Security plays an important role in the development of multiagent systems. However, a careful analysis of software development processes shows that the definition of security requirements is, usually, considered after the design of the system. This is, mainly, due to the fact that agent oriented software engineering methodologies have not integrated security concerns throughout their developing stages. The integration of security concerns during the whole range of the development stages could help towards the development of more secure multiagent systems. In this paper we introduce extensions to the Tropos methodology to enable it to model security concerns throughout the whole development process. A description of the new concepts is given along with an explanation of how these concepts are integrated to the current stages of Tropos. An example from the health care sector is used to illustrate the above.

Abstract. Security is a crucial issue for information systems. Traditionally, security is considered after the definition of the system. However, this approach often leads to problems, which translate into security vulnerabilities. From the viewpoint of the traditional security paradigm, it should be possible to eliminate such problems through better integration of security and systems engineering. This paper argues for the need to develop a methodology that considers security as an integral part of the whole system development process. The paper contributes to the current state of the art by proposing an approach that considers security concerns as an integral part of the entire system development process and by relating this approach with existing work. The different stages of the approach are described with the aid of a case study; a health and social care information system.

No abstract

SUMMARYTime-optimal solutions for a simple crane model have been arrived at by a number of authors using expensive analogue methods. The method described in this paper also gives time-optimal solutions but these can be obtained using a very simple microprocessor. These solutions are obtained quickly and inexpensively and thus can be used for comparison with other suboptimal solutions based on more complex and less idealized models. An extension of the method is also outlined which covers certain variants of the simple model.