International Journal Information and Communication Technology Research
Volume:13 Issue: 2, Spring 2021

A compact, high gain, series fed transverse slot array antenna is designed on the top plate of a shorted corrugated H-plane horn. Corrugations are used to reduce the wavelength inside the shorted horn in order to suppress the grating lobe problem associated with the transvers slot arrays. By etching eight transverse slots separated by one guided wavelength, the array is formed and a proper broadside radiation is successfully achieved. The proposed structure can also be interpreted as a flat horn with directive radiation which is also more compact compared to the conventional H-plane horns as the guided wavelength inside it is reduced. The antenna is simulated by HFSS simulator and optimized for maximum gain and best matching condition. The simulation results show that the reflection coefficient is less than -10 dB around 15.15 GHz with about 1.3 % fractional bandwidth and the gain and SLL are about 19 dBi and -13 dB, respectively. Also, the cross polar discrimination of the antenna is better than 25 dB in both E- and H-planes. Simulation results were confirmed by comparing them with those obtained by CST software.

In this paper, we propose an analytical model for dimensioning of Orthogonal Frequency Division Multiple Access (OFDMA) systems in 5G networks by considering Internet of Thing (IoT) application using stochastic geometry. In these systems, some communication is lost when the number of required subcarriers is greater than the number of the available subcarriers. We compute the upper bound of the lost communication probability for downlink. In such a system, the position of the receiving users is modeled by the Poisson point process (PPP). The number of subcarriers dedicated to each user depends on its Signal to Noise Ratio (SNR), position and the shadowing, hence for calculating the number of subcarriers, it is needed to use stochastic geometry. Since the focus of our work is on IoT application in 5G networks, a multi-group user system with each group of users having its own application and throughput requirement is considered. For having dimensioning in terms of subcarriers, we present concentration inequality for functions defined on PPP to calculate the upper bound of loss probability. The performance of the upper bound in different range of user intensity is investigated.

The increasing fascination with the Internet of Things has led to the extensive deployment of Low-power and Lossy Networks. IPv6 Routing Protocol over Low Power and Lossy Networks serves as the ideal routing protocol proposed by IETF for routing in IoT-LLNs. Routing attacks are one of the IoT challenges that can lead to network performance problems and often denial of service. The Destination Advertisement Object (DAO) insider attack is one of the most notable attacks in RPLs, and previous studies have not developed a complete method for its detection so as to separate the malicious node from the normal node. Using an anomaly-based intrusion detection system, this paper suggests three methods based on random, fixed, and dynamic threshold adjustment to prevent DAO insider attack and identify malicious nodes. The results showed that the proposed model has a detection rate of 100% and a very low rate of false alert.

Contextual feature extraction is studied for polarimetric synthetic aperture radar (PolSAR) image classification in this work. The contextual locality preserving projection (CLPP) method is proposed for generation of contextual feature cubes using limited training samples. The local information in neighborhood regions is used to extend the training set by including the spatial information. Then, a supervised transform is applied to the polarimetric-contextual feature cube to reduce the data dimensionality while preserves the local structures and settles the samples belonging to the same class close together. Finally, a guided filter is applied to the classification map to degrade the speckle noise.  The classification results on two real L-band PolSAR data from AIRSAR show superior performance of CLPP for PolSAR classification in small sample size situations.

Named Entity Recognition is a challenging task, specially for low resource languages, such as Persian, due to the lack of massive gold data. As developing manually-annotated datasets is time consuming and expensive, we use a multitask learning (MTL) framework to exploit different datasets to enrich the extracted features and improve the accuracy of recognizing named entities in Persian news articles. Highly motivated auxiliary tasks are chosen to be included in a deep learning based structure. Additionally, we investigate the effect of chosen datasets on performance of the model. Our best model significantly outperformed the state of the art model by , according to F1 score in the phrase level.

A smart contract is a computer protocol for creating or improving a contract which makes it possible to create valid transactions without intermediaries. The most important feature is security and speed, because this technology runs on a blockchain platform and its information will remain confidential. Despite these benefits, unfortunately, companies still use paper contracts. Knowledge-based companies can save time and money by implementing smart contracts with their customers in the form of robotic process automation and process management, and by reducing errors and risks in processes. Increase productivity in business. The purpose of this study was to present a smart contract model in knowledge-based companies based on Grounded Theory by Focusing on the robotic process automation strategies and process management using qualitative and quantitative paradigms. The analysis approach in this research is quantitative-qualitative. To collect data in the qualitative part, semi-structured interviews were used. In the quantitative part, the structural equation method was used. The sample size was calculated according to confirmatory factor analysis of 110 experts. Based on the data analysis, due to the abnormality of the data distribution, the partial least squares method was used with the help of Smart PLS software version 2.