Discrete-time systems review
Đã giới thiệu| English | Ý nghĩa |
|---|---|
| impulse response | impulse response |
| convolution | phép tích chập |
| causality | tính nhân quả |
| BIBO stability | BIBO stability |
| frequency response | đáp ứng tần số |
One list decides everything
- An LTI system is fully described by its impulse response $h[n]$: the output is convolution $y[n] = \sum_k h[k]\,x[n-k]$. Causality: $h[n] = 0$ for $n < 0$; BIBO stability: $\sum |h[n]|$ converges.

$h = [1,2,1]$ on a unit step gives $y = 1, 3, 4, 4, \ldots$ — it settles at $\sum h = 4$, the DC gain $H(1)$ ✓.
{exercise:uni-ee1201-settle}
The filter h = [1, 2, 1] is fed a unit step. Where does the output settle?
Độ khó 1/5 · 100 xu + phần thưởng nhỏ khi hoàn thành lần đầu tiên hôm nay
A step through any FIR settles at the tap sum Σh = 1+2+1 = 4, which is the DC gain H(1).
The recursive cousin
- The averager $y[n] = (y[n-1] + x[n])/2$ has $h[n] = 0.5^{n+1}$: its step response is $1 - 2^{-n}$ (0.5, 0.75, 0.875, …), because its one pole sits at $z = 0.5$.
- Exponentials are eigenfunctions: $H(e^{j\omega}) = \sum h[n]e^{-j\omega n}$ is the frequency response — the DTFT of the taps.
{exercise:uni-ee1201-pole} {exercise:uni-ee1201-bibo}
The averager y[n] = (y[n−1] + x[n])/2 has a pole at z = 0.5, so its step response is —
Độ khó 1/5 · 100 xu + phần thưởng nhỏ khi hoàn thành lần đầu tiên hôm nay
Each step halves the remaining distance to 1: 0.5, 0.75, 0.875, 0.9375 … = 1 − 2⁻ⁿ.
A discrete LTI system is BIBO stable exactly when —
Độ khó 1/5 · 100 xu + phần thưởng nhỏ khi hoàn thành lần đầu tiên hôm nay
BIBO stability is Σ|h[n]| < ∞; for rational systems this is exactly all poles inside the unit circle.