Casting uint16 to int16 in Simulink

Hi Mathworks Community

do you know a good implementation for casting an uint16 to an int16.

I'm using Simulink to program a dSpace MicroAutobox II and receiving two 8bit values via CAN-Interface representing a signed integer in Two's complement representation.

Example 0100 1001 0010 1010 = 18730; 1010 1000 0010 1010 = -22486

Is there a native method to do so? Yet I'm using a self-written MATLAB Code looking like this:

function n_actual = fcn(data1, data2)
% data2 contains bits 8 - 15, data1 0 - 7 little endian
%#codegen
tmp = uint16(data2)*2^8+uint16(data1);
if(tmp > 2^15)   % negativ value
                 % tip signed bit
                 % (2^16-1) subtracted by tmp for int16 
                 % (tmp+1)*(-1) add 1 for the previous
                 % step to stay in range of value and negate
    tmp = -((32767-int16(tmp-2^15))+1);
else
    tmp = int16(tmp);
end

n_actual = tmp;

The typecast methode simply takes the number and matches it on the range of values to cast to. -20 would then be 32767, therefore not interesting for my use, am I right.

Hopefully my English is not to bad to understand my issue.

Thanks for your reply

Matthias

[Matthias]

I forgot to mention why it's important to handle it as efficient as possible. The Autobox is a trip computer (real time) with few resources and therefore every avoided computation is a good computation.