Publications

Recently, multimedia streaming applications are deployed anywhere (homes, campuses, etc.). Delivering these applications over mobile ad hoc networks is a challenging task: first, mobile ad hoc networks are infrastructure less, with dynamic topology, and limited resources. Secondly these applications have strict quality of service (QoS) requirements given with delay, jitter, bandwidth, loss rate, etc. Several architectures have been developed. However, an important part like QoS in routing level is still omitted. In this paper, many different coding techniques for video streaming are reviewed and two routing protocols (reactive and proactive) are used to evaluate which of them can improve QoS for multimedia streaming applications. Results show that reactive protocol performs better than proactive protocol in terms of throughput and network load in high mobility, but roles are reversed in terms of loss rate and network load for large-scale networks. PNSR and Jitter are influenced by nodes numbers independently of used protocol.

La fiabilité figure parmi les objectifs les plus importants pour les fabricants en tant qu’indicateur de performance et en plus, elle est devenue un argument commercial important dans l’impact économique du cycle de vie et un paramètre clé de la qualité et d’aide à la décision dans l’étude de la plus part des composants, produits et processus. A ce propos, l’évaluation de la fiabilité des produits en phase de conception occupe une place de choix. Dans cet ouvrage, nous avons proposé une démarche d’anticipation des risques en conception qui s’articulent principalement autour du modèle DIC « outil de gestion de la connaissance » et leurs méthodes supports « outils SDF ». Cette méthodologie permet au concepteur de comprendre la conservation d’une valeur opérationnelle dans le temps qui est cependant associée à un risque ou une probabilité de défaillance qu’il faut tenir compte dès la phase de conception du produit. A cet effet nous avons procédé à des essais accélérés afin de capitaliser des connaissances sur ces risques potentiels.

Prediction of effective properties for multiphase composite is very important not only to analysis and optimization of material performance, but also to new material designs. In this paper, the effective elastic property of some complex particulate composites is analyzed and compared with numerical results, demonstrating the validity of the proposed approach. We propose the equivalent morphology concept for the numerical homogenization of random composites. In this study, this concept is extended for complex material. A home script based on Python codes is made to automate the generating of Representative volume element with various volume fraction.

The objective of this work is to present an analysis of RPL Routing protocol for Low-Power and Lossy Networks (LLNs) in the Internet of Things (IoT) and to determine the adequate working mode between storing and non-storing. Objects in LLNs networks have limited processing power such as battery and memory. The study examined the behavior of RPL towards some number of quality of service parameters in Wireless Sensor Networks (WSNs). Simulations are conducted under ContikiCooja Simulator. Results show that non-storing mode has better results in terms of delay and loss rate, but the storing mode consumes less energy.

The Tinibaouine region, located in north-eastern Algeria on the borders of the Batna-Belezma Mountains, is characterized by a semi-arid to arid climate with an average annual rainfall not exceeding 465 mm and an average annual temperature of around 22 C. This region is characterized by the cultivation of apricots as essential crop followed by that of olives, whose plots are all irrigated with the Tinibaouine spring water. These are 450 Ha of trees for apricot and 108 Hectare for olives which constitute the principal revenue of the citizens of this small village. This paper estimated the crop reference and actual evapotranspiration (ETO) respectively and the irrigation water requirement of apricot trees and olive trees. The long recorded climatic data, crop and soil data, effective water allocation and planning, the information about crop water requirements, irrigation withdrawals were computed with the Cropwat model which is based on the United Nations’ Food and Agriculture Organization (FAO), the Penman-Monteith method was used to estimate ETo. Crop coefficients (Kc) from the phenomenological stages of apricot and olive were applied to adjust and estimate the actual evapotranspiration ETc through a water balance of the irrigation water requirements (IR). The results showed that the annual reference evapotranspiration (ETO) was estimated at 3.71 mm/day. The irrigation requirements were estimated at 35800 m3/hectare for apricot, 6980 m3/hectare for olive, also Irrigation needs estimated on land at 14185, 05 m3/hectare for olive and apricot.

Temperature, solar insolation, shading and faults affect the performance of the photovoltaic array. Often, the PV arrays get shadowed, completely or partially, by the passing clouds neigh boring buildings, towers or by trees, and other utilities. The situation is of a interest in a case of the large PV power plants. In the case of the shading the characteristics of the PV module are more complex with the several peak values. Under such conditions, it is very difficult to determine the maximum power point (MPP). MATLAB-programmed modelling and simulation of photovoltaic module is presented here, by focusing on the effects of partial shading on the output of the photovoltaic (PV) module and Faults Bypass Diode. The proposed models facilitate simulating the dynamic performances of PV-based power systems and have been validated by means of simulation study. The southern part of Algeria, where the experimental system is mounted, is particularly well appropriate to photovoltaic systems. To evaluate the effectiveness of the proposed model, experiments have been conducted to compare the experimental and simulated current-voltage (IV) and power-voltage (PV) curves of a PV system under some predefined partial shading and faults bypass diode, using different PV technologies such as mono-crystalline and multi-crystalline.

Recently, researches in the interventional microrobots have taken the lion’s share in the field of biomedical devices. The aim of biomedical microrobots is to reach inaccessible areas of the human body and deliver drugs in high position. This work presents a new approach to elaborate a new physics-based model for novel self-propelled swimming microrobots. The robot is composed of an ellipsoidal head and hybrid tail that are propelled by a joint polymer metal composite actuator. Green’s function is used to solve the coupled elastic/fluid problems caused by the vibrating hybrid tail in a fluid. This method allowed producing the velocity of microrobot. For the control of the swimming microrobot in hazardous environment, the flatness-fuzzy-based control strategy is developed to eliminate the effect of nonlinear model and to generate the optimal trajectory of flat outputs. The fuzzy technique is aimed to adjust the law control gains in real time for improving the precision of the proposed microrobot in tracking the desired trajectory in fluid. The multi-objectives genetic algorithm is employed to optimize both the reference trajectory and the design parameters in order to enhance the time response and to minimize the dynamic tracking error of the trajectory. To achieve this, a numerical model based on accurate solutions of Navier–Stokes equations is developed. The results of the simulation show that the proposed design with ellipsoidal head gives better performance in comparison with that achieved by the conventional structure.

This paper addresses the problem of fault diagnosis from observed data containing missing values amongst the inputs. In order to provide good classification accuracy for the decision function, a novel approach based on support vector machine and extreme learning machine is developed. SVM mixture model is used to model the data distribution, which is adapted to handle missing values, while extreme learning machine enables to devise a multiple imputation strategy for final estimation. In order to prove the efficiency of our proposed method, we have developed the classifier using the condition monitoring observations from milk pasteurisation data. The experiments show that the proposed algorithm gives improved results compared to recent methods, essentially if the number of missing data is significant. The results show that our approach can perfectly detect dysfunction, identify the fault, and is strong in …