MINA - a tool for MSC-based performance analysis and simulation of distributed systems
Performance analysis can help to address quantitative system analysis from the early stages of the system development life cycle, e.g., to compare design alternatives or to identify system bottlenecks. This thesis addresses the problem of performance evaluation of distributed systems by employing a viewpoint where analytical and simulative evaluation techniques are unified in the MINA tool to make use of both techniques. We suggest a modelling tool chain to evaluate the performance of distributed systems like computer and communication systems based on an MSC description of the system. MSC-based performance evaluation of distributed systems is an approach that uses performance models, which are based on an MSC description of a system to evaluate system performance measures. To determine the system performance, these descriptions can be extended by notions for time consumption and resource usage and afterwards be included in a system performance model. Based on this unique model specification, analytical as well as simulative techniques can be applied to achieve either quick mean value results by queueing networks analysis or confidence intervals or transient measures by simulation. The applicability to real world systems and the advantages of the tool has been demonstrated by a large application example in the field of mobile communication systems, and its effectiveness has been evaluated by comparing it with other approaches. The experimental results show that the tool is scalable, the way it can model simple as well as complex systems. Moreover, it is straightforward and has the ability to find reasonable solutions in an efficient manner.