Scalability Analysis of the RADAR Decision Support Tool

This report presents a theoretical complexity analysis and empirical scalability analysis of the Requirements and Architecture Decision Analyser (RADAR) described in a separate paper [1]. We will refer to that paper as the RADAR paper and assume the reader is familiar with its content. Further information can be found in the tool website The analysis of a RADAR model involves four main steps: 1) Generating the design space 2) Simulating all solutions in the design space 3) Shortlisting the Pareto-optimal solutions 4) Computing expected values of information over the shortlisted solutions The time and space complexities of these steps are summarised in Table I. The first step, generating the design space, involves a single recursive traversal of the model's abstract syntax tree (AST). The time and space complexity of this step is thus O(m) where m is the model length measured as number of nodes in the model's AST.