02.10.2019 Views

UploadFile_6417

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

Optimal Equiripple Design Technique 363<br />

7.5.2 CONSTRAINT ON THE NUMBER OF EXTREMA<br />

Before we give the solution to this above problem, we will first discuss<br />

the issue: how many local maxima and minima exist in the error function<br />

E(ω) for a given M-point filter? This information is used by the<br />

Parks-McClellan algorithm to obtain the polynomial interpolation. The<br />

answer is in the expression P (ω). From (7.41) P (ω) isatrigonometric<br />

function in ω. Using trigonometric identities of the form<br />

cos (2ω) =2cos 2 (ω) − 1<br />

cos (3ω) =4cos 3 (ω) − 3 cos (ω)<br />

. = . .<br />

P (ω) can be converted to a trigonometric polynomial in cos (ω), which<br />

we can write (7.41) as<br />

P (ω) =<br />

L∑<br />

β(n) cos n ω (7.46)<br />

□ EXAMPLE 7.21 Let h(n) = 1 [1, 2, 3, 4, 3, 2, 1] . Then M =7and h(n) issymmetric, which<br />

15<br />

means that we have a Type-1 linear-phase filter. Hence L =(M − 1)/2 =3.<br />

Now from (7.7)<br />

n=0<br />

α(n) =a(n) =2h(3 − n), 1 ≤ n ≤ 2; and α(0) = a(0) = h(3)<br />

or α(n) = 1 [4, 6, 4, 2]. Hence<br />

15<br />

or β(n) =<br />

P (ω) =<br />

3∑<br />

α(n) cos ωn = 1 (4 + 6 cos ω +4cos 2ω +2cos 3ω)<br />

15<br />

0<br />

= 1 15<br />

{ 4+6cosω + 4(2 cos 2 ω − 1) + 2(4 cos 3 ω − 3 cos ω) }<br />

=0+0+ 8 15 cos2 ω + 8 15 cos3 ω =<br />

[<br />

0, 0, 8<br />

]<br />

.<br />

3∑<br />

β(n) cos n ω<br />

15 , 8 15<br />

From (7.46) we note that P (ω) isanLth-order polynomial in cos(ω). Since<br />

cos(ω) isamonotone function in the open interval 0

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!