Abstract: The methods for solving the Feature Interaction problems are of two weak points: low detecting ability and incomplete detecting methods. This paper discusses the approaches to detecting feature interactions in intelligent networks at the stage of service design. And based on the discussion, the model for denoting features is presented, called the combining model, which uses the FSM to denote the behavior of features and uses the temporal logic to represent the inhere property of features. Consequently, this paper presents the feature compounding theory and the feature interaction model at the stage of service design.

Key words: intelligent networks, feature compounding, feature interaction model