Volume 71 (2020): Issue 6 (December 2020)

The IoT environment includes the enormous amount of atomic services with dynamic QoS compared with traditional web services. In such an environment, in the service composition process, discovering a requested service meeting the required QoS is a di cult task. In this work, to address this issue, we propose a peer-to-peer-based service discovery model, which looks for the information about services meeting the requested QoS and functionality on an overlay constructed with users of services versus service nodes, with probably constrained resources. However, employing a plain discovery algorithm on the overlay network such as flooding, or k-random walk could cause high message overhead or delay. This necessitates an intelligent and adaptive discovery algorithm, which adapts itself based on users’ previous queries and the results. To fill this gap, the proposed service discovery approach is equipped with a reinforcement learning-based algorithm, named SARL. The reinforcement learning-based algorithm enables SARL to significantly reduce delay and message overhead in the service discovery process by ranking neighboring nodes based on users’ service request preferences and the service query results. The proposed model is implemented on the OMNet simulation platform. The simulation results demonstrate that SARL remarkably outperforms the existing approaches in terms of message overhead, reliability, timeliness, and energy usage efficiency.

Recent research has focused on finding ways to control hysteresis of dynamic comparators. The current proposed techniques are based on either geometrical dimension adjustment or digital control. The first case does not allow for post fabrication control, while the second has limited accuracy. This paper presents a new dynamic comparator design with external hysteresis adjustment using an analog voltage. This is achieved by proposing an architecture including control devices with a specific sizing. This is performed with no significant increase of the design complexity, keeping the power consumption as low as possible. The design is analyzed, showing that the proposed solution allows accurate hysteresis adjustment without affecting the inherent circuit properties. The dynamic comparator is also implemented using a 180 nm commercially available CMOS technology. The results show that a variation of 550 mV of the control voltage allows an accurate hysteresis adjustment ranging from 0 to 40 mV, according to the input conditions. Moreover, the simplicity of the circuit in conjunction with the use of dynamic technology have allowed the best performances to be achieved compared to the current state of the art, in terms of energy with an FoM equal to 116 fJ/decision and silicon area of 180 µm² .

To deliver steady electric power with decent quality, intelligent and healthy control schemes are greatly necessary for automatic generation control (AGC) of power systems therefore, this work presents the design of two degree of freedom based tilt integral derivative controller with filter for AGC. Firstly, a two-area reheat thermal model is considered, and the gain of the controller are optimized by adaptive differential evolution method. The advantage of the suggested approach is validated by equating results with latest published approaches such as symbiotic organism search algorithm and articial bee colony tuned PID controller. Further, the suggested method is extended to a three unequal area thermal model and the performance of results are equated with teaching-learning based optimization-PIDD and firey algorithm-PID for the identical model. Lastly, the experiment of the proposed scheme has been employed in real-time simulation using OPAL-RT, for validation of its viability and cogency.

The main purpose of this work is to propose a modern one-dimensional convolutional neural network (1 D CNN) configurations for distinguishing separate PD impulses from different types of PD sources while the parameters of these sources are changed. Three PD sources were built for signal generation: corona discharge, discharge in a void, and surface discharge. The reason for using separate PD impulses for classification is to develop a universal tool with the ability to recognize an insulation defects by analysing very few events in the insulation in a short range of time. Additionally, we found the optimal sample rates for the data acquisition for these network configurations. The necessity of signal filtering was also tested. The following configurations of a neural network were proposed: configuration for classification raw PD impulses; configuration for classification of PD impulses represented by power spectral density, for both filtered and unfiltered variants.

The presented experiments and studies are intended for photovoltaic applications of crystalline silicon. This work deals with chemical treatment of the surface of n-type silicon wafers with different resistivity to reduce their reflectivity. Chemical surface treatment of silicon is an alternative method to using the antireflection layer. Optical losses caused by the reflection of light from the surface of the solar cells significantly reduce their efficiency. The investigated samples were prepared by the electrochemical etching method in the solution based on hydrofluoric acid and ethanol. The analysis of the prepared samples is divided into two parts, namely experimental measurements, and theoretical modeling. Experimental measurements are performed using UV-VIS spectroscopy, spectroscopic ellipsometry and SEM microscopy. Theoretical modeling is based on the construction and optimization of theoretical model of optical response (reflectivity and ellipsometric parameters) to determine the effective refractive index and thickness of formed structure. Effective refractive index of studied samples in theoretical model of optical response is based on Looyenga effective medium approximation and Tauc-Lorentz dispersion model.

EHD (Electrohydrodynamic) printing is a promising technique for alternative fabrication of highresolution micro- and nanostructures without employment of any molds or photo-masks However, the printing precision can be easily influenced by the printing conditions, such as applied voltage, printing distance (the distance between the nozzle tip and the substrate), and flow rate. Unfortunately, up to now there was no work which analyzed those influencing factors in-depth and systematically by theory and numerical simulation. In this paper, the theory of EHD printing was presented and the effect of applied voltage, printing distance, and flow rate on the width of printed line was analyzed by numerical simulation. The simulation results showed that the width of printed lines is proportional to printing distance, nozzle size, and flow rate. However, it is inversely proportional to the applied voltage.

In this study, a three-way Wilkinson power divider (WPD) was proposed the circuit whose center frequency was selected as 1.9 GHz operates between 0.95-2.95 GHz frequencies and has a bandwidth of approximately 105%. The simulation and measurement results of the designed circuit were well aligned with each other below 10 dB return loss and 13 dB isolation. In addition, 0.4, 0.3 and 1.0 dB insertion loss were obtained in the output ways, respectively. The circuit is made more compact thanks to the curved structure of the output ways. It has been found that the proposed three-way power divider takes about 35% less space than the conventional circuit design.

In this paper, four differently shaped Wilkinson power dividers are presented by selecting the same physical length of two-section transmission lines, dual arbitrary frequency band Wilkinson power dividers can be achieved. The 2.4 GHz (WLAN) and 5.9 GHz (DSRC IEEE 802.11p) frequency bands are selected to complement the future development of multi-band, multi-standard transceivers. To improve physical separation and electrical isolation between the two output ports a parallel RLC circuit is employed. For verification, the simulated and measured performance results of dual-band Wilkinson power dividers implemented on the Rogers 4003C laminate are presented. The measurement results for the fabricated Wilkinson power dividers were in good agreement with theoretical simulation results and show dual-band characteristics.

This contribution presents, for the first time, the design of microwave lowpass filter with ultrawide stopband and low insertion loss based on the log-periodic zig-zag defected ground structure (DGS); the design is specifically evolved from the principle of the log-periodic line antenna and the slotline based DGS. Based on the log-periodic property, studies show that ultra-wideband suppression can be realized by increasing the flare angle or the angle subtended by each element of the slotline DGS. The developed filter therefore shows low insertion loss of 0.7 dB within the passband, sharp transition between the passband and the stopband and ultra-wide suppression band from 2.61-26.5 GHz with a 20 dB suppression level. The DGS architecture is simple and easy to realize, and the design is thus useful in many microwave applications. The fabricated prototype filter validates the study, both from simulations and from measurements.

An improved feedline configuration for dual-mode resonator filter is investigated in this paper. Based on the introduced topology, a dual-mode dual-band bandpass filter with center frequencies of 1.8 and 2.4 GHz is optimally designed, fabricated and tested. The introduced dual-band bandpass filter has simple structure and enables high selectivity to be realized due to two pairs of transmission zeros located near to the lower and upper passband, respectively. Both measured and simulated performances are presented with good consistency.