I’m pretty forgiving about reading mildly inaccurate physics stuff, especially when it’s in science fiction stories, but every once in a while I read a real zinger that nearly causes me physical pain. Recently I was reading an article about Wi-Fi when I came across this paragraph:
This was in an otherwise fascinating article about how engineers are developing light-based Wi-Fi, called Li-Fi, which would in essence just be a lamp in your room that would communicate to your wireless device through a light sensor. Fluctuations in the light, undetectable by the human eye, would carry the information.
You can see why I was in pain from the first sentence: “Since light travels much faster than Wi-Fi radio waves, data speeds are significantly faster.” The problem with this is that, in air, the speed of visible light and the speed of radio waves are basically identical and equal to the vacuum speed of light.
I point this out not to pick on the author of the article — writers are very busy, and sometimes a sentence gets poorly worded or some concept non-essential to the story gets honestly misunderstood, and we’ve all made mistakes like that at some point. (You’ll notice I don’t even link to the article.)
But I thought it was a good excuse to ask and answer the question: what does determine the data speeds of our communications devices, and why is light a better option than radio? And if it is a better option, why haven’t we done it already? So let’s take a look!
Let’s start again with the basics. Both light and radio are electromagnetic waves, which are combined oscillations of electric and magnetic fields, and their only internal difference is their frequency of oscillation. Visible light, the light we can see with our eyes, has frequencies that range from 400 terahertz (400 million million oscillations per second) for red and 790 terahertz (790 million million oscillations per second) for violet. The visible color of light directly depends on its frequency or combination of frequencies. Radio waves oscillate much slower: anything that oscillates at 300 gigahertz (300 billion oscillations per second) or slower is classified as a radio wave.
As an aside, I said that frequency is the only “internal difference” between light and radio. Obviously, light and radio waves appear very different from one another, but that difference arises due to how their respective frequencies interact with matter. Visible light is strongly emitted/absorbed by individual atoms, and therefore strongly interacts with it, whereas radio waves do not strongly interact with individual atoms and tend to pass through most matter with ease. This latter property is what makes radio waves so useful: you can listen to the radio or talk on your cellphone or use your Wi-Fi anywhere in your house, (mostly) regardless of any walls between your receiver and the source.
So how do we encode information onto a wave to transmit it? We begin by choosing a wave frequency to transmit over, which is usually called the carrier frequency. In an idealized form, the carrier frequency wave is a perfect sinusoidal wave:
The wavelength is the length of one full cycle of the carrier wave, and it is inversely proportional to frequency. Often the wavelength and frequency will both be given for a carrier wave, even though they are proportional; each of them gives a different intuitive feel for how the wave is oscillating.
An idealized perfect sinusoidal wave has no beginning or end: it keeps oscillating in exactly the same way forever, from the start of time to the end of time. To transmit information on a carrier wave, we distort it slightly, those distortions carrying the information we want. For traditional radio, an actual audio signal is encoded on the carrier; for internet signals, the distortions carry the binary 1’s and 0’s of the digitized information.
Let’s talk about traditional broadcast radio for a little while first. You have probably heard of the two ways of encoding information in radio: AM (amplitude modulation) and FM (frequency modulation). In amplitude modulation, the signal is used to create a change in the amplitude of the carrier wave, and in frequency modulation, the signal is used to create a change in the frequency of the carrier wave; this is illustrated below.
So what is the effect of such modulation on the frequency of the total transmitted wave? Our signal almost certainly consists of a range of frequencies; for example, if we’re transmitting an audio signal, the audible range of human hearing is between 20 hertz (Hz) and 20 kilohertz (20,000 hertz, or 20 kHz). We would expect, then, that the transmitted wave will now have a frequency spectrum broadened to about 20 kHz. If we were to make a plot of the power transmitted in our wave as a function of frequency, it might look something like this.
That is: since we have now modulated our carrier wave, it has a range of frequency components centered on the carrier frequency. The width of this range is called the bandwidth, referred to the width of the “band” of frequencies that our signal consists of. If we transmit a 20 kHz audio signal, we expect a bandwidth of around 20 kHz.
(This picture is a little oversimplified, because the bandwidth also is affected by other factors, including the “natural” bandwidth of the carrier wave. No carrier wave is a perfect sine wave, and it is affected by random fluctuations that cause its spectral width to increase.)
Now suppose we can to send multiple independent signals at the same time. The only way we have to distinguish between different signals is by using different carrier frequencies, which is why AM and FM radio stations are labeled by a number that represents that carrier frequency.
But in order to avoid interference between signals, we must choose carrier frequencies that are far enough apart in frequency so that their bands don’t overlap. Or, in picture form:
If frequency bands overlap, when you listen to one channel you’ll hear a little bit of the other, in what is known as crosstalk. The only other way to distinguish radio signals is by distance, as a radio transmission can typically only be recovered over a finite distance from the transmitter. If you’ve ever been on a road trip, and were listening to one radio station as it got fainter and fainter and started to hear another one on top of it, you’ve experienced this: you were between two distant radio stations which had similar carrier frequencies.
Certain radio frequency bands have been set aside for certain purposes. The AM radio band consists of signals between 530 and 1700 kHz. In North America, this band is broken up into 10 kHz wide channels, allowing 118 channels overall. The bandwidth used for each channel is usually less than 10 kHz and therefore can only transmit over a fraction of audible human frequencies. This is fine for talk radio but not so good for music, which is preferably transmitted in the FM band.
AM radio falls into what is called “medium frequency” radio; then there is “high frequency” radio which are frequencies used for shortwave radio communication, between 3 and 30 megahertz (millions of Hertz, or MHz). FM radio falls into the “very high frequency” radio designation, and this band is defined as frequencies from 30 to 300 megahertz. FM radio is allotted the frequencies between 87.5 to 108 MHz.
Now we start to see what a difference a carrier frequency can make! The allotted bandwidth for FM radio is 20 million Hertz, which is significantly larger than the allotted bandwidth for AM radio, which is about 1 million Hertz. This means that not only is it possible to fit more audio signals within that bandwidth at the same time, but more bandwidth is in principle available for each channel! The bandwidth of an FM station is often close to 300 kHz, which allows both mono and stereo audio signals to be broadcast simultaneously as well as other digital information (like sending the song information to be displayed on the screen of your car stereo, if it is capable).
So a higher carrier frequency basically gives us more room to fit more channels with a higher rate of data transfer. When we move away from audio signals and instead think of transmission of, for example, internet data, similar rules apply: if we want to transfer data at 300k bits per second, we would need a 300 kHz bandwidth available.
You might think that simply going to a higher bandwidth would be better, but there are practical limits to this. Very high frequency radio signals are impacted more by barriers than medium frequency signals, making them impractical for some communications situations. Visible light, we have noted, has a much higher frequency, but doesn’t pass through barriers at all, which means at the very least it could not replace free-space radio communications. You could imagine going to even higher ultraviolet or X-ray frequencies, but at that point the electromagnetic waves become harmful to humans, so they are only used in limited applications and not for communications.
Now what about computer Wi-Fi? There are a number of radio frequency bands used for Wi-Fi, so let’s just talk about the 2.4 GHz (2.4 gigahertz, or 2.4 billion Hertz) range. There are 14 channels used in this range, starting at 2.4 GHz and spaced 5 MHz apart, though each channel usually broadcasts over about 25 MHz. In the early days of Wi-Fi, it was rather common to run into problems where you and a neighbor would broadcast over similar channels, causing a drop in performance; higher-frequency Wi-Fi systems have reduced this problem. Wi-Fi broadcasts in the “ultra-high frequency” (300 MHz to 3 GHz) or “super-high frequency” (3 GHz to 30 GHz) radio bands. These signals can pass through walls, but are attenuated more in doing so, making them great for Wi-Fi in your home that usually can only run into trouble from your nearest neighbors. With little competition from neighbors, and a high carrier frequency, it can transmit data at higher bandwidths: the aforementioned 25 MHz, fast enough to stream movies and download data or play videogames with your friends online.
(I should note that, once we’re talking about Wi-Fi, the means of encoding information are much more sophisticated than AM or FM, but that’s beyond our discussion here.)
But even this may not be enough for modern internet users, as our demand for high data rates increases. This brings us back to the beginning of this post, where engineers are proposing to use visible light to transmit wireless data. Visible light encompasses a 390 THz range of frequencies, a roughly 1000 times larger than the entire radio frequency spectrum combined, at 300 GHz! This means that, if we stuck to somewhere around 14 distinct frequency channels for visible light, we could have data transmission rates more than 1000 times greater than what we have with Wi-Fi. The Li-Fi company quotes 2.25 GHz, which is roughly 1000 times greater than the 25 MHz we quoted as the current Wi-Fi rate.
The idea, as I understand it, would be to have an actual lamp that would broadcast data at the same time it illuminates the room. The fluctuations in the light signal would be too small for the human eye to detect. There are two possible limitations with this approach: since light doesn’t pass through walls like Wi-Fi, your computer would have to be in line of sight with the transmitter to receive data. Furthermore, you would have to be in the same room with the transmitter or have multiple transmitters if you wanted to use your computer everywhere in your home. And this doesn’t take into account one other limitation: even if your Li-Fi transmitter can transmit data really fast, your internet service provider may not be able to transmit data to the Li-Fi that fast! Your speed will still be limited by your ISP, and they tend to be very stingy about data rates.
But, to summarize: the rate at which data can be transmitted via electromagnetic waves is strongly dependent on the carrier frequency being used and the available bandwidth. Light may, or may not, become a solution for transmitting more data in an electromagnetic spectrum that is already crowded with applications, but it seems worth exploring!
