• Research makes knowledge grow

RESEARCH PROJECTS BUSINESS ENGINEERING


Refactoring Information Systems and New Cryptographic Systems

Refactoring Information Systems  


Overview

“The only constant is change.” This statement is especially apt in software development. For this reason there are now functions for refactoring in modern programming environments e.g. eclipse. These functions are currently directed exclusively at the consistent alteration of the programme system and disregard any existing databases that are generated and administered by this system.

The project develops concepts, theoretical foundations and prototypical tools which make it possible to include the consistent migration of available databases in the refactoring process.  In this way the modern technologies of agile software development are also made available and utilisable for the field of information systems.

Starting position

Agile technologies for developing programmes are asserting themselves more and more. However the large sector of information systems with very large databases is excluded from these developments to a considerable extent, since agility inevitably led to the migration of databases. Migrations like these can sometimes take hours or days. Frequent migrations – a consequence of agile development – severely interfere with the operational level.

For this reason the reconfiguration of information systems is usually confined to programmes. Data models and the data which go with them are not affected. Therefore system models and data models diverge more and more from each other. This situation reduces productivity in development in the long term and plays a considerable role in the premature “ageing” of the systems. There is therefore a huge demand for technologies and tools which support the frequent migration of productive databases without interfering with the system runtime, but these are not available.

Own preliminary work

Prof. Dr. Löwe has published numerous publications in the field of software reengineering. As a result of his former work as head of the Department of Software Reengineering at the Fraunhofer Institute for Software and System Technology (ISST) he has diverse practical experience in this field. At the ISST he was deeply involved in designing the Continuous Engineering research focal point. His “Habilitation” postdoctoral thesis "Evolution Patterns“ laid the first theoretical foundations for the proposed research approach. Due to his professional work at Talanx insurance company he has first-hand experience of the practical migration problems and solution methods.

Prof. Dr. König knows the practical problems to be solved in this project due to his former work at SAP AG. He has extensive knowledge of modern technologies that are used with manufacturers and users of large information systems.

The research team - Professors Michael Löwe and Harald König as well as research assistants Michael Peters and Christoph Schulz – was supported by resources from the FHDW Hannover from 1st November 2005 until 28th February 2007. The team has developed an initial theoretical design for the refactoring of information systems and published findings in two contributions to international congresses as well as two technical reports. In addition the first research partnerships have been initiated as a result of invitations to lecture at the TU technical university in Berlin, the University of Oldenburg and the University of Leicester.

Christoph Schulz’s proposed doctoral thesis on the subject of "The Refactoring of Information Systems" has been accepted at the TU technical university in Berlin, Faculty of Electrical Engineering.

Within the framework of two student projects and several degree theses since 1st January 2005 a tool has been developed for the analysis, transformation and extension of systems in the programming language Java, which can serve as a basis for the planned tool development.

Objectives

  • Extensive survey of currently existing demands, technologies and tools
  • Further development of the existing theoretical foundations to form a comprehensive concept which fulfils all the practical requirements
  • Formalising the existing research partnerships, if possible in the form of joint projects sponsored by third parties
  • Evidence of practical feasibility by developing a prototypical tool support
  • Initiation of development partnerships for the commercial application of the results.



New cryptographic systems


Overview

Current discussions in society and politics show the increasing significance of the question of data security. The Internet accounts for only one, albeit important, field of application. A classic procedure for encoding data is the asymmetrical RSA algorithm based on the number theory. Compared with symmetrical procedures the algorithm provides relatively few weak points, however it is also very slow. For this reason a new procedure is to be developed in this project based mainly on problems of diophantine, inhomogeneous approximation of real irrational numbers and no longer based, as was formerly the case, on the use of modular arithmetic and the problem of the prime factorisation of larger numbers as the most important security background.

Starting position

The RSA procedure still continues to be the commonest encoding procedure. Its security is based on the difficulty of splitting large numbers into their prime factors in an effective time period. However it is not yet proven that the RSA procedure cannot be attacked by a different method other than finding out faster and faster prime number factorising algorithms, for example with the number field sieve.

The elements of the continued fraction factorisation developments of real irrational numbers also display irregular behaviour; apart from estimating errors in rational approximation no quantitative statements can be made as to the exact values of the individual convergents. This gave rise to the idea of a new encoding algorithm. So far no encoding algorithms with methods of diophantine approximation have been developed.

Own preliminary work

Prof. Dr. Carsten Elsner has been doing research in the field of analytic number theory for nearly 20 years especially in the section of diophantine approximation. He already explored this theme in his degree dissertation "Theory of the approximation of complex numbers with algebraic numbers of limited degree” and also in his “Habilitation” postdoctoral thesis “Congruence conditions in the diophantine approximation of real irrational numbers”. Since then he has published countless publications in international journals on diophantine approximation. Prof. Elsner belongs to an international research network which takes him to regular lectures in Japan (Hirosaki University, Keio University).

FHDW Hannover’s cooperation with Prof. Dr. Volkhard Klinger has already resulted in a further development in new cryptographic algorithms. In this field, research assistant Michael Peters also did preliminary work with implementing some ideas. Furthermore research assistant Martin Schmidt will support the project within the framework of his degree thesis. Here it is planned to establish the theme of this degree thesis in the field of number theory, above all in the examination and evaluation of the new cryptographic algorithms.

Objectives

  • Complete implementation of the new algorithms
  • Documentation of the ideas used
  • Simulation of attacks on the security of the algorithm
  • Statistical tests
  • Comparison of advantages and disadvantages of the encoding technologies most commonly available on the market
  • Patenting the new cryptographic procedures