Skip to main content

Derive DTFT; DTFT periodic with 2π

Short answer

The DTFT pair is derived by periodic extension, DTFS, and limit N→∞: X(e^{jω})=Σ x[n]e^{-jωn}, x[n]=(1/2π)∫_{2π} X(e^{jω})e^{jωn}dω. Periodicity: X(e^{j(ω+2π)})=Σ x[n]e^{-j(ω+2π)n}=Σ x[n]e^{-jωn}=X(e^{jω}) because e^{-j2πn}=1 for integer n.

Derive Fourier Transform of discrete time aperiodic signal. Also, prove that DTFT is periodic with 2π.

Derivation of the DTFT

  1. Periodic extension: let x[n]x[n] be aperiodic and of finite length (−N1≤n≤N2-N_1 \le n \le N_2). Form x~[n]\tilde{x}[n] by repeating x[n]x[n] with a period NN large enough to avoid overlap. Then x~[n]=x[n]\tilde{x}[n] = x[n] over one period, and x~[n]→x[n]\tilde{x}[n] \to x[n] as N→∞N \to \infty.
  2. DTFS of the extension: with ω0=2π/N\omega_0 = 2\pi/N,
ak=1N∑n=⟨N⟩x~[n]e−jkω0n=1N∑n=−∞∞x[n]e−jkω0na_k = \frac{1}{N}\sum_{n=\langle N\rangle}\tilde{x}[n]e^{-jk\omega_0 n} = \frac{1}{N}\sum_{n=-\infty}^{\infty}x[n]e^{-jk\omega_0 n}

The second form holds because x[n]=0x[n] = 0 outside the chosen period. 3. Envelope: define X(ejω)=∑n=−∞∞x[n]e−jωnX(e^{j\omega}) = \sum_{n=-\infty}^{\infty}x[n]e^{-j\omega n}. Then ak=1NX(ejkω0)a_k = \frac{1}{N}X(e^{jk\omega_0}). 4. Synthesis: using 1/N=ω0/2π1/N = \omega_0/2\pi,

x~[n]=∑k=⟨N⟩akejkω0n=12π∑k=⟨N⟩X(ejkω0)ejkω0nω0\tilde{x}[n] = \sum_{k=\langle N\rangle}a_k e^{jk\omega_0 n} = \frac{1}{2\pi}\sum_{k=\langle N\rangle}X(e^{jk\omega_0})e^{jk\omega_0 n}\omega_0
  1. Limit N→∞N \to \infty: ω0→dω\omega_0 \to d\omega, kω0→ωk\omega_0 \to \omega, and the sum over NN terms covers Nω0=2πN\omega_0 = 2\pi, so it becomes an integral over 2π2\pi. Also x~[n]→x[n]\tilde{x}[n] \to x[n].

This gives

X(ejω)=∑n=−∞∞x[n]e−jωn,x[n]=12π∫2πX(ejω)ejωndωX(e^{j\omega}) = \sum_{n=-\infty}^{\infty}x[n]e^{-j\omega n}, \qquad x[n] = \frac{1}{2\pi}\int_{2\pi}X(e^{j\omega})e^{j\omega n}d\omega

Proof that the DTFT is periodic with period 2π

X(ej(ω+2π))=∑n=−∞∞x[n]e−j(ω+2π)n=∑n=−∞∞x[n]e−jωn e−j2πn\begin{aligned} X(e^{j(\omega+2\pi)}) &= \sum_{n=-\infty}^{\infty}x[n]e^{-j(\omega+2\pi)n} = \sum_{n=-\infty}^{\infty}x[n]e^{-j\omega n}\,e^{-j2\pi n} \end{aligned}

Since nn is an integer, e−j2πn=cos⁡2πn−jsin⁡2πn=1e^{-j2\pi n} = \cos 2\pi n - j\sin 2\pi n = 1. Therefore

X(ej(ω+2π))=∑n=−∞∞x[n]e−jωn=X(ejω)X(e^{j(\omega+2\pi)}) = \sum_{n=-\infty}^{\infty}x[n]e^{-j\omega n} = X(e^{j\omega})

More generally, X(ej(ω+2πr))=X(ejω)X(e^{j(\omega+2\pi r)}) = X(e^{j\omega}) for any integer rr.

Reason: discrete-time exponentials whose frequencies differ by 2π2\pi are identical sequences. Only 0≤ω<2π0 \le \omega < 2\pi (or −π≤ω<π-\pi \le \omega < \pi) needs to be considered. Low frequencies lie near 0,±2π,…0, \pm2\pi, \dots and the highest frequency is ω=±π\omega = \pm\pi.

✓ Answer checked by Claude Opus 5.5 (high effort)

All Signals and Systems chapter 3 questions →