Proceedings of the 3rd International Conference on Electronic Engineering and Renewable Energy Systems

Inhaltsverzeichnis

1. Frontmatter

2. Communication, Networks and Information Technology

   1. Frontmatter

   2. A Fuzzy Logic-Based Intrusion Detection System for WBAN Against Jamming Attacks

      Asmae Bengag, Amina Bengag, Omar Moussaoui, Blej Mohamed

      Abstract

      Wireless Body Area Network (WBAN) is a set of special nodes called medical sensors. These sensors are very useful and helpful for making the user able to connect everywhere and every time. However, they suffer from many problems like the low computing capacity, energy and memory space. In terms of security, WBAN systems are threatened by various types of attacks due to the wireless communication. This technology must have a robust mechanism to detect attacks for making medical applications more reliable and safety. In our work, we have focused on identifying jamming attacks from WBAN using the fuzzy logic system (FLS) that is one of the powerful mechanisms of artificial intelligence (AI) to determine different network cases. The proposed system used one of the fuzzy inference methods named Mamdani model and based on three network parameters Packet delivery ratio (PDR), received strength signal indicator (RSSI) and energy consumption amount (ECA). Our intrusion detection system is simulated by using MATLAB 9.0 and Castalia for analyzing the output result as jamming detection index (JDI).

   3. An Enhanced Approach Based on PCA and ACO Methods for Facial Features Optimization

      Chaimaa Khoudda, El Miloud Smaili, Salma Azzouzi, Moulay El Hassan Charaf

      Abstract

      An automatic system for facial expression analysis consists generally of three main phases: detection, feature extraction and classification. In this study, we focus on the extraction of face characteristics (feature extraction) as well as the optimization of the obtained results. The objective is to reduce the number of facial features by removing noisy and redundant data in order to ensure an acceptable facial recognition accuracy while guaranteeing an optimal selection of distinctive facial information. For this purpose, we suggest a new approach based on the combination of principal component analysis (PCA) and ant colony algorithm (ACO). The study was conducted by exploiting the database of the Olivetti Research Laboratory (ORL).

   4. The GPSR Routing Protocol in VANETs: Improvements and Analysis

      Amina Bengag, Asmae Bengag, Mohamed Elboukhari

      Abstract

      One of the most critical problems in VANETs is the frequent link breakage caused by the high velocity of vehicles. Due to the short connection lifetime between vehicles, the communication paths are frequently interrupted during the transmission of data packets between the source and destination vehicles, causing the search for a new route that increases the routing overhead and diminishes the PDR and the throughput. To manage those issues, several routing protocols have been proposed by considering important factors to improve the quality of service in VANETs. The GPSR (Karp and Kung in ACM MobiCom, pp. 243–254, 2000) for Greedy Perimeter Stateless Routing is the most popular position-based protocol. In this paper, we propose three new models to enhance this protocol that guide the selection of the next-hop vehicle based on some important metrics of the participating nodes. We have used a real urban scenario to evaluate the performance of our models, by varying the vehicle density and measuring the percentage of packet delivery ratio, throughput and routing overhead during the transmission of data packets.

   5. The Dynamics of a Population of Healthy Adults, Overweight/Obese and Diabetics With and Without Complications in Morocco

      E. N. Mohamed Lamlili, Wiam Boutayeb, Abdesslam Boutayeb

      Abstract

      Between 1980 and 2014, the global prevalence of diabetes in adult people has nearly doubled, while the number of adult people living with diabetes rose from 108 to 422 million. In Morocco, from 2000 to 2018, obesity and diabetes increased, respectively, from 13.6 to 20% and from 6.6 to 10.6%. Half of the the adult population in Morocco are overweight. In this paper, a compartmental model is proposed to describe and analyze the evolution dynamics of an adult population from different stages (P: no overweight and no diabetes, W: overweight/obesity, D: diabetes without complications and C: diabetes with complications). Asymptotical local stability is proved, and normal foreword sensitivity index is used to guide decision makers in adopting a pragmatic strategy to control the prevalence of diabetes by acting on the flows into and out of the population living with overweight/obesity.

   6. Dermatologist-Level Classification of Skin Cancer with Level Set Method and Isolation Forest

      Khalid Bellaj, Soumaya Boujena, Mohammed Benmir

      Abstract

      Cancer classification (CC) by machine learning (ML), a research area that combines ML and scientific computing techniques, has experienced considerable growth in recent years. This article reviews existing and emerging approaches at CC and presents some well-known results in a unified framework. We study the combination of segmentation methods (SM) with ML techniques for solving partial differential equations (PDE). As a concrete example of SM, improved by ML, we present a method that uses isolation forest (IF) to reduce the computational cost in our SM while maintaining accuracy. Finally, experimental results show that our method significantly outperforms existing solutions for CC.

   7. A Generalized Freeman Chain Code for Offline Arabic Character Recognition

      Mohammed Kadi, Youssef Douzi, M’barek Nasri

      Abstract

      This work goes through several stages: First, the Arabic character is preprocessed and thinned. And a preliminary classification of the character is done according to the number of its loops and the nature, number, and position of its complementary parts (Dots, Hamza, etc.). Then, the Freeman chain code of 8 directions is generated according to a new innovative algorithm that can represent the internal parts of the Arabic character if they exist and without redundant information. After that, a normalization process based on the generalization of the code of 8 directions to a code of 24 directions is done. Finally, and for comparison, the two algorithms, traditional and proposed, are implemented separately to recognize Arabic characters. After 1023 tests on a set of 300 printed characters, isolated and distributed over 10 fonts, the recognition rate obtained was always higher than that obtained with the traditional algorithm, from 90.7 to 100%. Also, the number of font combinations giving good results almost doubled, and the execution time when normalizing the generated code was less.

   8. Performing Spectrum Sensing Using a Deep Learning Algorithm for Cognitive Radio

      Omar Serghini, Hayat Semlali, Asmaa Maali, Najib Boumaaz, Abdallah Soulmani, Abdelilah Ghammaz

      Abstract

      Cognitive radio network is an intelligent technology, used to solve the scarcity of the radio spectrum by allowing the unlicensed users to have access to the licensed spectrum. We study spectrum sensing methods in cognitive radio networks, the problem in the process of spectrum sensing that the detection rate of the the primary user signal is low in the environment of low signal-to-noise. A spectrum sensing algorithm based on multilayer perceptron is proposed which does not require a prior knowledge of the traffic characteristics of the licensed user. The performance of the proposed model is evaluated in terms of accuracy and loss for different epoch number also in terms of detection and false alarm probabilities. Simulation results have shown that the proposed approach provides good detection compared to the classic technique based on energy detection.

   9. A Retrospective on OOADARE as an Automated Object-Based Approach for Requirements Engineering

      Amal Khalil, Hajar Lamsellak, Zineb Bougroun, Mohammed Saber, Mohammed Ghaouth Belkasmi

      Abstract

      In agile methods, the communication between customer and development team is ensured by requirements, most often presented in an unstructured textual format, which frequently involves redundancy or non-precision information. This leads, in practice, to poor system quality, especially if we use classical approaches such as scenario-based approach. Yet, OOADARE approach is introduced in this way, using semi-structured text models in the form of user stories and constraint story cards (CSC), to automate the object-oriented transformation of requirements into a class diagram. However, the approach failed to capture all the elements needed to construct a correct and complete class diagram. This paper, at that point, proposes templates in natural language, which are part of the same perspectives of CSCs proposed by OOADARE, namely semi-structured text, to fill these gaps and ensure the completeness of the class diagrams thus generated.

   10. Computational Analysis of Human Navigation Trajectories in the VR Magic Carpet ™ Using K-Means

       Ihababdelbasset Annaki, Mohammed Rahmoune, Mohammed Bourhaleb

       Abstract

       In this research, we use unsupervised machine learning clustering techniques, notably K-means (Jain in Pattern Recogn Lett 31:651–666, 2010 [1]), to explore human navigation using the VR Magic Carpet (Berthoz and Zaoui in Dev Med Child Neurol 57:15–20, 2015 [2]). This is a variant of the Corsi Block Tapping task (CBT) (Corsi in Human memory and the medial temporal region of the brain. McGill University, 1972 [3]) that was carried out within the experimental framework of virtual reality. The participant’s trajectory was captured as raw spatial data and afterward kinematically evaluated. Our previous research (Annaki et al. in Digital technologies and applications. ICDTA 2021. Lecture notes in networks and systems, vol 211. Springer, Cham, 2021 [4]) found three distinct groups. However, the classification remained unclear, suggesting that they include both types of people (ordinary and patients with cognitive spatial impairments). Based on this premise, we used K-means to distinguish patients’ navigation behavior from that of healthy people, highlighting the most significant differences and validating the feature on which our previous analysis was based.

   11. The Management of Approaches in the Decentralized Architecture of the PBM

       Essaid Ammar

       Abstract

       The quality of service contract between the various stakeholders mainly aims to facilitate the presentation of information, and this interpretation allows on the one hand an easy to understand human reading, and on the other hand a possible processing by machines. After the development of these two input models, we deduce our output ontology which makes it possible to describe the agreements of the SLA, agreements make it possible to guarantee a certain quality of service to customer requests. In this paper, this artificial intelligence has required more sophisticated information management, hence the need for a global management and supervision approach that supports all these criteria (diversity of equipment, multimedia flow, complex system, user mobility, Terminal mobility, etc.). Network managers (operator, access provider, etc.) have understood that good management means mastering all the software and hardware resources of the network, hence the possibility of guaranteeing a certain quality of service for each customer range, which subsequently results in more agreement retention and customer gain.

   12. COVID-19 SEIAR Model with Sensitivity Analysis

       Mohamed Derouich, E. N. Mohamed Lamlili, Abdesslam Boutayeb

       Abstract

       Beside the unexpected toll of mortality and morbidity caused by COVID-19 worldwide, low- and middle-income countries are more suffering from the devastating issues on economic and social life. This disease has fostered mathematical modelling. In this paper, a SEIAR mathematical model is presented to illustrate how policymakers may apply efficient strategies to end or at least to control the devastating wide spread of COVID-19.

   13. Toward Multi-label Attribute Estimation on Multiple Faces Using CNN

       Mohammed Berrahal, Mostafa Azizi

       Abstract

       In the past few years, there has been a dramatic increase in the number of intelligent or cognitive applications. These applications can understand, learn, and interact with people in a more natural way than traditional applications. Many of these new applications are based on artificial intelligence (AI) technology, namely the ability of computers to learn and work on their own. One of the earliest examples of an intelligent application is the facial recognition system, based on deep learning models; this system can detect, estimate, and classify human faces based on facial attributes. Our approach is focusing on the estimation of attributes in images or videos. We train the first component, a YOLOv5 network, on Wider Face datasets, to detect faces, crop them, then pass the resulted data to the second component where we train a CNN multi-label network on CelebA datasets. The obtained model is capable of estimating two levels of face images, colored ones, and sketches. In addition, we make a comparison between YOLOv5 and SSD-ResNet (Single-Shot MultiBox Detector) face detectors.

   14. New Design of an X-Band 22 Patch Array Antenna with Circular Slots for Nanosatellites

       Nabil EL Hassainate, Ahmed Oulad Said, Zouhair Guennoun

       Abstract

       In order to meet the requirements in terms of size, weight, high gain, wide bandwidth and circular polarization of antennas used for nanosatellites, a 2\(\,\times \,\)2 circular polarization X-band patch array antenna with circular slots is designed for a 1U CubeSat. This article describes the antenna and presents its simulated performance. The antenna is formed by two layers of substrate, a planar array 2\(\,\times \,\)2 of patch truncated at the corner and having a cut circular slot located on the top, a ground plane encapsulated between the two layers and a feed network on the bottom. The overall performance of the antenna is improved by the introduction of cut circular slots on four antenna elements. The presented antenna is excited by a suitable feed network. It offers a bandwidth of 740 MHz in the frequency range of 8.60–9.34 GHz with a maximum gain of 13.30 dB. It has a return loss of − 27.70 dB and an axial ratio of 1.693 dB.