Design methodologies for FPGA using Matlab and Simulink.

EasyChair Preprint 590

Abstract

The first method for designing digital systems was the edition in schematic form; since then, the methods of design have advanced enormously, especially in recent years. In this paper a brief description of the design methods will be made, until reaching the advanced methods supported by Matlab and Simulink. The latter are fast and flexible, it allows to explore different architectures, and to check different data formats and operators at different points of the system. Each FPGA (Field Programmable Gate Array) supplier has its own tool to design and verify using Simulink and Matlab; usually in fixed-point arithmetic, although lately floating-point libraries have been included. It is also possible to design using tools from Matlab and Simulink, from models in fixed point or floating point arithmetic. In any case, from the fixed point model, the system can be automatically generated in a hardware description language, which can be implemented on an FPGA. The techniques over Simulink and Matlab are based on models edition and verification techniques. The verification techniques are "co-simulation" and "FPGA in the loop". These methods are being used in industry and research because they are fast and flexible; these tools does not seem to have a special follow-up in the academic world, which may leave these techniques of the university environment unattached.

Keyphrases: FPGA, MATLAB, Metodología de diseño, Simulink, VHDL, Verilog, punto fijo, punto flotante

@booklet{EasyChair:590,
  author    = {Santiago Tomás Pérez Suárez},
  title     = {Design methodologies for FPGA using Matlab and Simulink.},
  howpublished = {EasyChair Preprint 590},
  year      = {EasyChair, 2018}}