In this research paper, Giuseppe Peano and Cantor set fractals based miniaturized hybrid fractal antenna (GCHFA) is proposed that operates for biomedical applications. The proposed GCHFA is designed ...by merging Giuseppe Peano and Cantor set fractals that help in achieving better performance characteristics as well as miniaturization. Firefly algorithm (FA) has been employed to optimize the feed position of the designed antenna. The substrate material selected for the proposed GCHFA is low‐cost, commercially available FR4 epoxy whose thickness is 1.6 mm and relative permittivity is 4.4. A data set of 65 GCHFAs with different geometrical parameters is generated for the realization of two different bioinspired approaches. For the performance evaluation of fabricated prototype, vector network analyzer is used. The experimentally observed resonant frequencies are 2.4400 and 5.8115 GHz, and at these resonant frequencies, S (1,1) < −10 dB. The designed antenna is suitable for Industrial, Scientific, and Medical bands of biomedical applications. Moreover, the behavior of the proposed GCHFA is nearly omnidirectional. A comparative study of three different artificial neural networks (ANNs) is also done to evaluate the most suitable ANN type for the analysis of proposed GCHFA. The optimized, simulated, and experimental results depict that they are closely matched.
Full text
Available for:
FZAB, GIS, IJS, IZUM, KILJ, NLZOH, NUK, OILJ, PILJ, PNG, SAZU, SBCE, SBMB, UL, UM, UPUK
In this paper, a compact Giuseppe Peano, Cantor Set, and Sierpinski Carpet fractals based hybrid fractal antenna (GCSA) is designed and developed for Industrial, Scientific and Medical (ISM) band ...applications. The proposed GCSA is a hybrid fractal design which is created by fusing Giuseppe Peano, Cantor set, and Sierpinski carpet fractals together. The optimization of the microstrip line feed position is performed by using a Firefly Algorithm (FA). The substrate material employed for proposed GCSA is a low-priced, easily available FR4 epoxy of thickness 1.6 mm. By varying the geometrical dimensions of the radiating patch, a data set of 58 GCSAs is randomly generated for the realization of Artificial Neural Network (ANN) and FA approaches. The designed structure is fabricated, and then measured results are evaluated. The proposed GCSA is capable of resonating at 2.4450 GHz with S(1, 1) < -10 dB. The measured bandwidth of the operating ISM band is 101 MHz. The quantitative performance of three different ANN types reveals that Feed Forward Back Propagation ANN (FFBPN) shows minimum error in comparison to other two ANN types. The simulated, experimental, and optimized results show a good match that specifies the preciseness of the measurement.
This paper deals with the development of a virtual instrument for fault diagnosis in fractal antenna array using Lab‐VIEW software. Faults in antenna array are considered on the basis of the ...radiation pattern. In this study, theta and gain values of radiation patterns for each fault are used in Lab‐VIEW for curve fitting. An artificial neural network (ANN) has been developed for fitted data points using the Leavenberg Marquard algorithm in MATLAB software and mean square error (MSE) is minimized. The designed ANN model has been embedded in the virtual instrument. The proposed virtual instrument system gets test patterns as input and generates output for several faults present in antenna array. Simulated and measured results of the fractal antenna array are validated experimentally. This virtual instrument model has not been developed for fractal antenna array so far.
Full text
Available for:
FZAB, GIS, IJS, IZUM, KILJ, NLZOH, NUK, OILJ, PILJ, PNG, SAZU, SBCE, SBMB, UL, UM, UPUK
In this paper, the design of a Metasurface incorporated Frequency Reconfigurable Planar Antenna (MS-FRPA) for Wireless Applications is presented. The structure of the projected MS-FRPA consists of a ...patch with a metasurface placed one above the other with no gap between them. The MS is composed of an array of alternately placed dual split ring resonators arranged periodically in both horizontal and vertical directions. Frequency reconfiguration is achieved by rotating the MS relative to the designed patch antenna. The projected reconfigurable antenna is constructed on Rogers RO4350B material with thickness 1.524 mm. High Frequency Structure Simulator software is employed for analysis of the structure. The results clearly reveal that frequency tuning is achieved in 4.35 to 5.33 GHz with a fractional tuning range of 20.2%. The proposed structure provides appreciable realized gain with stable radiation patterns at all rotation angles. Further, the measured outcomes of the developed prototype show good correlation with the simulated outcomes.
This manuscript presents the design of an antenna based on nested square shaped ring fractal geometry with circular ring elements for multi-band wireless applications. The impedance bandwidth and ...reflection coefficient of the antenna are improved with the design of different iterations from the 0th to 2nd. The performance parameters of the antenna like reflection coefficient, VSWR, bandwidth, bandwidth ratio, and current density are improved in the final iteration. It also achieves the enhanced bandwidth greater than 3 GHz at three resonant frequency bands and exhibits additional frequency band at 2.4 GHz. Likewise, the frequency band of designed fractal antenna shifts towards the lower end and helps in achieving the miniaturization of antenna. The proposed fractal antenna is designed and fabricated on a low-cost FR4 glass epoxy substrate and investigated using HFSS software. The proposed antenna is optimized for generating different parameters, and the last geometry is fabricated and tested. Further, these parameters are compared with the experimental results and found in good agreement with each other. Due to the multi-band behaviour and improved bandwidth, the proposed fractal antenna can be considered as a good candidate for several wireless standards.
This paper reveals Koch Like Sided Sierpinski Square Carpet Multifractal Antenna (KLSSSCMA) having wideband features and comprises a new hybrid geometry using Sierpinski carpet and Koch curve ...geometries. The antenna is designed using square patch having dimensions 29.4 mm × 29.4 mm over Roger RT/Duroid substrate having height equal to 3.18 mm and dimensions 48.48 mm × 48.48 mm. The proposed antenna is analyzed and simulated using FEM (Finite Element Method) based Ansoft HFSSTM v.13 software successfully up to fourth iteration. Simulation results show that KLSSSCMA has wideband characteristics from 2 to 8 GHz that covers S and C bands for WiMAX, WPAN, WLAN and Wi-Fi applications. The projected antenna is capable of achieving a good impedance bandwidth with S11 less than -9 dB over the whole range of BW and also exhibits almost stable radiation patterns.
Full text
Available for:
IJS, IZUM, KILJ, NUK, PILJ, PNG, SAZU, UL, UM, UPUK
•Evaluated deep networks vs. supervised learning for rice leaf blast detection.•Models were developed at variable spatial resolution (16x16 to 256x256 pixels)•Deep networks had transformative ...potential than machine and supervised models.•AlexNet256x256 excelled intricate leaf blast pattern recognition (98.54% accuracy)•Inception_Res_V2224x224 has high accuracy (99.09%), balancing precision and recall.
The widespread prevalence of leaf blast disease poses a significant threat to rice (Oryza sativa L.) ecosystems globally, and inflicts substantial grain yield losses. The pervasive impact of leaf blast disease in rice ecosystems is compounded by the challenge of accurately identifying its symptoms, which often bear visual resemblance to those of other diseases. This complexity frequently leads to misdiagnosis by experts, underscoring the critical need for précised, reliable and robust diagnostic techniques for timely initiation of its control measures. In modern agriculture, automatic disease detection has played a pivotal role by revolutionizing the way plant diseases are identified, monitored and managed in different ecosystems.
Therefore, we developed four deep learning convolutional neural networks (CNNs) viz. AlexNet, Inception, Xception and Visual Geometry Group (VGG), machine learning models viz. decision tree (DT) vis-à-vis random forest (RF) and support vector machine (SVM) based supervised learning for automatic detection of leaf blast disease of rice. These models of moderate to high architectural complexity were developed and evaluated at variable spatial resolution (16 x 16 to 256 x 256 pixels) using a data-set of 2467 diseased + 2976 healthy plant images partitioned at 80:20 training-to-testing ratio. An independent validation was performed against 7200 images data-set (20 images captured from field and each rotated at 360°) as well as against all data-set (9667 images). These results revealed that AlexNet256x256 model has exceptionally high accuracy (∼98.54 %), which decreased substantially at lower spatial resolutions. Notably, AlexNet256x256 model entails ∼17.78 million trainable parameters, emphasizing its overwhelming significance for learning intricate patterns of rice leaf blast detection. As a function of number of epochs, we observed decreased loss during training (R2 = 0.8280–0.9593) and validation phase (R2 = 0.8506–0.9962), whilst increased accuracy during training (R2 = 0.9498–0.9933) and validation (R2 = 0.6814–0.9447). Similarly, Inception_Res_V2224x224 model exhibited exceptionally higher accuracy (∼99.09 %), precision (∼92.59 %), recall (100 %) and F1-score (∼96.15 %), highlighting model’s proficiency in making correct predictions across training and validation data-sets. Nonetheless, Xception224x224 model achieved highest training accuracy (exceeding ∼99.84 %) with precision and recall reaching perfect scores of 100 %. Additionally, F1-score (∼99.53 %) and sensitivity (∼98.67 %) and specificity (100 %) underpin Xception224x224 model’s high ability to correctly classify diseased instances, while minimizing false positives. Although, VGG model performed excellently, its accuracy was lower than the Xception224x224 model. These results revealed relatively inferior performance of machine and supervised learning models in accurately detecting leaf blast disease of rice.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ, ZRSKP
In this manuscript, the design of a high isolation Multiple Input Multiple Output (MIMO) Hybrid Fractal Antenna (HFA) has been presented for 5 G and other wireless standards in a wider operational ...frequency range from 1 to 30 GHz. The fractional ground plane is used for the assembly of the antenna in the initial phase, further, the T-stub has been exploited on the ground plane to expand the isolation between the radiating elements. The dimensions of the proposed antenna are optimised using the Particle Swarm Optimisation Gravitational Search Algorithm (PSOGSA). The performance parameters of the projected antenna have been measured and analysed. The obtained experimental results are highly correlated with simulated ones. Performance diversity parameters of designed antenna like CCL, DG, ECC, TARC, and MEG fall within working limits in the frequency range of interest. The developed antenna retains high isolation between −30 to−50 dB in the desired frequency range. The proposed model of an antenna can be useful for 5 G 3.5 GHz band, 5 G NR (New Radio) frequency bands (3.3-5.0 GHz), LTE band 46 (5.15-5.925 GHz), European Union 5 G frequency band (5.9-6.4 GHz), 5 G 26 GHz frequency band and other wireless standards in the frequency range from 1 to 30 GHz.
Full text
Available for:
BFBNIB, GIS, IJS, KISLJ, NUK, PNG, UL, UM, UPUK
In this work, an Artificial Neural Network (ANN) and Particle Swarm Optimization (PSO) techniques are used to compute the feed position of projected Minkowski, Giuseppe Peano and Koch curves based ...hybrid fractal Antenna (MGKA). The structure of projected MGKA is obtained by the combination of three different fractal geometries i.e. Minkowski, Giuseppe Peano and Koch. Different fractal geometries are combined to enhance the performance characteristics and to achieve better miniaturization. The projected MGKA consists of 1.6 mm thick, low cost FR4 substrate whose relative permittivity is 4.4 and loss tangent value is 0.02. A set of 70 MGKAs with different dimensions is designed to obtain a data dictionary for the implementation of ANN and PSO techniques. The prototype is fabricated and then tested with Vector Network Analyzer (VNA). From the measured results, it is depicted that the resonant frequencies are 2.44 GHz and 5.84 GHz along with S (1, 1) values −18.23 dB and −12.71 dB. The fabricated prototype also provides excellent gains of 4.69 dB and 15.11 dB and operates within the Industrial, Scientific and Medical (ISM) frequency bands i.e. (2.4–2.5 GHz) and (5.82–5.85 GHz). The results of proposed ANN model and PSO techniques are well matched with the simulated and experimental values.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZRSKP
The manuscript investigates the utilization of Koch curves and hexagonal ring-shaped geometry for an ultra-wideband fractal antenna, achieved through a combination of DGS, parasitic elements, and ...EC-SRR. The antenna's properties are investigated without and with DGS, Parasitic element, and EC-SRR in the hexagonal ring-shaped geometries for broader band characteristics. Simulation results show that the antenna designed without DGS resonates at five distinct frequencies, while with DGS, Parasitic element, and SRR, it resonates at six frequencies. An enhanced bandwidth of 10.42 GHz (99.43%) is revealed in the final antenna geometry, and the proposed antenna resonates at six frequencies, 3.4, 5.8, 8.5, 11, 14.2, and 15 GHz, with reflection coefficients of − 24.33, − 38.10, − 28.35, − 27.39, − 32.32, and − 15.93 dB, respectively. Combining a defected ground structure, parasitic element, and EC-SRR increases frequency bands and enhances the BW and reflection coefficient. With an overall dimension of 30 mm × 24 mm, the proposed antenna is suitable for wireless applications in the frequency ranges of 2.40–3.89 GHz and 5.33–15.75 GHz. The proposed antenna is fabricated, and the results are measured. It is found that simulated and measured results are in good agreement with each other.
Full text
Available for:
EMUNI, FIS, FZAB, GEOZS, GIS, IJS, IMTLJ, KILJ, KISLJ, MFDPS, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, SBMB, SBNM, UKNU, UL, UM, UPUK, VKSCE, ZAGLJ
PDF
