0

I'm taking a signals processing class and am trying to wrap my head around why the following is true.

$$\int_{t_0}^{t_0 +T}cos(\frac{2\pi}{T}kt)cos(\frac{2\pi}{T}mt) = \begin{cases} \frac{T}{2} & \text{if $k = m$} \\ 0 & \text{if $k \neq m$} \end{cases}$$

What does it mean(physically) for two sinusoids when they are multiplied with each other? Is there an easy way of visualizing that? What is the significance of $\frac{T}{2}$ when $k = m$?

Why is the above equation true physically/intuitively for signals?

FafaDog
  • 1,193
  • 1
  • 12
  • 25

1 Answers1

0

OP Here -

Thanks to Semiclassical's answer I've decided to answer my own questions.

(Also for more insight on the trig identities check out this)

So we have $$\int_{t_0}^{t_0 +T}\cos(\frac{2\pi}{T}kt)\cos(\frac{2\pi}{T}mt)$$

We will use the following trigonometric formula $$\cos(A)\cos(B) = \frac{1}{2}\left(cos(A+B) + cos(A - B)\right)$$

Applying that identity $$\int_{t_0}^{t_0 +T}\cos(\frac{2\pi}{T}kt)\cos(\frac{2\pi}{T}mt)dt = \frac{1}{2}\int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(k+m)) + \cos(\frac{2\pi t}{T}(k-m))\right)dt$$ $$ =\frac{1}{2} \left( \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(k+m))\right)dt + \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(k-m))\right)dt \right)$$

Case 1: $k = m$

$$ = \frac{1}{2} \left( \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(2k))\right)dt + \int_{t_0}^{t_0 +T}\cos(0)dt \right)$$

Notice that the first term contributes $0$ because we are integrating cosine over an entire period and the "positive areas" and "negative areas" negate each other, while the second term is integrating $1$ over the period. Thus, $$ = \frac{1}{2} \left( \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(2k))\right)dt + \int_{t_0}^{t_0 +T}\cos(0)dt \right) \\ = \frac{1}{2} \int_{t_0}^{t_0 +T}dt = \frac{T}{2}$$

Case 2: $k \neq m$

From $$=\frac{1}{2} \left( \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(k+m))\right)dt + \int_{t_0}^{t_0 +T}\left(\cos(\frac{2\pi t}{T}(k-m))\right)dt \right)$$

Both terms contribute $0$ since they are integrating a cosine wave over a full period with "negative" and "positive" areas

And of course, a similar argument can be made for sine waves.

FafaDog
  • 1,193
  • 1
  • 12
  • 25