Convert data to sfixed in workspace

Hey.

please, how can i convert the datatype of a variable in workspace from DOUBLE to Sfixed9_EN5.

the variables Lc(i) will be used as inputs of a simulink model.

Thank you.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
global Lc1;
global Lc2;
global Lc3;
global Lc4;
global Lc5;
global Lc6;
global Lc7;
global Lc8;
global Lc9;
global Lc10;
R=1/2;  
Mo=2;
numerrmin =1000;     
SNRdB=1:0.5:3;
BER_bit=zeros(size(SNRdB));
codewordH=[0,0,0,0,0,0,0,0,0,0];
modulatedsig=1-2*codewordH;
for k=1:length(SNRdB)
    fprintf('SNR:%d\n',SNRdB(k))
    frame=0;
    numerr=0;
    rat_err=0;
    EbN0_lin=10^(SNRdB(k)/10);
    EsN0_lin=EbN0_lin*R*log2(Mo);
        Es= 1 ; 
    N0=Es/EsN0_lin;
    sigma=sqrt(N0/2); 
        while(numerr < numerrmin)
            fprintf('frame : %d\n',frame)
        frame = frame + 1;
        bruit= randn(size(modulatedsig));
        receivedsig =modulatedsig+sigma.*bruit;
        Lci = (4.*(receivedsig./N0));
        Lc1=Lci(1);
        Lc2= Lci(2);
        Lc3=Lci(3);
        Lc4=Lci(4);
        Lc5=Lci(5);
        Lc6=Lci(6);
        Lc7=Lci(7);
        Lc8=Lci(8);
        Lc9=Lci(9);
        Lc10=Lci(10);

%         Lc1=typecast(Lc1, 'fixdt(1,9,5)');
%         Lc2=typecast(Lc2, 'sfix9');
%         Lc3=typecast(Lc3, 'sfix9');
%         Lc4=typecast(Lc4, 'sfix9');
%         Lc5=typecast(Lc5, 'sfix9'); 
%         Lc6=typecast(Lc6, 'sfix9');
%         Lc7=typecast(Lc7, 'sfix9');
%         Lc8=typecast(Lc8, 'sfix9');
%         Lc9=typecast(Lc9, 'sfix9');
%         Lc10=typecast(Lc10, 'sfix9');
        Lc1=timeseries(Lc1);
        Lc2=timeseries(Lc2);
        Lc3=timeseries(Lc3);
        Lc4=timeseries(Lc4);
        Lc5=timeseries(Lc5);
        Lc6=timeseries(Lc6);
        Lc7=timeseries(Lc7);
        Lc8=timeseries(Lc8);
        Lc9=timeseries(Lc9);
        Lc10=timeseries(Lc10);
        %sim('LDPCMatlabSimulink','SrcWorkspace','current')
        s=sim('LDPCMatlabSimulink');
        x1=getdatasamples(x1,1);
        x2=getdatasamples(x2,1);
        x3=getdatasamples(x3,1);
        x4=getdatasamples(x4,1);
        x5=getdatasamples(x5,1);
        x6=getdatasamples(x6,1);
        x7=getdatasamples(x7,1);
        x8=getdatasamples(x8,1);
        x9=getdatasamples(x9,1);
        x10=getdatasamples(x10,1);
        SMS=[x1 x2 x3 x4 x5 x6 x7 x8 x9 x10];%%%%%[0 0 0 0 0 0 0 0 0 0];%%
        %iteration=it;
        
           [Num1,Rat1] = biterr(codewordH,SMS);
           numerr=numerr+Num1;
           rat_err=rat_err+Rat1;
        if  frame>2
         break ;
       end
         BER_bit(k)=rat_err/frame;
        end
end
semilogy(SNRdB, BER_bit,'v-r','LineWidth',2);
grid on;
xlabel('SNR,(dB)')
ylabel('(BER)')

Answer 1

[Akash]

Hi Taamouch,

You can convert a variable 'x' of datatype 'double' to the fixed-point data type 'sfix9_En5' by using the 'fi' function as follows:-

x_fix = fi(x, 1, 9, 5);

In this function:

• 1 specifies that the number is signed.
• 9 is the word length.
• 5 is the number of fractional bits.

For more detailed information, you can refer to the MATLAB documentation on the 'fi' function:-https://www.mathworks.com/help/fixedpoint/ref/embedded.fi.html#d126e96527:~:text=example-,a%20%3D%20fi(v%2Cs%2Cw%2Cf),-returns%20a%20fi

You can also utilize the following MATLAB documentation that demonstrates a model for converting signals between built-in and fixed-point data types:-

https://www.mathworks.com/help/fixedpoint/ug/double-to-fixed-point-conversion.html