Autonomic Systems, Coping Strategies adn Dream Functions

The main challenge for autonomic systems operating in remote environments is how to deal with uncertainties. This paper develops a framework of coping strategies for autonomic systems to deal with uncertain situations, new environments, conflicting sensory inputs and diverse mission objectives. Important aspects of the coping strategies include learning mechanisms, for which the concept of dream functions is presented. More sophisticated autonomic systems are likely to involve collaboration and interaction with other units (autonomic and human). For successful mission goals it is suggested that a fairness and collaboration approach to resource allocation will provide better results than self-interest strategies. Meta-games and Hyper-games provide a mechanism for encapsulating fairness parameters. The paper argues ethics has to be considered explicitly when developing more sophisticated autonomic systems.