Title :
Control theory-based foundations of self-controlling software
Author :
Kokar, Mhtzyslaw M. ; Baclawski, K. ; Eracar, Yonet A.
Author_Institution :
Northeastern Univ., Boston, MA, USA
Abstract :
The authors´ control theory based paradigm gives a framework for specifying and designing software that controls itself as it operates. Based on this paradigm, their self controlling software model supports three levels of control: feedback, adaptation, and reconfiguration. We believe this model can lead to software systems with an impressive capability for responding, adapting, and reconfiguring. Of course, self-controllability does not come for free. The application´s functionality must be supplemented with some redundancy to implement the mechanisms of self-adaptability: evaluation, model estimation, adaptation, and reconfiguration. However, we can reduce this overhead and improve overall system performance by: evaluating the behavior based on a sample of feedback iterations rather than on every iteration; generating more efficient interfaces between components at runtime; and constructing more efficient component organizations, scheduling algorithms, and evaluation algorithms at runtime
Keywords :
adaptive systems; control engineering computing; feedback; redundancy; self-adjusting systems; adaptation; component organizations; control theory based foundations; control theory based paradigm; evaluation algorithms; feedback; feedback iterations; model estimation; reconfiguration; redundancy; scheduling algorithms; self controlling software; self-adaptability; software model; software systems; system performance; Control systems; Feedback; Graphical user interfaces; Hardware; Object oriented modeling; Open loop systems; Operating systems; Software algorithms; Software design; Software systems;
Journal_Title :
Intelligent Systems and their Applications, IEEE
DOI :
10.1109/5254.769883