Data Flow Oriented Test Selection for Lotos

A data flow oriented test selection method for selectively generating abstract tests from a LOTOS specification is presented. This method is based on static data flow analysis to identify maximal IO-df-chains, which exhibit the associations between each specification output and those specification inputs that influence the output through definitions and uses of variables. The maximal IO-df-chains are then used to guide the derivation of test sequences. A new set of inference rules is introduced for this purpose. As an illustration, the test selection method is applied to the alternating bit protocol.