Simulink Fixed Point

Simulating Fixed-Point Designs

Once a model has been configured with fixed-point data types, you can simulate its performance to explore implementation tradeoffs between limited range and precision for fixed-point embedded systems. Simulink Fixed Point provides debugging capabilities that let you switch to a floating-point representation of your model to perform benchmarking of the design, identify modeling errors, and determine ideal minimum, maximum, and overflow values.

Using automatic scaling features, you can access these values and change the scaling of signals, optimizing precision while ensuring that their range is covered. Data type ranges can be specified using design minimums and maximums or by using simulation results. Output data types can lock their output scaling to prevent them from being autoscaled. Otherwise, scaling is optimized automatically.

Sharing Models with Fixed-Point Data Types

Simulink Fixed Point makes it easy to share models that include fixed-point data types. Because data types can be globally overridden, you can switch between fixed-point and floating-point representations without changing the model’s structure. The model remains compatible with Simulink, Stateflow, DSP System Toolbox, Communications System Toolbox™, Computer Vision System Toolbox™, and other products supported by Simulink Fixed Point. As a result, you can use one model throughout your design process. You can keep models synchronized, reuse the same models for multiple purposes, and circle back in the development process without breaking the model.

Top right: The Fixed-Point Advisor, showing an initial fixed-point design converted from a floating-point model. Bottom left: The Fixed-Point Tool, which contains a more optimized scaling using global data type override, logged min/max and overflow values, automatic scaling, and comparison of floating-point to fixed-point signal responses.