Ste

ANNALS OF THE UNIVERSITY OF CRAIOVA

Series: AUTOMATION, COMPUTERS, ELECTRONICS and MECHATRONICS
ISSN: 1841-0626

    Year 2007 Volume 4(31) no. 3

  1. DATA MINING DECISION TREES ALGORITHMS OPTIMIZATION (pp. 3-8)

    Laviniu Aurelian Badulescu
    University of Craiova, Faculty of Automation, Computers and Electronics, Software Engineering Department


    Abstract: Decision Tree classifiers are relatively fast compared to other classification methods and are easily interpreted and comprehended by humans. One major aim of a classification task is to optimize its classification accuracy. A good selection of an attribute used to split the data set at each Decision Tree node will optimize the accuracy of the classification. In this paper, we study the optimization of the Decision Trees induced with 14 attribute selection measures over four well-known databases in order to find the optimum attribute selection measure.
    Keywords: databases, decision trees, classification, error rates, machine learning.

  2. APPLICATION OF FRACTAL DIMENSIONS SPECTRUM IN MEDICAL SCREENING (pp.9-14)

    Daniela Alexandra CRISAN 1), Catalina POPESCU-MINA 2), Dumitru POPESCU 3), Justina Lavinia STANICA 1)
    1) Romanian-American University, Bucharest
    2) University of Bucharest
    3) University “Politehnica” of Bucharest


    Abstract: Fractal properties of objects concerned the attention of many researchers in the last decades; important results were obtained in medical screening field. The similarity between a breast tissue and a synthetically generated fractal suggests that fractal properties may be used in describing mammographic lesions. When categories a mammar anomaly, one important feature that the radiologist has to observe is the contour’s shape. His experience shows that a regular contour is associated to a benign case, while an irregular shape characterizes a malign lesion. So, it is expected that malign lesions will have higher dimensions that benign ones. The analyzed mammograms were grey-color images; the fractal image spectrum was used. Experiments were done on 30 cases of mammografic lesions BI-RADS 4 (which normally leads to biopsy); the results show that fractal dimension is a pertinent tool for classifying mammar lesions.
    Keywords: image processing, medical imaging, fractal dimension, box-counting, CADM

  3. TESTING STRATEGIES FOR MOBILE APPLICATIONS (pp.15-20)

    Eugen Dumitrascu, Nicolae-Iulian Enescu, Gheorghe Marian
    University of Craiova, Faculty of Automation, Computers and Electronics, Craiova, Romania


    Abstract: In this article, we present the specific features of mobile applications as distributed applications. We enumerate the series of evaluation’s indicators of quality for mobile applications. We also present the testing model IVV (Integration Verification and Validation) and the IADT method (Inspection Analysis Demonstration and Test). However, we refer especially to the functional type testing of GUIs (Graphical User Interfaces) of some software applications and in particular of a mobile application.
    Keywords: mobile application, verification strategy, IVV, IADT

  4. A SEISMIC DATA ANALYSIS METHOD USING FUZZY NEURAL NETS (pp.21-26)

    Ganea Eugen, Brezovan Marius, Ganea Simona
    Department of Software Engineering, Faculty of Automation, Computers and Electronics, University of Craiova


    Abstract: Seismic stratigraphy is a complex geological processing of seismic data. A seismic facies is a stratigraphic region that has a characteristic reflection pattern that is distinguishable from those of the other areas. The goal of this paper is to describe a rule based system in order to detect regions in three-dimensional seismic images. For this, the attributes which are extracted from the seismic images will be interpreted through a rule-based classifier. The method for detecting these regions will use the texture characteristic and the model of fuzzy neural networks. Seismic facies are identified in a volume of seismic data and a set of initial textural attributes representative of the volume of seismic data is calculated.
    Keywords: seismic image, rule-based system, seismic data analysis, texture, fuzzy logic, neural nets.

  5. IMPROVING E-LEARNING and KNOWLEDGE MANAGEMENT IN SME-S (pp. 27-32)

    Ileana Hamburg , Thorsten Busse
    Institute for Work and Technology, FH Gelsenkirchen, Germany


    Abstract: Through including e-Learning, particularly 2-0 and social knowledge management for example in Communities of Practice (CoP) could improve training processes and facilitate a fast transfer of knowledge by meeting small and medium sized-companies (SMEs) needs of flexibility, to be oriented to their business and be easier integrated into their workflow with a suitable personalized content at a lower cost to the organization can be achieved. In this paper after the presentation of some reasons of the low use of e-Learning in European SMEs examples of training strategies including e-Learning 2.0 for improving this situation and the building of CoP are given. It is important, however, to help SMEs to develop and to implement such training strategies because many of them have not always the resources and knowledge to do this alone.
    Keywords: e-Learning 2.0, SMEs, Strategies, Guidelines

  6. INTRODUCTION IN OBJECT TRACKING (pp.33-38)

    Andrei Ingeaua, Dumitru Ingeaua, Stefan Udristoiu
    University of Craiova


    Abstract: This article in intended as a short reference for some of the techniques that can be used in object tracking. Object tracking has many application in various fields starting from robotics, military applications, etc. A real interes for us is tracking the object in real time because of it’s intensive use in visual servoing. This article will present some basic notions related to object tracking and video sequences, a short theoretical breviar for principal methods and algorithms like HMM(Hidden Markov Model), and Viterbi algorthm.
    Keywords: vision, image recognition, object tracking.

  7. SCALABLE APPLICATION WITH WINDOWS OVERLAPPED I/O COMPLETION PORTS (pp.39-44)

    Marin Lungu
    Computer Dept., Faculty of Automation, Computers and Electronics University of Craiova


    Abstract: The paper will address the thorny issue of building a truly scalable client / server architecture using Windows Sockets. A comparative approach is taken towards the possible architectures for writing such a server, the limitations involved in each alternative are presented and the reasons behind choosing the Input Output Completion Ports (IOCP) solution are given. Some subtle details are pointed regarding the steps taken for tuning the server performance level by enhancing the Operating System’s throughput. As an illustration, the solution presented implements the server architecture described and two separate messaging clients built for the Pocket PC platform and the Windows platform.
    Keywords: Winsock Programming, IOCP, Client/Server Architectures, Communication networks.

  8. ONLINE CERTIFICATE STATUS PROTOCOL FOR HIGH-VOLUME PKI ENVIRONMENTS (pp.45-50)

    Marius Marian
    Department of Automation, University of Craiova


    Abstract: More and more human activities and businesses are now common to take place via Internet. The experience has proved that most network-related applications are affected by security attacks given a relatively moderate amount of time. Public-key cryptography permits to easily obtain integrity and confidentiality for the data exchanged between network applications, and also authentication of the involved parties. The verification of public-key certificates is one essential step when using public-key technology. The Online Certificate Status Protocol (OCSP) is meant to provide a timely and secure solution of getting certificate status information. The paper presents an OCSP implementation respecting most of the standardized lightweight OCSP profile, meant to accomodate the certificate revocation checking service for large scale public-key infrastructures.
    Keywords: public-key infrastructures (PKI), online certificate status protocol (OCSP), certificate validation.

  9. E-LEARNING SYSTEM COMPONENTS. CASE STUDY (pp.51-56)

    Adrian Neatu, Dan Tusaliu, Robert Popescu
    University of Craiova, Faculty of Automation, Computers and Electronics, Computer and Communications Engineering Department


    Abstract: Our society has become a very dynamic one nowadays. Everything is constantly improved, amended, adapted to the newly developed technologies. It would only seem natural for the learning process to undergo the same modifications. The module presented in this paper could be integrated into a larger e-learning project, its goal being that of adding information to existing electronic courses and facilitating the dissemination via the Internet.
    Keywords: e-learning, ICT, open source, graphical annotation, Java.

  10. IMPLEMENTATION OF AN HEURISTIC ALGORITHM FOR EVALUATING FRAGMENTATION IN DISTRIBUTED DATABASES (pp.57-62)

    Adrian Runceanu
    University Constantin Brâncusi Târgu-Jiu


    Abstract: The design of distributed database is an optimization problem and the resolution of several sub-problems as data fragmentation (horizontal, vertical, and hybrid), data allocation (with or without redundancy), optimization and allocation of operations (request transformation, selection of the best execution strategy, and allocation of operations to sites). There are some different approaches to solve each problem, so this means that the design of the distributed databases becomes hard enough. There are many researches connected to the data fragmentation being presented both for relational database and respectively object-oriented database. The paper presents the implementation of a heuristic algorithm conceived before that uses an objective function who takes over information about the administrated dates in a distributed database and it evaluates all the schemes of the database vertical fragmentation.
    Keywords: vertical fragmentation, distributed databases.

  11. COMPARISON OF SHAPE-BASED IMAGES RETRIEVAL METHODS (pp.63-70)

    Andreea Soimu*, Anca Loredana Ion**, Dorian Cojocaru*
    University of Craiova, Faculty of Automation, Computers and Electronics *Mechatronics Department, **Software Department


    Abstract: Shape is an important visual feature and it is one of the primitive features for image content description. However, shape content description is a difficult task. Because it is difficult to define perceptual shape features and measure the similarity between shapes. This paper presents methods for content-based visual retrieval on the shape characteristic inside an image database. More shape signatures have been used to derive Fourier descriptors, all these being described in the paper. A new function was also suggested in order to improve the existent signature methods, this being called the area signature.
    Keywords: image processing, shape, Fourier descriptors, geometric moments

  12. STACKS ARCHITECTURE OR FILE REGISTER ARCHITECTURE IN THE HIGH-PERFORMANCE EMBEDDED CELLULAR COMPUTING? (pp.71-76)

    Marius Stoian , Gheorghe Stefan
    Department of Electronics, Politehnica University of Bucharest


    Abstract: Stack oriented architectures are compared with register file oriented architectures in order to decide what is the best for building cellular programmable machines, used on SoC as a high performance embedded computing machine. Area & power vs. computing performance are investigated considering two kind of ``users": (i) a compiler, which is another machine, and (ii) a human mind, writing hand coded programs. We conclude that the a good candidate as the most appropriate architecture for a processing element in the embedded cellular computing is the stack oriented architecture.
    Keywords: Stacks architecture, file register architecture, embedded cellular computing

  13. A NEW RAISING NICHE: MOBILE SOFTWARE. AN EXAMPLE (pp.77-82)

    Dan-Costin Tusaliu, Adrian-Gabriel Neatu, Cosmin Selaru
    University of Craiova


    Abstract: According to the Mobile Games report, the market for mobile games continues to expand quickly, with half of all U.S. wireless subscribers now owning phones capable of downloading games. A full 27 percent play games on those devices, including purchased downloads as well as free demos and pre-loaded games, compared to 20 percent last year. A mobile game is a video game played on a mobile phone, smartphone, PDA, handheld computer or any type of handheld or wireless device. Mobile games are played using the technologies present on the device itself. For networked games, there are various technologies in common use. Examples include text message (SMS), multimedia message (MMS) or GPRS location identification. More common, however, are non networked applications that simply use the device platform to run the game software. The games may be installed over the air, they may be side loaded onto the handset with a cable, or they may be embedded on the handheld devices by the OEM or by the mobile operator. This paper is focused in the development of a golf game using J2ME technology.
    Keywords: Java, Mobile Software, Game, Engine, Interface, Pixel, Graphics

  14. SIMULATIONS OF SOME FUNDAMENTAL TERNARY FUNCTIONAL UNITS (pp.83-88)

    Oleg Cernian
    University of Craiova Computer and Communications Engineering Department


    Abstract: The future development of computer technology implies approaching of non-binary logic as a possible alternative for nowaday binary logic. In this work it is discussed simulation of different fundamental sequential ternary logic devices on a ternary simulator based on HDL-Aldec platform. All presented structures were captured and the appropriate functional experiments were run on that ternary simulator proving the correctness of their operation.
    Keywords: Ternary logic, Ternary simulation, Tristables, T-type gates, Ternary registers, Ternary frequency dividers

  15. OFFSET VOLTAGE AND CAPACITOR MISMATCH EFECTS IN PRECISION AMPLIFIER (pp.89-94)

    Andrei DANCHIV*, Mircea BODEA#, Claudius DAN§
    *Infineon Technologies Romania
    #, § “POLITEHNICA” University of Bucharest, Electronics, Telecomm. and IT Dept.,


    Abstract: This paper develops an analytical model for a standard topology, fully differential, switched capacitor amplifier; including the base amplifier offset voltage and common mode range, and capacitor mismatch effects. The amplifier is designed in a 0.6 µm process and the analytical model accuracy is compared with the simulation results.
    Keywords: offset voltage, fully differential, correlated double sampling

  16. A CMOS CONTINUOUS-TIME CURRENT MODE FILTER TECHNIQUE FOR LOW SUPPLY VOLTAGE APPLICATIONS (pp.95-100)

    Elena Doicaru*, Mircea Bodea**
    *University of Craiova, Faculty of Automation, Computer Science and Electronics
    ** POLITEHNICA University of Bucharest, Faculty of Electronics, Telecommunication and Information Technology


    Abstract: This paper presents a current mode (translinear) technique for low voltage, continuous-time, analogue filter implementation. Firstly, the state-space description of current-mode filter is presented. Then the current-mode integrator based on translinear approach is presented and analyzed. Finally, as a test vehicle for the proposed technique, a low voltage, third order, continuous-time filter, for low power applications structure implementation is described.
    Keywords: Analogue signal processing circuits, translinear circuits, CMOS analogue filters, low power and low voltage circuits.

  17. ABOUT GM-C HIGH FREQUENCES CONTINOUS-TIME ACTIVE FILTER REALIZATION. EXEMPLE IN THE CMOS TECHNOLOGY.(pp. 101-106)

    Elena Doicaru, Vintila Filipescu
    University of Craiova, Faculty of Automation, Computer Science and Electronics


    Abstract: In this paper are presented the state-space description of the Gm-C filter and the modality to synthesize these filters in order to obtain good performances. A CMOS transconductor in the 1mm CMOS process for high frequencies is also presented. The transconductor has a good linearity (1% relative error for 2Vp-p input signals with a 5V power supply) and high dc gain (ł40dB) combined with a large band to avoid errors in the filter characteristic. The transconductor has also a very good dynamic range. The good performances of transconductor were emphasized by SPICE simulation. With this transconductor a low pass filter was synthesised. The filter performances were also emphasized by SPICE simulation.
    Keywords: transconductor, high frequency, active Gm-C filters.

  18. A VERSATILE SRAM MEMORIES TESTER (pp. 107-112)

    Vintila Filipescu, Elena Doicaru
    University of Craiova, Faculty of Automation, Computer Science and Electronics


    Abstract: This paper presents a tester able to verify the correct work of the static RAM memories with a maximum capacity of 512Kwords x 8bits. The wide range of work frequencies this can use, the blocking of the work scheme not only at the end of the test but also in the case of a fault, the visualizing of the address logical combination where the fault was detected, all these made the presented tester be a very useful tool both in industrial and educational field.
    Keywords: SRAM, memory, block, clock, testing, fault.

  19. PROBABILISTIC MODELS FOR SPEECH AND SOUND ANALYSIS (pp.113-118)

    M. Vacher1, D. Istrate2, J.F. Serignat1
    1LIG, UMR 5217, CNRS-INPG, Grenoble, France
    2RMSE-ESIGETEL, Fontainebleau, France


    Abstract: Sound signal carry rich information which can be extracted and used by analysis systems in different modes. The medical remote monitoring systems may help elderly people to live home in security. We have already proposed an environmental sound analysis algorithm to detect distress sound or sentences. The sound and speech recognition modules based on Gaussian Mixture Models and Hidden Markov Models are detailed. A real time implementation on an embedded PC is proposed and evaluated. This implementation is flexible either for the hardware target (desktop, laptop or embedded PC) or on the alarm sending (E-Mail, SMS or remote monitoring center).
    Keywords: Signal processing algorithms, Gaussian distributions, Markov models, Pattern recognition, Real time computer systems, Embedded systems, Speech Analysis

  20. IMPROVED TEMPERATURE BEHAVIOR CMOS INTEGRATED VOLTAGE REFERENCE NANOSTRUCTURE (pp.119-124)

    Cosmin Popa, Anca Manolescu, Anton Manolescu
    Faculty of Electronics, Telecommunications and Information Technology, University Politehnica of Bucharest


    Abstract: A new high precision superior-order curvature-corrected integrated voltage reference nanostructure will be presented. In order to improve the temperature behavior of the circuit, a double differential structure will be used, implementing the linear and the superior-order curvature corrections. An original CTAT (Complementary To Absolute Temperature) voltage generator will be proposed, using exclusively MOS transistors biased in weak inversion for a low power operation of the integrated nanostructure, having two great advantages: an important reducing of the circuit silicon area and an improved accuracy (matched resistors being replaced by matched MOS active devices). The superior-order curvature-correction will be implemented by taking the difference between two gate-source voltages of subthreshold-operated MOS transistors, biased at drain currents having different temperature dependencies: (Proportional To Absolute Temperature) and , respectively.
    Keywords: temperature dependence, superior-order curvature-correction technique

  21. DIGITAL DIFFERENTIAL PROTECTION RELAY (pp.125-130)

    Traian-Titi Serban
    University of Craiova, Faculty of Automatics, Computers and Electronics


    Abstract: The paper presents an original implementation of a digital differential protection relay for single phase circuits, exhibiting additional features for a better protection of the operator, of the load and of the power line. The main hardware modules and some software routines are described
    Keywords: digital relay, differential protection, overcurrent protection, current transducer, insulated instrumentation