Amin Alqudah

Stochastic-based robust dynamic resource allocation for independent tasks in a heterogeneous computing system

Heterogeneous parallel and distributed computing systems frequently must operate in environments where there is uncertainty in system parameters. Robustness can be defined as the degree to which a system can function correctly in the presence of parameter values different from those assumed. In such an environment, the execution time of any given task may fluctuate substantially due to factors such as the content of data to be processed. Determining a resource allocation that is robust against this uncertainty is an important area of research. In this study, we define a stochastic robustness measure to facilitate resource allocation decisions in a dynamic environment where tasks are subject to individual hard deadlines and each task requires some input data to start execution. In this environment, the tasks that cannot meet their deadlines are dropped (i.e., discarded). We define methods to determine the stochastic completion times of tasks in the presence of the task dropping. The stochastic task completion time is used in the definition of the stochastic robustness measure. Based on this stochastic robustness measure, we design novel resource allocation techniques that work in immediate and batch modes, with the goal of maximizing the number of tasks that meet their individual deadlines. We compare the performance of our technique against several well-known approaches taken from the literature and adapted to our environment. Simulation results of this study demonstrate the suitability of our new technique in a dynamic heterogeneous computing system. Calculating stochastic task completion time in heterogeneous system with task dropping.A model to quantify resource allocation robustness and propose mapping heuristics.Evaluating immediate and batch mappings and optimizing queue-size limit of batch mode.Analyzing impact of over-subscription level on immediate and batch allocation modes.Providing a model in the batch mode to run mapping events before machines become idle.

Face detection using Viola and Jones method and neural networks

Human face detection and recognition is a hot topic and an active area of research. It is common in several fields such as image processing and computer vision. It is the primary and the first step in wide range of applications such as face recognition, personal identification, identity verification, facial expression extraction, and gender classification [1]. In this paper, a multi stage model for face detection is integrated based on Viola and Jones algorithm, Gabor Filters, Principal Component Analysis, and Artificial Neural Networks (ANN). This model was trained and tested using CMU (Carnegie Mellon University) data set [2]. The model showed an enhanced performance in terms of face detection rate.

Rainfall estimation from spaceborne and ground based radars using neural networks

Neural network (NN) is a nonparametric method to represent the relation between radar measurements and rainfall rate. The relation is derived directly from a dataset consisting of radar measurements and rain gauge measurements. Tropical Rainfall measuring Mission (TRMM) Precipitation Radar (PR) is known to be the first observation platform for mapping precipitation over the tropics. TRMM measured rainfall makes a significant contribution to the study of precipitation distribution over the globe in the tropics. Ground validation (GV) is a critical component in the TRMM system. However, the ground sensing systems have quite different characteristics from TRMM in terms of resolution, scale, sampling, viewing aspect, and uncertainties in the sensing environments. In this paper a novel hybrid NN model is presented to train ground radars for rainfall estimation using rain gauge data and subsequently the trained ground radar rainfall estimation to train TRMM/PR observation based neural networks. This hybrid NN model provides a mechanism to link between gauges on the ground, the ground radar observations and the TRMM/PR observations. The dual-polarization radar measurements from a ground WSR-88DP site in Dallas-Fort Worth region and local rain gauge data will be used for the demonstration purpose. The performance of the rainfall product derived for TRMM PR is then compared against TRMM standard rainfall products. In addition, a direct gauge comparison study is done to examine the improvement brought in by this hybrid neural networks approach.

Control of variable speed drive (VSD) based on diode clamped multilevel inverter using direct torque control and fuzzy logic

The deployment of Direct Torque Control using fuzzy logic in variable speed drive that is based on multilevel inverter has increased by researchers in the last few years. A fuzzy logic in variable speed drive as type of controlling will be employed to control the speed and torque of industrial applications. Our proposed approach is based on Simulator (MATLAB/SIMULINK) that is used to identify the performance of Direct Torque Control in terms of Speed, Torque, Transient response and the current ripple. The goals are to reduce the overall current ripple and, hence to improve the transient response.

Precise and accurate decimal number recognition using Global Motion Estimation

Precise and accurate automatic recognition of decimal numbers is essential for many applications. Motion Estimation (ME) is a basic component in any video compression technique used to account for the temporal redundancy in image sequences. Global Motions are often modelled by parametric transformations of two dimensional images. The process of estimating the motion parameters is called Global Motion Estimation (GME). In this paper, we propose a new way of using GME for the purpose of off-line machine-print decimal digit. We show that the proposed approach is able to achieve very high recognition rates.

Efficient k-class approach for face recognition

Display Omitted A novel k-class approach system for accurate face recognition is introduced.The system is fully automatic: every step in the system is automatically performed.The system does not take long time to recognize a picture from a large database.The system achieves high recognition rates of 96.9% for Rank 1 and 99.5% for Rank 10. In this research work, a new k-class approach for efficient and accurate face recognition called kCAFRe is established. The kCAFRe system has three stages: preprocessing, training, and testing. In the training phase, four reference pictures (classes) are constructed for each person. During testing, the correlation coefficient (CC) is calculated between the picture under test and each of the reference pictures. For each person, one accredited class is chosen. Subsequently, the accredited class for the person with the highest CC is selected. Results were given for four image databases. A recognition ratio of 97% has been obtained for the Libor Spaceks set.

Fully automated smart wireless frost prediction and protection system using a fuzzy logic controller

A smart fuzzy logic controller system is presented to protect the crops from frost damage that occurs every year. The system is a fully automated system to predict the frost and to protect the crops using wireless sensor network technology. The sensors are used to collect crops data and transmit these data to the fuzzy controller. After that, the fuzzy controller will decide the proper action to be taken in order to protect the crops. The frost protection mechanism used here is a solid fuel burner that generates an artificial smoke cloud. The conducted simulations have shown that the system can successfully handle the frost problem at critical weather situations. This is achieved by keeping the ground surface temperature above the freezing temperature, and therefore saving the crops from being injured. In order to reduce the system energy consumption requirements, different approaches are presented during the simulation process.

Forward vehicle collision mitigation by braking system based on artificial bee colony algorithm

AbstractDecision-making is a crucial step in vehicles’ safety systems, which determines the right time for the system to intervene and to take the required action, depending on the current situation of the vehicle. The warning/intervention system becomes active when the driving situation considered unpleasant while still giving the control to the driver. Taking the action of warning, braking and determining the best time for this intervention are important decisions in collision mitigation systems. In this work, a novel system is designed based on artificial bee colony algorithm to enhance the decision-making in such systems. This work c oncentrates on collision mitigation by braking systems only, for front-to-rear accidents. This paper studies cases when the hosting vehicle approaches moving and stationary objects. This work is simulated, and the results are obtained and analyzed which proved the contribution to this work in reducing the collision speed and the stopping distance, in comparison with previous related works.