A self-adapting algorithm for context aware systems

This paper presents a self-adapting algorithm that can automatically detect the changes in a system execution context and decide how the system should react. The self-adapting algorithm is characterized by a closed feedback loop with four phases: monitoring, analyzing, planning and execution. The monitoring phase uses the RAP (Resources, Actions, Policies) context model to represent in a programmatic manner the raw data collected about the system´s self and execution environment. In the analysis phase, the context entropy concept is used to evaluate the context situation for detecting the context changes and determining the degree of respecting a predefined set of policies. The planning phase uses a reinforcement learning based technique to explore all possible system´s states and select the adaptation action that should be executed by the system as a response to the context changes. The execution phase modifies the system behavior by enforcing the adaptation actions selected in the planning phase.