Output Response
∀ n 3 0
y(n) = can s0 + ∑(m = 0 to n-1) ca n- 1 - m bx(m) + dx(n).
The response can be decomposed into the sum of the zero input response (response if input = 0)
can s0
and the zero state response (response if initial state s0 = 0)
∑(m = 0 to n-1) ca n - 1 - m bx(m) + dx(n)
From this, you can see that the input/output behavior of the system is linear if the initial state is zero. Specifically, if
- input sequence x1 produces output sequence y1, and
- input sequence x2 produces output sequence y2
then for all u, w ∈ Reals,
- input sequence wx1 + ux2 produces output sequence wy1 + uy2
This property is called superposition.