panji
New Member
Posts: 3
|
Post by panji on Dec 26, 2004 22:20:19 GMT -5
Hi everybody, I have tried to verify that the DFT basis function are orthogonal. If X = (exp(-j*2*pi*n*k/N|k=0:N-1) and Y=(exp(-j*2*pi*m*k/N|k=0:N-1), Hence, by "Orthogonal" it is mean that Sigma{k=0:N-1}[(exp(-j*2*pi*n/N) (exp(j*2*pi*n/N)] = Sigma{k=0:N-1}[(exp(-j*2*pi*(n-m)/N)] This will be 1 if n=m and 0 if not (n not= m). But I have failed to verify that Sigma{k=0:N-1}[(exp(-j*2*pi*n/N) (exp(j*2*pi*n/N)] = Sigma{k=0:N-1}[(exp(-j*2*pi*(n-m)/N)]=0, for n not=m Is there anyway to see basis function of DFT is orthogonal especially for n not= m?
|
|
|
Post by Baraa Alyousef on Feb 10, 2005 14:42:03 GMT -5
HI Mr. as U know the result of 2*x (whatever x was) is even, then if we apply that on the situation of n not= m we will have an even number in the exp(j*even*pi) we know that exp(x) = cos(x) + j sin(x) if x=even*pi then sin () = 0 and cos = 1 ..... simple subistitution in DFT formula U will get the answer of Ur question ....
sincerely. Bara'a Alyousef
|
|