Hello,
I'm trying to understand why lower bandwidth, in LoRa modulation, would mean lower data rate. In my mind, narrowing the bandwidth would mean "squeezing" the signal frequencies, that were spread all over the bandwidth range, in a narrower space of frequencies. I've done a simple drawing to express what I mean. In my mind option A would make sense. The number of symbols (where symbols means the black dots to me) stays the same. Or what happens is more like B? But why?
Thank you, I have already seen the video. A part of the answer to my question would then be: given a fixed spreading factor, and decreasing the bandwidth, would result in less symbols available. But how does that affect the time the signal takes to send a chirp? Doesn't it just mean I have less characters available to be sent?
I'm not sure that an intuitive explanation is accurate or helpful, but one way of look at the reduction of bandwidth is that the receiver has a more difficult task to discriminate between neighboring frequencies, the closer they are together.
In general with communications, lower bandwidth = lower data rate. It is a property that become evident when taking the Fourier transform of a changing signal.
Buried in the video comments is the Matlab code to simulate transmission and reception, which will run on GNU Octave.
Bandwidth is a direct function of the data rate. The extreme is about a one Hz bandwidth that takes around 15 minutes to send one bit, but can be received halfway around the world on extremely low frequency radio.
By the way, Visual Electric also produced this fun video, which explains the bandwidth issue in regard to the history of the first successful transatlantic cable.
Thank you. I also thought about the intuition of discriminating between frequencies. But, if I'm not wrong, when for example we increase the Spreading factor, keeping constant the bandwidth, we actually fit more frequencies in the same bandwidth span (with the drawback of a longer message, since we have to fit more symbols). And that doesn't seem a problem.
I knew that message duration = 1/f in general. I don't see however how this concept applies to LoRa, which is quite a different modulation, since it is digital.
Thanks already for the time spent helping the noob student
Yes you are right they are still waves. If I'm not getting it wrong, you mean that the fact that it takes more time is based on the fact th frequency is lower? To me it seems more like in this way: given a central frequency, lets say 433 Mhz:
with 125 BW we have a span of frequencies between 433 +- 125/2
with 62.5 BW we have a span of frequencies between 433 +- 62.5
If we say that the frequency gap between each "chip" or "symbol" (the black dots in the figure) stays constant, then of course I understand that there is less "space" for the symbols. So the conclusion would be: fixed spreading frequency means also fixed frequency gaps between the chips. My (humble) impression is that that's not the right way to see it, because when we increase the SF we are able to have smaller frequency gaps in the same bandwidth (if this last concept is correct)
