Issues in communication during architecture design in modern software engineering : A Systematic Literature Review

Abstract: The purpose of this research is to define which instruments with their characteristics exist to mitigate the communication issues in modern software development as well as the exact issues that emerge during software architecture design with their characteristics and validity. A Systematic Literature Review (SLR) is used as a major method of the research. Also, this thesis presents the vision of an ideal tool as well as an attempt to approbate the findings from the SLR in a small surrogate project of the ‘Battleship’ game while applying usual non-functional requirements from highly-collaborative industry projects. Additionally, we make the emphasis on simplicity and ease of use of the tools as well as the importance of the popularisation and promotion of both found and future tools to facilitate verification and adoption in the industry. The most popular Agile approaches (Scrum, Kanban, Extreme Programming, DevOps) do not specify, how these aspects of planning and documentation have to be managed and there are no existing additional widely-used and accepted tools for this purpose. In this work, a significant amount of attention is paid to the verification of the found tools in practice with feedback. For the tools which are not verified in practice, this work does a comparison against the vision of an ideal tool to understand. This helps to understand whether the given tool can be quickly adopted in practice, if further research is continued, or whether certain changes are required. Unfortunately, this work cannot present any existing tools which conform to our vision and are popularised, heavily verified in practice or relatively widely used in the industry. The closest solutions found were introduced in two industry papers concentrating on repurposing the existing Agile practices to mitigate the planning and documentation issues. During an SLR, a few tools were found, which we believe can be relatively easily taken to verification in practice or promoted to be used in industry on an everyday basis. However, we admit that further research of the found tools, which do not satisfy our requirements, can be used to reevaluate them in future to check for the desired properties again. Furthermore, we define a roadmap for future research for both the search for a tool and the development of a new instrument.