# kaiser

Kaiser-Bessel window.

## Syntax

w = kaiser(n,beta)

w = kaiser(n,beta,type)

## Inputs

n
Length of the kaiser window.
Type: integer
Dimension: scalar
beta
Mainlobe width control factor. See Comments.
Type: double
Dimension: scalar
type
Type of the window.
Type: string ('symmetric' (default) or 'periodic')

## Outputs

w
Window coefficients.

## Examples

Symmetric kaiser example:

kaiser(6,3*pi,'symmetric')
w = [Matrix] 6 x 1
0.00061
0.17041
0.83535
0.83535
0.17041
0.00061
Periodic kaiser example:
kaiser(5,3*pi,'periodic')
w = [Matrix] 5 x 1
0.00061
0.17041
0.83535
0.83535
0.17041

The window type depends on the application. If the window is used to preprocess data for an FFT then the type should be 'periodic.' If the window is used to design an FIR filter then the type should be 'symmetric.' The difference is that the periodic window omits the right endpoint so that consecutively positioned windows are not duplicating the right endpoint of the 1st window at the left endpoint of the 2nd window. The two endpoints of the symmetric window have equal values.

If beta = pi*alpha then 2*sqrt(1+alpha^2) is the width of the mainlobe of the Fourier transform. A typical beta value is 3*pi.