Unable to perform assignment because the left and right sides have a different number of elements.

I'm trying to use this LMS algorithm to estimate the impulse response however i'm getting th following error.

"Unable to perform assignment because the left and right sides have a different number of elements.

Error in lms2 (line 10)

y(n) = sum(w.*xN); % Output signal "

The coefficents are as follows:

x is 38500x1 double

N is 200

THIS IS THE CODE BELOW.

function [y,e,J,c,w] = lms2(x,d,N,mu)

M = max(size(x));

y = zeros(1,M);

w = zeros(1,N); % initialise weight vector

for n = N:M

xN = x(n:-1:n-(N-1));

y(n) = {sum(w.*xN)}; % Output signal

e(n) = d(n) - y(n); % Error signal

w = w + mu*e(n)*xN; % filter coeffients update

c(n-N+1,:) = w(1,:); % matrix containing the coefficient vectors

end

J = e.^2; % learning curve

Answer 1

[Deepak Meena]

Hello Victor,

My understanding is that your problem is with the syntax and your understanding of least mean square algorithm is correct.

Now answering your question y(n) and {sum(w.*xN)} are of different data types

When you try this in command window:

>> y = zeros(1,1000);
>> y(12)
 
ans =
 
     0
 

so here y(n) is a scalar integer. But on the other side

 >> w = zeros(1,100);
 >> xN = x(100:-1:1);
 >> ans = w.*xN;

Here w : 1x100

xN: 100x1

Now when you check the dimension of "ans" in workspace you will find it will 100x100 double array .This is because " .*" stands for element wise multiplication .When we multiply a 1x100 array with 100x1 array element wise it produce 100x100 array. To read further about element wise multiplication consider this document.

Similarly, when you do

>> ans  = sum(w.*xN);

here "ans" would be 100x1 matrix as it performs row wise operation. Addition to that when we add curly bracket with it create cell array. So on the LHS we have a scaler and on the RHS we have Cell array, that's why that error is coming up.