How Much Bandwidth Does America Really Need?

Now here's a contentious question: how much bandwidth does America really need?

If you ask broadband providers, more often than not their answer will essentially amount to whatever speed their technology can deliver.

I don't necessarily blame them for doing this as how can you answer that question with a higher number that instantly marginalizes your technology or existing infrastructure? For example, if a satellite provider were to say 1Gbps is the optimal speed then they just cut themselves out of the game.

But as I've argued before we can't let private interests drive this discussion about what's needed for the public good.

So how much bandwidth does America really need?

Let's start by taking a look at a recent ITIF report entitled, "The Need for Speed: The Importance of Next-Generation Broadband Networks."

In it, ITIF makes a strong case for the need to support the deployment of "next-generation broadband networks."

They define "next-generation broadband" as at least 20Mbps and ideally 50Mbps or upwards downstream and at least 10Mbps or higher upstream.

While I definitely agree that getting us from 5Mbps to 50Mbps will enable the creation of a host of new applications, what's odd is that their own report suggests that 50Mbps is too low a goal to set for next-generation broadband, even in the near-term.

Take this section, for example:

"It is not a stretch to envision a household of the near-future having at least the following demands for high-speed broadband Internet access: Mom engaged in a videoconference for her home-based business; dad watching a live HDTV football game; daughter using the computer to access streaming video of a college course lecture; son playing a real-time interactive game; relatives, perhaps grandparents, in town with one downloading an episode of a high-definition movie and another connected to an uninterruptible medical video feed to a remote monitoring facility. In the background home appliances are being monitored and video home security devices are sending video feeds back to the home security company's emergency operations management center. This home would easily consume more than 90Mbps of aggregate bandwidth (both directions): 15Mbps per HDTV stream x 2 HDTV streams, 80Kbps for gaming, 18Mbps for high-definition two-way video conferencing (requiring 18Mbps both upstream and downstream), 15Mbps for a video course lecture, and 10Mbps for home security and home-based monitoring uses."

So by their own calculations the household of the near-future will require at least 90Mbps of symmetrical bandwidth.

This is especially interesting as it legitimizes the goal of achieving a 100Mbps Nation by 2015 as not being based in idealism but instead reflecting the very realistic demands for bandwidth of what a truly networked world will look like.

But given that broadband equals infrastructure, we can't only be thinking about things in terms of a few years into the future. We need to look ahead at least twenty so that we're not sinking money into investments that will be outdated before their lifespan is done.

Let's visit another section from ITIF's report:

"Within the next decade, 2160P 'QuadHDTV' will come onto the market, driving demand for bandwidth consumption up to 64Mbps for good-quality 2160P. On the technology frontier lies ultra high-definition video (UltraHD), which Japan is currently experimenting with. UltraHD video, operating at 7680x4320 resolution and requiring 256Mbps, will bring cinematic quality video to wall-sized video displays, and will eventually become as common and affordable as the 1080P HDTV sets of today. UltraHD television and video will require substantial amounts of and become a leading consumer of broadband going forward."

So within the next decade we'll be moving from HD video to QuadHD, which will at a minimum increase our demand for bandwidth four-fold as all video applications will have four times as much resolution. So now we've gone from 90Mbps symmetrical to 360Mbps.

Looking a bit further ahead, in the not too distant future we're going to be living in an UltraHD world, which has 16 times the resolution of HD, bringing our demand for bandwidth up from 90Mbps to almost 1.5Gbps.

And remember, ITIF's calculations of future household bandwidth demands admit that they're only taking into account the applications we know about today and that there will undoubtedly be a host of new applications that are created as a result of having this much capacity that we can't imagine today.

So really these numbers should be considered as the baseline goals not the end-game of what we'll need, and it's likely our demands will grow even larger.

Let's review:

- To fully support the technologies we have today to deliver HD video a household needs at least 90Mbps of symmetrical bandwidth.

- To fully support the technologies of ten years from now to deliver QuadHD video a household will need at least 360Mbps.

- To handle the technologies of twenty years from now to deliver UltraHD a household will need at least 1.5Gbps.

With all this in mind it would be irresponsible to set goals for next-generation broadband at anything less than 100Mbps, and really we should leave the 100Mbps Nation behind and start talking about a 1Gbps Nation and beyond.

And given these goals, we should be focusing government subsidies on networks capable of supporting these speeds moving forward.

Getting everyone today's broadband is not good enough. We need to set bold goals that reflect the growing demands of these clear technological trends. Otherwise we're going to set ourselves up to be on the lagging edge of broadband for decades to come.