Autonomic Computing
Competency Leader
Autonomic systems are often typified as consisting of large numbers of distributed autonomic, often resource-constrained embedded, systems. These types of systems could be hoped to evolve dynamically but as Baresi et al. (2006) point out, in these domains open-world assumptions about how a piece of software might be used are dominant. Designers cannot fully predict how a system behaves and how it will interconnect with a continuously changing environment. Therefore open assumptions must be built in and software must adapt and react to change dynamically, even if such change is unanticipated.
Continuously evolving autonomic or adaptive systems may be expected to cycle continuously as a control loop (Dobson et al., 2006)
L. Baresi, E. D. Nitto, and C. Ghezzi. Towards open-world software: Issue and challenges. In Software Engineering Workshop, 2006. SEW ’06. 30th Annual IEEE/NASA, pages 249–252, April 2006.
Mike Hinchey
Printer-friendly version