Latest issue

International Review on Computers and Software - November 2007 - Papers

Quality-driven Software Development within MDA Approach
A Reverse Engineering Approach for Specifying Semantic Web Service with Respect to MDA
Four Reasoning Models for C3 Metamodel
Behavior Metamodeling for Software Architecture Description
A Non-Linear Registration Method Guided by B-Splines Free-Form Deformations for Three-Dimensional Facial Reconstruction
Quality-on-Request Phonocardiogram Compression Algorithm Using the Discrete Cosine Transform: Comparison Between the Linear and the Lloyd Quantizers
Round Data Mailer Real-Time Protocol and its Message Delivery Performance
Seamless Handover and QoS Mechanisms for Efficient DVB-H and Mobile IPv6 Integration
Thermal and Visible Image Fusion: A Machine Learning Approach
Adaptive Image Watermarking Using Multiple Transforms
Optimized FPGA Mapping of a Bit-serial Square Root Operator with Minimum Output Delay
Application-oriented Testing of Embedded Processor Cores Implemented in FPGA Circuits
A Microcontroller Based System for an Automatic Control of pH and Acidity Degree of Chemical Products
Identification and Prediction of Wastewater Treatment Parameters

International Review on Computers and Software - Papers

Quality-driven Software Development within MDA Approach
by I. Dubielewicz, B. Hnatkowska, Z. Huzar, L. Tuzinkiewicz

Abstract - The paper proposes extension of MDA called Quality Model Driven Architecture (Q-MDA) which can be used for quality control in software develop¬ment process. The proposed Q-MDA approach, being an extension to the MDA, allows for considering quality requirements while conducting the design process. The main idea of presented extension consists of two elements. The first one is explicit decomposition of a model at given abstraction level on two interrelated models: software artifacts, which reflect functional require¬ments and quality model. The second element is simultaneous transformation of these two models. Such a procedure ensures that the transformation decisions should be based on the quality assessment of the created models.

Keywords: MDA, Quality Assurance, Quality Model.

A Reverse Engineering Approach for Specifying Semantic Web Service with Respect to MDA

by D. Amar Bensaber, D. Benslimane, M. Mimoun

Abstract - Among the many challenges of using web services are the problems of specification, search, discovery, selection, composition, integration, substitution and evolution. Semantics, especially as supported by the use of ontologies, and related Semantic Web technologies, are likely to provide better qualitative and scalable solutions to these requirements. To address the need for semantically defined web services, OWL-S and WSML have been proposed as competing semantic web service languages. Consequently, we propose in this work a model driven approach for semantic web service development. The methodology is divided into three main steps. In the first step we reverse engineered WSDL documents into UML profile models that enable the use of high-level graphical models as an integration platform for semantic web services. In the second step, suitable domain ontologies are used for the semantic annotation of the UML models. Finally, in the third step a conversion tool will generate automatically the OWL-S description from these UML models. The UML profile provides flexibility as it can expresses multiple semantic web service concepts.
Copyright © 2007 Praise Worthy Prize S.r.l. - All rights reserved

Keywords: WSDL, UML, MDA, Reverse engineering, Semantic web service.

Four Reasoning Models for C3 Metamodel
by A. Amirat, M. Oussalah

Abstract - The architecture is considered to be the driving aspect of the development process; it allows specifying which aspects and models in each level needed according to the software architecture design. Early Architecture Description Languages (ADLs), nearly exclusive, focus on structural abstraction hierarchy ignoring behavioural description hierarchy, conceptual hierarchy, and metamodeling hierarchy. In our approach these four hierarchies constitute views to appropriately “reason about” the architecture of a system described using our C3 metamodel. C3 is defined to be a minimal and complete architecture description language. In this paper we provide a set of mechanisms to deal with different levels of each type of hierarchy, also we introduce our proper structural definition for connector types used to instantiate any connexion elements deployed at the architectures and application levels.

Keywords: Conceptual, Configuration, Connector, Behaviour, hierarchy, Metamodeling, Structure.

Behavior Metamodeling for Software Architecture Description
by A. Bessam, M. Tahar Kimour, A. Melit

Abstract - Software architecture specification captures system structure, by identifying architectural components and connectors, and required system behavior by specifying how components and connectors are intended to interact. In order to have a complete system specification, a rigorous behavior description is needed. Many techniques for behavior specification exist. Behavioral concepts and their use in architectural specification are in a fast evolution and become so numerous, so it becomes difficult to learn and to handle them. Development of a software system requires developing or having a well fitting metamodel, and thus deciding what should be or should not be modeled about the system. For these purposes, we present in this paper, a generalized metamodel of behavioral aspects, that indexes the various architectural behavior concepts in classes, in a generic way. We show that our proposed metamodel allows having a general, a unified and an adaptable view of behavioral concepts required in software architecture description.

Keywords: Modeling framework, Behavior metamodeling, architecture description languages, component-based software architecture, extensibility.

A Non-Linear Registration Method Guided by B-Splines Free-Form Deformations for Three-Dimensional Facial Reconstruction

by A. Kermi, I. Bloch, M. T. Laskri

Abstract - The aim of facial reconstruction is to estimate the face of an unknown individual from the shape of his/her skull to aid in recognition and identification. This paper presents a 3D computerized facial reconstruction method based on recent techniques of medical imaging. It is based on a reference 3D image from which skull and skin are extracted, and an image of the skull of the unknown head for which the skin should be reconstructed. The facial reconstruction is obtained by deforming the 3D image of the reference skull to the one of the unknown skull. This transformation is based on a non-linear registration method guided by B-splines Free-Form Deformations (FFD). Then, the same deformation is applied to transform the reference skin to obtain a new skin that we consider near the unknown skin. This method has been evaluated on sets of skull/skin data, extracted from 3D-MRI of individual heads having similar anthropological features.

Keywords: 3D facial reconstruction, skull, skin, 3D-MRI, non-linear registration, Free-Form Deformations.

Quality-on-Request Phonocardiogram Compression Algorithm Using the Discrete Cosine Transform: Comparison Between the Linear and the Lloyd Quantizers

by R. Benzid, M. Benyoucef, L. Saidi

Abstract - In this paper, we propose an efficient compression method for phonocardiogram (PCG) signals. It is based on an automatic thresholding of the discrete cosine transform (DCT) coefficients obtained after transforming the original signal into the frequency domain. The threshold finding is achieved using the peak signal to noise ratio (PSNR) as the controlled quality criterion. The Linear quantization and the iterative Lloyd algorithms are compared when applied on the non-zero DCT resulting coefficients. In order to enhance the storage efficiency the lossless two-role encoder is used. The experimental results obtained on a set of normal and pathological different PCG waveforms, point out the efficiency of the developed method.

Keywords: Phonocardiogram Compression, Discrete Cosine Transform, Quality-on-Request.

Round Data Mailer Real-Time Protocol and its Message Delivery Performance

by M. Naghibzadeh, M. Sabeghi, S. Mirshokraie, H. Abachi

Abstract - The scope of real-time systems is ever expanding. New applications like automotive control, Internet multimedia and remote surgery are joining with traditional real-time systems like aircraft control and process control. The Nature of real-time applications is also changing towards distributed or even Grid-enabled ones. Communication protocols to handle the in-time delivery of messages between different processes of such application are an essential part of such systems. Token Ring, Controller Area Network, Time Division Multiple Access, and the newly designed Round Data Mailer are developed to fulfill the requirements of distributed real-time system’s communication. In this paper, we will focus on describing the characteristics of the Round Data Mailer and will compare its performance against both Token Ring and Controller Area Network protocols. Two important criteria of this comparison are end-to-end delay and freshness of delivered data. The results show that Round Data Mailer performs better and delivers fresher data to the consumers.

Keywords: Real-time communication protocol, Round Data Mailer, data freshness.

Seamless Handover and QoS Mechanisms for Efficient DVB-H and Mobile IPv6 Integration

by I. Miloucheva, M. Wetterwald, J. Mödeker, K. Jonas, D. Hetzer

Abstract - Seamless handover and QoS provision of applications in converged DVB-H and Mobile IPv6 networks are discussed. The goal is to support efficiently interactive mobile broadcast/multicast services using unidirectional broadcast media (DVB-H) combined with bidirectional mobile access technologies (UMTS, WLAN, WIMAX) in heterogeneous Mobile IPv6 environments. In particular, the focus is:
- Handover strategies based on usage of Link Layer Tunneling Mechanisms (IETF RFC 3077) for emulation of bidirectional connectivity of DVB-H access networks in heterogeneous Mobile IP environment;
- Network discovery and handover optimization for DVB-H access networks considering IEEE 802.21;
- QoS mechanisms of interactive multicast streaming and download services using DVB-H access networks.

Keywords: DVB-H, IPDatacast, seamless handover, QoS, Mobile IPv6, IEEE 802.21, Link Layer Tunneling, unidirectional connectivity, return channel.

Thermal and Visible Image Fusion: A Machine Learning Approach
by A. Mujahid Khan, M. Fayyaz, A. M. Gillani

Abstract - Both in military defense and civilian applications, an increasing interest is being shown in fusing information from infrared and visible sensors in order to get precise situational assessment. This paper presents a novel pixel-based infrared and visible image fusion algorithm. In this study, linear and nonlinear features in image data are dealt independent of each other. Image is first transformed to wavelet transform domain, splitting its linear features from nonlinear ones, and fusing linear and nonlinear features using different set of fusion rules. The technique successfully incorporates a linear classifier to fuse the linear features and an RBF kernel-based Principal Component Analysis to fuse the nonlinear features in data. Evaluation of the proposed technique through an image database of 32 registered image pairs shows that the proposed method is quite effective for online applications.

Keywords: Image Fusion, Thermal & Visible images, Support Vector Machines (SVM) and Kernel Principal Component Analysis (K-PCA).

Adaptive Image Watermarking Using Multiple Transforms
by S. A. M. Gilani, A. N. Skodras

Abstract - A novel scheme for the watermarking of color images is presented in this communication. The first objective is to find the most suitable alternative to RGB color space, which is known to be highly correlated. Color spaces linearly related to RGB color space with uncorrelated components are found to be most suitable for watermarking applications. Second objective is to make the scheme adaptive to the color image. This is achieved by keeping the PSNR in a predefined quality range, while tuning the watermark strength. The length of the watermark is kept constant for each of the color channels. In order to select the suitable space for watermark casting each of the color channels is transformed through Discrete Wavelet Transform and then the Discrete Hadamard Transform is applied to the approximations only. By this multiple transformation we are in a position to select the most appropriate coefficients for the watermarking, which are close to the low frequencies and uncorrelated in nature. For increased security the coefficients are randomized by means of a private key. Watermark detection is fast and blind, i.e. only the watermark generation and coefficient randomization keys are needed, and not the original image. The presented results demonstrate the robustness of the method against some common image processing attacks such as compression, scaling, uniform or gaussian noise addition, median filtering, cropping and multiple watermarking. Finally, the proposed scheme is also tested against StirMark benchmark.

Keywords: Adaptive Watermarking, Color spaces, DWT, DHT, Multiple transforms, IPR protection.

Optimized FPGA Mapping of a Bit-serial Square Root Operator with Minimum Output Delay
by S. Simard, J.-G. Mailloux, R. Beguenane

Abstract - We propose an FPGA implementation of a bit-serial square root operator based on an nonrestoring algorithm which optimally maps to the underlying hardware, and has minimal output delay. The most significant bit (MSB) of the square root is obtained right with the next clock cycle after input of the first bit of the radicand. Synthesis results for different radicand widths in the range of 8 to 64 bits targeting two different Xilinx families of FPGAs have shown this design to consume a number of 4-input LUTs at most twice the given width, and maximum clocking frequencies in excess of 100 MHz are possible in all cases. The proposed operator has been designed for use in advanced control loops, and has been found suitable for the intended purpose.

Keywords: Bit Serial, Square root, Nonrestoring, FPGA, Delay.

Application-oriented Testing of Embedded Processor Cores Implemented in FPGA Circuits
by M. Wegrzyn, A. Biasizzo, F. Novak

Abstract - This paper deals with application-oriented testing methods for FPGA circuits. These methods are useful both for manufacturing testing and the detection of the incorrect behaviour of applications in cases when Single-Event Upset (SEU) phenomena occur. After a brief introduction to the test approaches for the detection of SEU-induced faults in FPGAs, the paper focuses on the problem of testing embedded processor cores, and a new approach to the efficient instruction testing of microprocessor cores is presented.

Keywords: functional test, embedded processor-core test, single-event upset, built-in self-test.

A Microcontroller Based System for an Automatic Control of pH and Acidity Degree of Chemical Products
by L. Khouni, A. Khellaf, L. Saad-Saoud, D. Mokeddem

Abstract - In this work, a microcontroller based system for an automatic control of pH and acidity degree of chemical products is implemented. The acquisition chain consists of a commercial sensor TACUSSEL, an analogue signal processing unit and an Analogue to Digital Converter built in the microcontroller.
A micrometric magnetic field pump is used to titrate products. The microcontroller drives the different parts of the system in real time. It measures the pH values, and then computes the different points of the differential pH curve using the gradient method. The maximal value corresponds to the equivalence point.
Extensive analysis on known products such as the commercial vinegar, milk, acetic acid and water, shows that our results are in complete concordance with those using the Gran method.

Keywords: Microcontroller Based System, Equivalence point, pH control techniques.

Identification and Prediction of Wastewater Treatment Parameters
by S. Boulahbel, A. Khellaf, L. Saad Saoud , D. Mokeddem

Abstract - In this paper, neural networks are extensively used to identify and predict wastewater process parameters. Three methods are implemented to control the system operations. The difference neural network inputs are excluded in the first method and considered in the second one. The third one is implemented using the extended Kalman filter.

High performance is obtained by our approach, which considers input difference effects and, therefore, fits better to overcome the complex wastewater problem.

Keywords: Wastewater, Identification, Prediction EKF and Neural Networks.