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.