Static Analysis of Digital Filters

We present an Abstract Interpretation-based framework for automatically analyzing programs containing digital filters. Our framework allows refining existing analyses so that they can handle given classes of digital filters. We only have to design a class of symbolic properties that describe the invariants throughout filter iterations, and to describe how these properties are transformed by filter iterations. Then, the analysis allows both inference and proofs of the properties about the program variables that are tied to any such filter.

• Type: Conference
• Year: 2004
• Where: ESOP
• Authors: Jérôme Feret
• Keywords: Abstract Interpretation-based Framework, Digital filters, ESOP 2004, Programming Languages, filter Iterations