An Actual Communication Network Performance Evaluation Method

The queuing system and network theories may be hardly applied to solve performance evaluation problems for actual large-scale computer systems and networks. This is true especially for high performance communication networks such as wireless, satellite or photonic networks. Firstly, the reason is that the theoretical assumptions can hardly be fulfilled in actual networks. Secondly, the methods are too complex and hardly understandable for actual network designers. Thirdly, the well-known methods do not teach too much possible designers of their solutions, in contrast to e.g. simulation modelling. Fourthly, case studies of performance evaluation solutions for large-scale communication networks can hardly be found in the literature. The paper presents an approximate method developed early in the computer network era by a team of the first Polish network designers and is a proposal for an approach for solving significant actual large-scale network design problems.