Line of Sight
Communications Latency: How Low Can You Go?

Communications Latency: How Low Can You Go?

Speed, speed, speed...it's the mantra in modern communications, and most any kind of latency has become intolerable.

Latency has become a critical factor in many communications applications these days. Latency, of course, is the short delay that occurs due to the time it takes a signal to travel from one point to another, either in free space or some medium. In the past, latency did not matter that much. After all, most electronic signals travel at the speed of light or a bit less, and for most of us that’s fast enough. We can never go faster than that anyway. Let’s face it, there’s no such thing as zero latency. But less would be welcome.

Some of the newer communications applications require or would like less latency. Factory control systems often need short and deterministic timing to be successful. The new 5G wireless systems have been created to keep latency to a minimum. The target has been to get below 10 ms, down to about 1 ms. Self-driving and other connected cars with vehicle-to-vehicle (V2V) communications must have low latency to function safely and effectively. A more recent demanding application involves financial trading operations.

Electronic trading is mostly automated, so buying and selling happens super-fast. The faster, the better, as you can exploit your competition and take advantage of arbitrage opportunities. To make the most profit, communication delays must be very short. A few milliseconds make all the difference between success and failure.

Banks and other financial institutions use fiber-optic links for electronic trading. While fast, the optical networks also have latency. The speed of light slows down when passing through glass fiber. The index of refraction and IR signal wavelength determine the amount of delay. A general rule of thumb is a delay of 5 µs/km or about 8 µs/mile for single-mode fiber. Of course, other factors like electrical-to-optical conversion time and vice versa, SERDES, and packet format and length will have an impact on the overall transmission time.

Right now, a typical link from New York to London has a one-way latency in the 30- to 40-ms range. But that’s not fast enough for many. It’s led to research into alternative solutions. Then radio waves were rediscovered. Radio waves travel faster in free space than in a fiber cable.

Some organizations have adopted the unexpected use of shortwave (SW) radio in the 3- to 30-MHz part of the spectrum. Signal propagation is by sky wave, which is refracted from the ionosphere back to earth and in some cases reflected back up for another bounce. The result is that a small signal can easily travel around the globe using multiple skips. Latency must be less than 30 ms to be useful. It’s not known what frequencies are being used.

Just remember that at these low frequencies, bandwidth is severely limited to only a few kilohertz max. And propagation is variable depending on how the ionosphere changes during different times of the day. Still, it’s working. Will we be seeing giant Yagi’s on bank rooftops? A new use for the HF spectrum other than ham radio?

We are living in an age where the speed of light is not fast enough. But what can we do about it?

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish