Code generation concepts in Paragon for generating multiple HDLs

This work presents the code generation concepts implemented in a mixed-signal/technology behavioral modeling environment called Paragon, which was developed at the University of Arkansas. This thesis is mainly focused on the design, analysis and algorithms used in the code generator modules of VHDL-AMS, MAST and VTB. The model information is saved in a centralized database, which makes use of XML and MathML formats as an intermediate representation. Two different approaches are explained for generation of behavioral models in both VHDL-AMS and MAST with illustrative examples. The code generator is also capable of supporting MAST imported models, which effectively translates MAST models into various HDLs. Generation of Resistive Companion models for the VTB environment is implemented using the User-Defined Device (UDD) tool developed at the University of South Carolina.