How is data conveyed over waves such as WiFi?

1.46K views
0

How can a wavelength possibly carry enough information to load say, a video? How can one wave carry instructions for hundreds of thousands of pixels?

In: Technology

>one wave

One wave *frequency* travelling at the speed of light. By changing either the frequency or the amplitude of waves, you can convey ideas like “on” and “off”. For example: you can program an antenna to understand that a wave peak above X volts is a 1, and below X volts is a 0, over a given period of time (typically miniscule fractions of a second). Because light is fast [source required], you can communicate a LOT of 1s and 0s over a short period of time. Those 1s and 0s are translated and compiled as code which executes as whatever data the application is trying to run, like a movie.

Edited to fix mobile autocorrect

Data is just bits, 1’s and 0’s. A WiFi signal is just radio waves (with some extra quirks) and has a theoretical max of 2.4GHz or even 5GHz frequency. These numbers indicate how many 1’s or 0’s per second can be transmitted and/or received. So, in one second 2.400.000.000 or 5.000.000.000 bits can be received. A video of 10 seconds, with a bitrate of 8MB/s, needs only 64.000.000 (8x8x1.000.000) bits to be received for playback.

This sort of radio transmission generally works in a way where you basically take the base radio wave, then add the data on top. Another station then recieves the transmission, subtracts the base wave, and is left with the data.

Modern WiFi uses 2.4GHz or 5GHz as its carrier (base frequency) meaning that it can send a 1 or a 0 from 250 million all the way up to 3.3 billion times per second. A video is just a frame of images, and an image is just a table of pixels. One pixel, normally encoded as RGBA (Red, Green, Blue, Alpha/Transparency) requires 64 ones and zeros (16 for each of R, G, B, and A, hence 16-bit color depth) meaning that you can send of 4 million pixels per second at 250 Mbps (million ones or zeroes per second). A 1920×1080 display only has 2 million pixels, so the 250Mbps WiFi connection can transfer the video at 2 frames per second.

2 frames per second isn’t that much, so thats why we use compression. H265, the High Efficiency Video Codec (HEVC), along with other compression algorithms, use many tricks to reduce the amount of data that needs to be sent. One trick they use is to basically say “This pixel and the next 30 after it are this color” which allows you to not have to send the color of the next 30 pixels, saving data. Another trick is to say “This square of pixels on the screen all moved one pixel to the left this frame, compared to the last” which allows you to not have to resend that entire square of pixels. Using tricks like this, H265 can hit an average of 62% reduction in data size for 1080p video.

We can give the wave a little nudge when we send it out, and the thing on the other side can tell that the wave was nudged. We do really, really complicated nudges for fast stuff like WiFi, but it’s still just a bunch of nudges.

Edit: Source, decades of schooling and experience as an electrical engineer. Still all just nudges tho.

A laymen’s example in my opinion of all these is that wifi is like you and me communicating silently by flashlight.

If you agree to read the flashlight every second and I wanna send you a message, I could keep it on to give you a 1 and keep it off to tell you 0, and you could record that binary and get a message.

If I was a computer, I could flicker this light billions of times per second. If you were a computer, you could read this light billions of times per second. Also, since we’re computers now, we can use wavelengths that humans can’t see.

Boom! you and me can communicate a lot very quickly by using light that people can’t see