Wireless Video: Demand vs. Capacity

The forecast for internet video use from Cisco’s annual Visual Network Index continues to astonish. It estimates that, by 2021, 82% of all internet traffic will be video—up from an estimated 73% in 2016. That’s all internet traffic, both wired and wireless. Predictions for wireless video traffic are no less impressive: 78% in 2021 over 60% in 2016.

This got us to thinking about how such massive amounts of video will make it to their intended addresses, especially wirelessly delivered video. The study doesn’t cite underlying data and assumptions made that would reveal how it measures demand vs. capacity.

Today, we have more efficient compression, LTE networks and Wi-Fi technical improvements. Future next-generation 5G networks are on their way. If wireless viewers are now consuming video over Wi-Fi and wireless carrier networks, what happens when more UHD (including future 8K resolution), virtual reality and augmented reality content attempt to reach wireless devices? Add to that the idea that just about every “thing” will be connected to the internet and communicating with other “things” wirelessly, and the question of adequate capacity looms.

Today’s solution is using the aforementioned technical improvements and allocating more radio spectrum to wireless providers and Wi-Fi. One could question the wisdom of this approach given the greater reliance on spectrum to deliver more data-intensive content and IoT traffic.

To date, most of the brain power and money is going to treating the symptom, not the problem. Naturally, there is more than one problem, but one of the most foundational problems is how spectrum is used. I’s mostly divided up into little chunks, and spectrum licensees use their chunks for their purposes within the confines of rules governing how it can be used and how the users keep from encroaching on each other’s territory.

In recent times, the privilege of using some of that spectrum for commercial use has gone to the highest bidder. Government auction of spectrum use is a tough habit to kick. Most of the proceeds from these auctions go directly into a government’s treasury. As long as governments can haul in billions of dollars in spectrum auctions, there’s little incentive to pump government funds into finding a more efficient way to use it.

But what happens if, one day, there isn’t enough of it?

There are commercial and government entities losing some access to spectrum that are working on finding new, more efficient ways of using it. On the commercial side, terrestrial broadcasters have a lot of to lose and they’ve already lost a good chunk of it. Terrestrial broadcasters have been moving off of 600 MHz (ITU Region 2) and 700 MHz (ITU Regions 1, 2 and 3) bands to make more room for wireless operators for several years now. Recent incremental gains in improving transmission efficiency and employing HEVC video compression may not be enough as they struggle to hang on to their current spectrum access.

In Europe, in conjunction with the DVB-T2 digital terrestrial standard, engineers are working on a system that uses wideband signals over a swath of spectrum. The technology, dubbed Wideband reuse-1 (WiB), was presented in a technical paper by Swedish-based Teracom at IBC 2016. To paraphrase (and surely oversimplify), WiB would allow coverage of the entire UHF band from a single wideband transmission per transmission site. It would allow for the re-use of spectrum within a wideband, combined with a number of techniques to cancel interference. Its developers also claim it will significantly reduce network costs and increase capacity.

There are other early-stage efforts to change how the radio spectrum is used, such as the U.S government’s DARPA “Spectrum Collaboration Challenge.” This encourages science teams to “develop radio networks that can autonomously collaborate to thrive without spectrum allocation.” No doubt there will be some time before solutions are realized, but in the meantime, at least some scientists are trying to solve the capacity problem at the core level.