Im getting errors at fixed point conversion step, when im converting my matlab code into verilog code

im getting erros at fixed point conversion step when im converting my matlab code into verilog code. The error was at function replacements step in the fixed point conversion step. it asking me for function replacement for covariance matrix, eigen values and eigen vectors. how can i overcome that error. my matlab code has no errors. can anyone help me with this error. here im attaching the image of function replacement error for your reference.

THANK YOU

Answer 1

[Samhitha]

Hi @Kavya T,

The error is because certain functions in your MATLAB code (like cov, eig, exp, sind) do not have native fixed-point support in HDL code generation. The HDL Coder Fixed-Point Conversion step requires you to provide a replacement or alternative implementation for these functions, since they can't be directly translated into hardware (Verilog)

1. Use lookup tables for sind, HDL Coder supports lookup table replacement for functions like sind. You can define a lookup table in MATLAB:

angle = 0:1:360; % example LUT angles 
sine_table = sind(angle); 

2. If eig and cov are computed on static data or data that changes slowly, precompute them in MATLAB and pass the results as parameters/constants to your HDL design.

For more details, look into following documentation:

https://www.mathworks.com/help/hdlcoder/fixed-point-conversion-basics-and-type-specification.html

Hope this helps!