Claims and Supporting Evidence for Self-Adaptive Systems – A Literature Review

A self-adaptive system aims at anticipating changes which occur in a complex environment and automatically deal with them at run time. Although a lot of work has been done on self-adaptive systems over the past decade but researchers have lack of knowledge on how the research results have contributed in improving of complex software systems. In addition, no systematic study has been performed so far on the claims along with the evidences associated with self-adaption. The concrete objectives of this thesis are to: (1) assess the quality of current research in self-adaptive systems (2) identify the focus of the research (3) what the expected claims are of self-adaption and to what extent these claimed benefits exist and (4) identify potential areas for future work in the field. For this purpose, various categories of 73 papers pertaining to SEAMS (2008-2010) were studied to obtain data related to 20 different items and after the concerted efforts, some research questions were framed for the collection of data. The study description criterion were modified on the basis of results obtained out of the collected data from time to time which helped to modify the performance cycle through a continuous monitoring, resulting in to produce the accurate results at the end. This approach also enables the researcher how critically the papers can be studied/analyzed and how systematically the required data has been extracted from the papers through research questions and data items. The extracted data was then subsequently used to achieve all above mentioned objectives. Numerous limitations during the study like observation of changed results due to human intervention or in case of conflict of opinions were faced. Secondly, this existing approach only dealt with SEAMS community and the results obtained from the extracted data may be different in case of dealing with other software engineering communities like ICAC (International Conference on Autonomic Computing) and SASO (Self-Adaptive and Self-Organizing), which would be taken as a challenge for future work.