# Calculating impulse response in convolution.

1. Aug 27, 2009

### ace_terabyte

1. The problem statement, all variables and given/known data
y(t) = $$\int ^{t}_{-\infty} e^{-(t-\tau)} x(\tau-2)d\tau$$
Find h(t), given that the continous time linear time invariant signal has input x(t) and output y(t).

2. The attempt at a solution
I was wondering if I can simply manipulate the expression until i get it in the form y(t) = x(t) * h(t) but I think that that approach is simply wrong. I just have no idea on how to approach this problem and was wondering if you can guide me in the right direction.

2. Aug 28, 2009

### CEL

Convolution is represented by a product only in the frequency domain: Y(s)=X(s)H(s).
In the time domain convolution is defined by the integral
y(t) = $$\int ^{t}_{-\infty} x(\tau) h(t-\tau)d\tau$$

3. Aug 28, 2009

### ace_terabyte

Thanks for that, the question asks me to find out h(t) and does not give me any other information. I assume it is the same in both frequency and time domain.

4. Aug 28, 2009

### CEL

NO, in the time domain you have h(t), in the frequency domain you have H(s), the Laplace transform of h(t).
Compare your equation with mine and you can obtain h(t).

5. Aug 29, 2009

### ace_terabyte

y(t) = $$\int ^{t}_{-\infty} e^{-(t-\tau)} x(\tau-2)d\tau$$

So going by what you said:

y(t) = $$\int ^{t}_{-\infty} (\frac{1}{e})^{(t-\tau-2)} x(\tau)d\tau$$

= $$\int ^{t}_{-\infty} (\frac{1}{e})^{(t-\tau)} (\frac{1}{e})^{(2)}x(\tau)d\tau$$

=$$(\frac{1}{e})^{(2)} \int ^{t}_{-\infty} (\frac{1}{e})^{(t-\tau)} x(\tau)d\tau$$

so is h(t) = $$(\frac{1}{e})^{(t-\tau)}$$ ?

6. Aug 29, 2009

### CEL

No, $$h(t-\tau) =(\frac{1}{e})^{(t-\tau)+2}$$

7. Aug 29, 2009

### ace_terabyte

Thanks a lot CEL, I appreciate it.