Show that c_1,c_2 Exist for Dirichlet Kernel Integral

  • Context: Graduate 
  • Thread starter Thread starter Zaare
  • Start date Start date
  • Tags Tags
    Kernel
Click For Summary
SUMMARY

The discussion focuses on proving the existence of positive constants \( c_1 \) and \( c_2 \) such that \( c_1 \log n \le \int_{-\pi}^{\pi} |D_n(t)| dt \le c_2 \log n \) for the Dirichlet kernel \( D_n \). The user has derived an upper bound \( \int_{-\pi}^{\pi} |D_n(t)| dt \le \frac{1}{\pi} \left( \frac{1}{2} + 4n \right) \) but seeks a more precise approximation for \( |\sum_{N=1}^{n} \cos(Nt)| \). The natural logarithm is specified, and the challenge lies in improving the approximation by considering negative terms in the Taylor expansions of \( \cos(Nt) \).

PREREQUISITES
  • Understanding of Dirichlet kernels in Fourier analysis
  • Knowledge of Taylor series expansions for trigonometric functions
  • Familiarity with integral calculus and properties of logarithms
  • Experience with inequalities and bounding techniques in mathematical proofs
NEXT STEPS
  • Research techniques for improving approximations in Fourier series, particularly for \( \cos(Nt) \)
  • Study the properties and applications of Dirichlet kernels in signal processing
  • Explore advanced integral inequalities and their proofs in mathematical analysis
  • Learn about the convergence properties of Fourier series and their implications on integrals
USEFUL FOR

Mathematicians, students in advanced calculus or analysis courses, and researchers working on Fourier analysis or approximation theory will benefit from this discussion.

Zaare
Messages
54
Reaction score
0
First the problem:
If [tex]D_n[/tex] is the Dirichlet kernel, I need to show that there exist positive constants [tex]c_1[/tex] and [tex]c_2[/tex] such that
[tex]c_1 \log n \le \int\limits_{ - \pi }^\pi {\left| {D_n \left( t \right)} \right|dt} \le c_2 \log n[/tex]
for [tex]n=2,3,4,...[/tex].

The only thing I have been able to do is this:
[tex] \left| {D_n \left( t \right)} \right| = \frac{1}{\pi }\left( {\frac{1}{2} + \left| {\sum\limits_{N = 1}^n {\cos \left( {Nt} \right)} } \right|} \right) \le \frac{1}{\pi }\left( {\frac{1}{2} + n} \right)[/tex]
Which is not good enough.
Any suggestions would be appreciated.

Edit:
By "log" I mean the natural logarithm.
 
Last edited:
Physics news on Phys.org
The integrals over ##\cos(Nt)|## are all constantly ##4##, hence ##\int_{-\pi}^{\pi}|D_n(t)|\leq \dfrac{1}{\pi}\left(\dfrac{1}{2}+4n\right)## by the triangle inequality. So the real problem is to find a qualitatively better approximation for ##|\sum_{N=1}^{n}\cos(Nt)|##, i.e. using the negative terms in the Taylor expansions of ##\cos(Nt)##.
 

Similar threads

  • · Replies 1 ·
Replies
1
Views
3K
Replies
6
Views
2K
  • · Replies 16 ·
Replies
16
Views
5K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 15 ·
Replies
15
Views
3K
  • · Replies 5 ·
Replies
5
Views
4K
  • · Replies 3 ·
Replies
3
Views
4K
  • · Replies 5 ·
Replies
5
Views
3K
  • · Replies 1 ·
Replies
1
Views
2K
  • · Replies 3 ·
Replies
3
Views
2K