mm2 main ground board for reducing the mutual coupling, i.e. correlation between the two elements. The radiation and isolation performance of MIMOIFA is examined in regards to the radiation patterns, envelope correlation coefficient (ECC) and peak gain. The proposed MIMO-IFA shows quasi-omni-directional pattern, under 0.05 ECC and about 2.7 dBi peak gain at 2.4 GHz. Therefore, it is a better and cheap candidate for 2.4 GHz industrial scientific medical (ISM) band including WLAN as well as wireless mobile devices.
Uludağ University Journal of The Faculty of Engineering, 2016
An artificial neural network (ANN) trained by different learning algorithms implemented to comput... more An artificial neural network (ANN) trained by different learning algorithms implemented to computing the operating frequency of E-shaped patch antennas (EPAs) is designed in this study. The ANN model is built on a multilayered perceptron (MLP) based on feed forward back propagation (FFBP). A data pool is firstly constituted for training and testing the ANN model through 144 EPA simulations using the moment method-based HyperLynx® 3D EM software in terms of the operating frequency. The ANN model is then trained via 130 data, and the accuracy of the model is tested through 14 data of simulated EPAs. The ANN is trained by 8 different learning algorithms to achieve a robust model. A benchmark which compares the learning algorithms against each other according to percentage error is revealed. The validity of the ANN is corroborated by simulated and measured data reported in the literature. It shows that the ANN model trained by Levenberg–Marquardt learning algorithm computes the closest ...
2019 International Artificial Intelligence and Data Processing Symposium (IDAP), 2019
Due to the latest developments in communication systems, the technology of many devices developed... more Due to the latest developments in communication systems, the technology of many devices developed for wireless communication systems have also changed. Therefore, the designs of antenna with properties that are lightweight, cost-effective, small-sized and a wide operating frequency range are very important to use in the mobile wireless communication systems. However, the use of computer-added antenna analysis techniques that can obtain accurate results in order to design antenna designs quickly and effectively has gained importance. In this study, a surrogate model has been proposed to determine the optimal feed point of rectangular microstrip antennas. Twenty-three antennas with different dimensions and dielectric constant values presented in the literature were used for the creation of the surrogate model. The predicted feeding results by the SM are being compatible with the optimal feed points of the antennas determined in the simulation.
2017 10th International Conference on Electrical and Electronics Engineering (ELECO), 2017
In this study, a compact ultra-wideband (UWB) multiple input multiple output (MIMO) antenna syste... more In this study, a compact ultra-wideband (UWB) multiple input multiple output (MIMO) antenna system is designed for mobile terminals. The antenna system consists of two symmetrical 8×8 mm2 square antenna elements fed by a 50 Ohm microstrip transmission line (MTL). The UWB MIMO antenna operates at large frequency range of 2.5–13.3 GHz with an isolation level of about 9 dB less than 5 GHz and 15 dB greater than 5 GHz in average. In fact, the isolation level is almost sufficient for MIMO operation especially for high frequency. Nevertheless, various techniques are then investigated, and additional parasitic structures (PSs) such as monopole, V-shaped, plus-shaped, T-shaped PSs and neutralization line are thus designed on the square UWB MIMO for further reducing the mutual coupling. Among the PSs, T-shaped PS shows better return loss and isolation performance. Moreover, the proposed UWB MIMO antenna with T-shaped structure is investigated in terms of resonant bandwidth, radiation pattern...
─ An artificial neural network (ANN) based estimator is presented for notch antenna analysis in t... more ─ An artificial neural network (ANN) based estimator is presented for notch antenna analysis in terms of the operating frequency. The notch antenna is formed by loading an asymmetric slot on one side of a rectangular patch. Architecture of the estimator is constructed over an ANN model trained with the simulated data of the notch antennas. In order to constitute a data pool for training and testing the ANN model, 96 notch antennas with seven antenna parameters are simulated with respect to the operating frequency. Antenna parameters including the patch dimensions, height and relative permittivity of the substrate are used as input vector of the ANN model. The simulated data of 80 notch antennas are employed to train the ANN model. The estimator is corroborated through testing with the remainders 16 antenna data, verifying with antenna data in the literature and validating with a prototyped notch antenna data. The results show that the estimator simply and fast computes the operating...
2018 International Conference on Artificial Intelligence and Data Processing (IDAP), 2018
In this study, a real-time system which counts the number of people with the help of a camera was... more In this study, a real-time system which counts the number of people with the help of a camera was demonstrated. The system can send the number of people to a mobile application via Internet of Things (IoT) and monitor simultaneously. This work is carried out in the main library of Inonu University. Background subtraction method was used to recognize moving humans on the visual field of the camera. According to motion information of humans, a counter was used to count the number of people in the saloon by determining whether going inside or outside. The counter will inform the users about what percentage of the saloon is empty. Matlab and Thingspeak combination help to send counter information to internet environment. A mobile application was used to track the counter information from Android and iOS smartphones. The results were presented in Matlab environment and mobile application simultaneously. Thanks to this work, students will not have to look for a place to work when the library is crowded and will not bother other working students. It is believed that this project will serve all students.
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
In this study a synthetic aperture radar-based reconstruction technique is applied to focus the u... more In this study a synthetic aperture radar-based reconstruction technique is applied to focus the unwanted disruptive effects in the two-dimensional ground penetrating radar data. The formulation of the applied Matched Filter Algorithm (MFA) which completely removes the hyperbolic effects is presented in detail. To test the capability of the proposed technique an experimental measurement setup is constructed in the laboratory and MFA is operated to gathered real GPR data. Thanks to the applied technique, high-resolution focused B-Scan GPR images are successfully obtained.
Journal of Intelligent Systems with Applications, 2018
In this study, an image processing techniques (IPTs) and a Sugeno-typed neuro-fuzzy system (NFS) ... more In this study, an image processing techniques (IPTs) and a Sugeno-typed neuro-fuzzy system (NFS) model is presented for classifying the wheat grains into bread and durum. Images of 200 wheat grains are taken by a high resolution camera in order to generate the data set for training and testing processes of the NFS model. The features of 5 dimensions which are length, width, area, perimeter and fullness are acquired through using IPT. Then NFS model input with the dimension parameters are trained through 180 wheat grain data and their accuracies are tested via 20 data. The proposed NFS model numerically calculate the outputs with mean absolute error (MAE) of 0.0312 and classify the grains with accuracy of 100% for the testing process. These results show that the IPT based NFS model can be successfully applied to classification of wheat grains.
2021 International Conference in Advances in Power, Signal, and Information Technology (APSIT), 2021
Novel chaotic map (CM) based on pi number socalled Pi-map is introduced for image encryption sche... more Novel chaotic map (CM) based on pi number socalled Pi-map is introduced for image encryption scheme (ImES). The chaotic behavior of the Pi-map is appreciated with respect to Lyapunov Exponent (LE) and bifurcation graphs. The LE which is stable and highest is also compared those reported elsewhere. The image encryption performance of Pi-map is corroborated on a simple ImES consisting of only two operations which are permutation and diffusion. The ImES with Pi-map is crypto-studied through various analysis and, they are even crosschecked with the others in the literature. It is shown that ImES is eminent and reliable thanks to the capability of the Pi-map regarding randomness and complexness.
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
In this study, a compact C-shaped printed antenna with reconfigurable band-notched characteristic... more In this study, a compact C-shaped printed antenna with reconfigurable band-notched characteristic is designed for ultra-wideband (UWB) operation. The basic structure is a C-shaped radiator with a very small size of 8×8 mm2 fed by a microstrip transmission line (MTL) with a matching line overall on 8×26 mm2 substrate under which a portion of 8×26 mm2 ground conductor is used. The capability of reconfigurable band-notched characteristic is gained by inserting a pin diode between the edges of C-shaped radiator. The antenna design is printed on a double sided copper with FR4 substrate. Based on measured result, the antenna provides an impedance bandwidth of 3.1-12.9 GHz (at VSWR=2) when it operates at ON state (full UWB). At OFF state, it passes the band of 3.0-3.5 GHz and 5.6-11.7 GHz, thus stop the band of 3.5-5.6 GHz corresponding to WIMAX and WLAN standard. Therefore, the proposed UWB antenna stands out over the literature in terms of geometry’s simplicity, compactness, near omnidirectional pattern, high efficiency as well as band-notched configurability.
Bu calismada, UHF bandinda calisan H sekilli kompakt mikroserit antenlerin (HKMA) rezonans frekan... more Bu calismada, UHF bandinda calisan H sekilli kompakt mikroserit antenlerin (HKMA) rezonans frekansi, cok katmanli algilayicilara (CKA) dayanan yapay sinir aglari (YSA) ile hesaplanmistir. Fiziksel ve elektriksel parametreleri farkli 216 adet HKMA’nin simulasyonu, FDTD (Finite Difference Time Domain) yontemini kullanan bir elektromanyetik simulasyon programi ile yapilmis ve her bir anten icin rezonans frekansi degeri belirlenmistir. Bu antenlerin 196’si, YSA modelinin egitilmesinde kullanilirken geriye kalan 20’si ise test edilmesinde kullanilmistir. Olusturulan YSA modelinde, egitim ve test icin ortalama yuzde hata degerleri, sirasiyla, % 0,68 ve % 0,84 olarak elde edilmistir. Ayrica, YSA yapisinin dogrulugunu ve gecerliligini gostermek icin, literaturde sunulan bazi calismalarda verilen simulasyon ve olcum sonuclari ile bu calismada tasarlanan HKMA’nin olcum sonuclari da test amaciyla kullanilmis ve ortalama yuzde hata degerleri, sirasiyla % 1,18 ile % 1,20 olarak elde edilmistir. ...
2018 XXIIIrd International Seminar/Workshop on Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2018
Because of increasing the use of smartphones, it has become easier to identify location of any us... more Because of increasing the use of smartphones, it has become easier to identify location of any user. The most popular technique for outdoor positioning is the GPS signal which is commonly used in smartphones and transport vehicles. However, position detection can not be achieved indoor with GPS. Therefore, in this study, a location determination based on WiFi signal strengths was performed indoor where user could not correctly receive the GPS signal. The data includes the strengths of seven WiFi signals that provide information about four different rooms. Based on the WiFi signal strength values coming from seven different sources to smartphone, the position of the user at which room can be determined. In this study, classification was achieved for the determination of the indoor room. Six different Machine Learning (ML) methods were applied to the classification. These methods are Artificial Neural Networks (ANN), K-Nearest Neighbors (k-NN), Decision Trees (DT), Naive Bayes (NB) Classifier, Extreme Learning Machine (ELM) and Support Vector Machines (SVM). Successful results were obtained from all the methods and these results were compared with each other.
In this paper, a color feature-based classification of the wheat grains into bread and durum usin... more In this paper, a color feature-based classification of the wheat grains into bread and durum using artificial neural network (ANN) model with bayesian regularization (BR) learning algorithm is presented. Images of 200 wheat grains are taken by a high resolution camera in order to generate the data set for training and testing processes of the ANN-BR model. Data of 3 main colour features (R, G and B) for 200 wheat grains (100 for durum and 100 for bread) are acquired for each grain using image processing techniques (IPTs). Features of R, G and B are separately determined by taking arithmetic average of the pixels within each grain. Several colour features of R/TRGB, G/TRGB, B/TRGB, R-G, G-B and R-B where TRGB is the total of R+G+B are reproduced. Then ANN-BR model input with the 9 colour parameters are trained through 180 wheat grain data and their accuracies are tested via 20 data. The ANN-BR model numerically calculate the outputs with mean absolute error (MAE) of 0.0060 and classi...
In this study, classification of two types of wheat grains into bread and durum was carried out. ... more In this study, classification of two types of wheat grains into bread and durum was carried out. The species of wheat grains in this dataset are bread and durum and these species have equal samples in the dataset as 100 instances. Seven features, including width, height, area, perimeter, roundness, width and perimeter/area were extracted from each wheat grains. Classification was separately conducted by Artificial Neural Network (ANN) and Extreme Learning Machine (ELM) artificial intelligence techniques. Then the performances of models are compared each other. The accuracy of testing was calculated 97.89% and 96.79% for ANN and ELM, respectively.
A microstrip meander line with a single H-shaped Defected Ground Structure (HDGS) is employed to ... more A microstrip meander line with a single H-shaped Defected Ground Structure (HDGS) is employed to design a novel compact lowpass filter with sharp roll off and wide stop band. Two parametric studies were conducted to analyze the sensitivity of the HDGS position and the size effects on the filter response. The design is fabricated to have a 6 GHz cutoff frequency using Rogers TMM 10i substrate, with an area of 7.6 x 6 mm. The measured results are in a good agreement with simulation prediction. Index Terms – Defected ground structures, compact filter, lowpass filter and microstrip meander line.
2020 IEEE XXVth International Seminar/Workshop Direct and Inverse Problems of Electromagnetic and Acoustic Wave Theory (DIPED), 2020
In this study, a dual-element inverted multiple input multiple output (MIMO) F antenna (IFA) oper... more In this study, a dual-element inverted multiple input multiple output (MIMO) F antenna (IFA) operating at 2.4 GHz band is designed by means of computational electromagnetic for WLAN wireless devices. The shape of IFA element resembles two superimposed E-shaped strips. IFA element with a small size of 10 x13mm2 operates at the frequency range of 2.28-2.62 GHz. They are orthogonally placed at the corners of a 50 x100mm2 main ground board for reducing the mutual coupling, i.e. correlation between the two elements. The radiation and isolation performance of MIMOIFA is examined in regards to the radiation patterns, envelope correlation coefficient (ECC) and peak gain. The proposed MIMO-IFA shows quasi-omni-directional pattern, under 0.05 ECC and about 2.7 dBi peak gain at 2.4 GHz. Therefore, it is a better and cheap candidate for 2.4 GHz industrial scientific medical (ISM) band including WLAN as well as wireless mobile devices.
Uludağ University Journal of The Faculty of Engineering, 2016
An artificial neural network (ANN) trained by different learning algorithms implemented to comput... more An artificial neural network (ANN) trained by different learning algorithms implemented to computing the operating frequency of E-shaped patch antennas (EPAs) is designed in this study. The ANN model is built on a multilayered perceptron (MLP) based on feed forward back propagation (FFBP). A data pool is firstly constituted for training and testing the ANN model through 144 EPA simulations using the moment method-based HyperLynx® 3D EM software in terms of the operating frequency. The ANN model is then trained via 130 data, and the accuracy of the model is tested through 14 data of simulated EPAs. The ANN is trained by 8 different learning algorithms to achieve a robust model. A benchmark which compares the learning algorithms against each other according to percentage error is revealed. The validity of the ANN is corroborated by simulated and measured data reported in the literature. It shows that the ANN model trained by Levenberg–Marquardt learning algorithm computes the closest ...
2019 International Artificial Intelligence and Data Processing Symposium (IDAP), 2019
Due to the latest developments in communication systems, the technology of many devices developed... more Due to the latest developments in communication systems, the technology of many devices developed for wireless communication systems have also changed. Therefore, the designs of antenna with properties that are lightweight, cost-effective, small-sized and a wide operating frequency range are very important to use in the mobile wireless communication systems. However, the use of computer-added antenna analysis techniques that can obtain accurate results in order to design antenna designs quickly and effectively has gained importance. In this study, a surrogate model has been proposed to determine the optimal feed point of rectangular microstrip antennas. Twenty-three antennas with different dimensions and dielectric constant values presented in the literature were used for the creation of the surrogate model. The predicted feeding results by the SM are being compatible with the optimal feed points of the antennas determined in the simulation.
2017 10th International Conference on Electrical and Electronics Engineering (ELECO), 2017
In this study, a compact ultra-wideband (UWB) multiple input multiple output (MIMO) antenna syste... more In this study, a compact ultra-wideband (UWB) multiple input multiple output (MIMO) antenna system is designed for mobile terminals. The antenna system consists of two symmetrical 8×8 mm2 square antenna elements fed by a 50 Ohm microstrip transmission line (MTL). The UWB MIMO antenna operates at large frequency range of 2.5–13.3 GHz with an isolation level of about 9 dB less than 5 GHz and 15 dB greater than 5 GHz in average. In fact, the isolation level is almost sufficient for MIMO operation especially for high frequency. Nevertheless, various techniques are then investigated, and additional parasitic structures (PSs) such as monopole, V-shaped, plus-shaped, T-shaped PSs and neutralization line are thus designed on the square UWB MIMO for further reducing the mutual coupling. Among the PSs, T-shaped PS shows better return loss and isolation performance. Moreover, the proposed UWB MIMO antenna with T-shaped structure is investigated in terms of resonant bandwidth, radiation pattern...
─ An artificial neural network (ANN) based estimator is presented for notch antenna analysis in t... more ─ An artificial neural network (ANN) based estimator is presented for notch antenna analysis in terms of the operating frequency. The notch antenna is formed by loading an asymmetric slot on one side of a rectangular patch. Architecture of the estimator is constructed over an ANN model trained with the simulated data of the notch antennas. In order to constitute a data pool for training and testing the ANN model, 96 notch antennas with seven antenna parameters are simulated with respect to the operating frequency. Antenna parameters including the patch dimensions, height and relative permittivity of the substrate are used as input vector of the ANN model. The simulated data of 80 notch antennas are employed to train the ANN model. The estimator is corroborated through testing with the remainders 16 antenna data, verifying with antenna data in the literature and validating with a prototyped notch antenna data. The results show that the estimator simply and fast computes the operating...
Uploads
Papers by Abdurrahim Toktas