Welcome to Journal of Theoretical and Applied Information Technology

Author:

Source:

Abstract

The major source of communication in present world is digital media. As it is quite easy to manipulate a digital media, it becomes essential to protect the digital information by legitimate means. Digital Watermarking has evolved as one of the latest technologies for digital media protection. Many techniques based on spatial and frequency domain have been developed in the recent past and are being used for effective watermarking. However, there is always a tradeoff between robustness and imperceptibility features of watermarking offered by these techniques. This paper offers a technique based on Backpropagation Neural Network to train a given cover image to produce a desired watermark image. At the end of the training, the entire trained neural network weights has been successfully hidden within the cover image itself. This makes it possible to supply only the cover image without any external weight files. By extraction techniques, the weights can be derived from the cover image and used to reconstruct the trained Neural Network again which in turn converts the cover image into desired watermark image. The technique for hiding the weights into the cover image has been designed in such a way that it does not produce visual deterioration of the original cover image. This method is extra secure as it leads to watermarking indirectly.

Author:

Source:

Abstract

There are a number of different video compression standards. Most of them are variations of the work of two different committees. The International Telephony Union’s Specialist Group on Coding for Visual Telephony and the International Standards Organization’s motion picture expert group, known as MPEG. The most popular standards are the ISO’s MPEG-2 and the ITU’s H.261. MPEG-2 has been chosen to simulate video application. We study link utilization, WAN clouds, and message sources for compressed data and uncompressed data. The study includes Video application. . The bandwidth used for simulation runs depends on the type of application under study. The bandwidth is divided into five percentages, 50%, 60%, 70%, 80%, and 90% (traffic intensity) for the purpose of study.

Author:

Source:

Abstract

Sending information secretly and communicating covertly have been of great interest for ages. On the other hand, text documents have been widely used and consequently various methods for hiding information in texts (Text Steganography) have been developed so far. In this paper a new method is proposed for hiding information in Persian and Arabic Unicode texts. In Persian and Arabic, some letters are connected together in a word. But there are two characters, zero width non joiner (ZWNJ) and zero width joiner (ZWJ) characters, which are respectively prevents the Persian and Arabic letters from joining or forces them to join together. In this method by using these two special characters, the information is hidden in Persian and Arabic Unicode text documents. This method has a high hiding capacity because it hides one bit in each letter. Also this method does not make any apparent changes in the original text and have a perfect perceptual transparency.

Author:

Source:

Abstract

Clustering algorithms gives general description of the clusters listing number of clusters and member entities in those clusters. It lacks in generating cluster description in the form of pattern. Deriving pattern from clusters along with grouping of data into clusters is important from data mining perspective. In the proposed approach reduct from rough set theory is employed to generate pattern. Reduct is defined as the set of attributes which distinguishes the entities in a homogenous cluster. It is observed that most of the remaining attributes in the cluster has same value for their attribute value pair. Reduct attributes are removed to formulate pattern by concatenating most contributing attributes. Proposed approach is demonstrated using benchmarking mushroom dataset from UCI repository.

Author:

Source:

Abstract

Neural networks (NNs) have been increasingly used in recent years for the solving complex nonlinear problems. NNs are seen as an attractive alternative to process based modeling approaches, as they are able to extract an underlying relationship from the data when knowledge of physical process is lacking. The paper evaluates the predictive power of a model, which emulates an army commander on the battlefield when encountered with various situations, using two different neural network configurations – the multilayer perceptron (MLP) and the probabilistic neural network (PNN). The proposed model may prove effective to defence scientists and commanders for Battlefield decision making, strategy development and resource management. Empirical results support the potential of PNN as a better classifier, in training data as well as test data, compared to MLP.

Author:

Source:

Abstract

Reactive power dispatch (RPD) is one of the important tasks in the operation and control of power system. This paper presents a Differential Evolution (DE) - based approach for solving optimal reactive power dispatch including voltage stability limit in power systems. The monitoring methodology for voltage stability is based on the L-index of load buses. The objective is to minimize the real power loss subjected to limits on generator real and reactive power outputs, bus voltages, transformer taps and shunt power control devices such as SVCs. The proposed algorithm has been applied to IEEE 30-bus system to find the optimal reactive power control variables while keeping the system under safe voltage stability limit and is found to be effective for this task. The optimal reactive power allocation results obtained using DE are compared with other methods. It is shown that the objective function value is less than those of other methods.

Author:

Source:

Abstract

Most current intrusion detection systems are signature based ones or machine learning based methods. Despite the number of machine learning algorithms applied to KDD 99 cup, none of them have introduced a pre-model to reduce the huge information quantity present in the different KDD 99 datasets. Clustering is an important task in mining evolving data streams. Besides the limited memory and one-pass Constraints, the nature of evolving data streams implies the following requirements for stream clustering: no assumption on the number of clusters, discovery of clusters with arbitrary shape and ability to handle outliers. Traditional instance-based learning methods can only be used to detect known intrusions, since these methods classify instances based on what they have learned. They rarely detect new intrusions since these intrusion classes has not been able to detect new intrusions as well as known intrusions. In this paper, we propose some clustering algorithms such as K-Means and Fuzzy c-Means for network intrusion detection. The experimental results obtained by applying these algorithms to the KDD-99 data set demonstrate that they perform well in terms of both accuracy and computation time.

Author:

Source:

Abstract

This paper presents a reinforced genetic approach to a defined d-resource system optimization problem. The classical evolution schema was ineffective due to a very strict feasibility function in the studied problem. Hence, the presented strategy has introduced several modifications and adaptations to standard genetic routines, e.g.: a migration operator which is an analogy to the biological random genetic drift.

Author:

Source:

Abstract

The sensorless DTC of Brushless AC (BLAC) machine using Luenberger observer is proposed in this paper. In Direct Torque Control (DTC), accurate rotor position information is not essential. The speed is estimated by Luenberger observer, which used to improve the performance of dynamic tracking and accuracy of the whole system. The speed regulation is realized by a conventional PI regulator. Simulation results shows small overshooting and good dynamic of the speed and torque, low ripples in torque and flux. So we verify the effectiveness of the proposed observer to monitor the drive.
 

Author:

Source:

Abstract

Principal Component Analysis (PCA) and Independent Component Analysis (ICA) are the techniques that deal with extracting the independent components from linear mixtures of Gaussian and non-Gaussian data at the input respectively. PCA is a classical method that deals with the second order statistics of data. It is also known as Karhunen-Loeve Transform or the Hotelling Transform in some application areas. ICA is a generalization of PCA that takes into account the higher order statistics also. This paper presents a simplified view of information theoretic approaches to the problem of Principal Component and Independent Component Analysis. With the help of these techniques we find a linear representation of multivariate data so that the components are as statistically independent as possible. Such representations capture essential features of the data in many applications. This paper summarizes major approaches that are based on information theoretic concepts and includes the applications of ICA.

Author:

Source:

Abstract

In this paper, the authors present the induction motor speed control with DTC (direct torque control). The DTC in its conventional form uses algorithms to select the components of the voltage inverter. This paper proposes to replace the conventional selector switches statements of the voltage inverter by a selector based on neural networks ANN (Artificial Neural Network), which is able to manage in the same way the switches states, without resorting to complex programming. The speed loop regulation is carried out by a fuzzy controller giving exceeding performance in comparison with a classic PI regulator. The design of the controller is based on the experience without knowing the mathematical model of the Motor. The simulation results in Matlab/Simulink show that the speed static error and accelerating error is near zero, also the torque and flux ripples are respectively in the range of 2.5% and 5%.

Author:

Source:

Abstract

The core aim of Agent Based University Planning project of Sakarya University is designing and implementation of a system that is based on Distributed Artificial Intelligence technique for scheduling and planning to face the problems that often takes place in a university administration division. This report particularly focuses on course planning, resource allocation, class timetabling, and office allocation, exam scheduling and other administrative and organizational bureaucracy of a university. Our goal is to define a hierarchical multi-agent architecture for the university administrative functions and design and implement a personnel agent for each entity stumble upon in that division. Designing information storage and distribution agents for the university environment is the key objective of this project.

Author:

Source:

Abstract

The day is not far away, where, computer’s peripherals performing tasks by taking commands from the most natural form of communication: THE HUMAN VOICE. Yes, this idea of developing a voice operable hardware device by using the speech recognition technology is to develop a basic application which demonstrates a hardware chip responding to the commands given by voice along with a GUI. Several years of research has already been done on the speech recognition technology and in our exploration we found that all the speech-recognizing engines were based on the HMM concepts although they may be written in different programming languages. HMM is the “Hidden Markov Model”. This concept aims at building and manipulating Hidden Markov Model (HMM). HMM is primarily used for speech recognition applications. After exploring the options available for selecting a speech engine, which recognizes words, the best results were observed in a program written in Java. To show the hardware operation we have made use of the 8051 micro controller and embedded “C” code for its operation. So by using this as foundation we can build many applications like for ex: operating a printer through voice and any other output device connected to the computer. The advantage of our work in this paper is that our application is speaker-independent that can recognize continuous human speech regardless of the speaker and that can continually improve their vocabulary size and recognition accuracy.