Scott T. Wilkinson, Ph.D., Vice President – Product Management and System Engineering, Hitachi Telecom (USA) Inc.

When speaking about next generation PON (or even current generation PON), the first question always asked by potential customers or cocktail party attendees is, “What would someone do with that amount of bandwidth?” Is there really a need for 100mbps to the home?

Fiber to the home in Asia has allowed carriers there to offer 100mpbs bidirectional service for years, but here in the United States there are still many carrier executives who don’t believe that you or I need anything approaching that speed. Last week at the SCTE Cable-Tec show, an entire panel of cable industry luminaries spent a session convincing an attentive audience that the coming 50mbps over DOCSIS 3.0 deployments were more than their customers would want and certainly more than they could use.

If that’s a true statement, then what are those Asian customers (and a few lucky customers in the United States) doing with their 100mbps of bandwidth?

Video is usually used as the justification for larger bandwidth to the home, and video certainly is a good reason to want more bandwidth. But the main uses for higher bandwidth to the home are the same applications that people use now, just in larger quantities and at larger scales. E-mail attachments, downloading and sending files, online applications, gaming, music, and the other applications that many people with low-speed access use every day are still the primary applications when consumers have access to much higher speeds.

Take exchanging photographs as an example. As higher and higher pixels counts on cameras proliferate without corresponding software expertise required to reduce the size of the resulting pictures, photo attachments on e-mails become increasingly unwieldy. For example, my brother-in-law always purchases the “best” camera available and enjoys sharing his family photos with all of us around the country. His e-mails have grown from a few hundred kbps to multiple mbps over the last few years. Downloading e-mail messages from him on our high-tier 20mbps cable modem can take several minutes, which may not seem like a lot but can be quite frustrating when a few dozen of such e-mails come in after a family vacation. Upstream sharing of photographs is even worse. In my house, we use an online photograph printing service to store copies of our most important photographs. Uploading 50 or so photos can take hours.

Neither of those services – e-mailing or posting photographs – is generally considered a reason for deploying 100mbps to the home. However, a 100mbps connection could potentially reduce the time spent on photography exchange by a factor of five or higher. These applications and other applications that Internet users employ every day are the drivers for higher speed bandwidth, not necessarily any special new bandwidth-heavy applications.

Another example is online backup. Dell Computers recently began offering their online backup service for free to anyone who buys a new Dell. However, primarily due to the limitations of the U.S. broadband infrastructure, they limit the amount of backup storage to a small fraction of the hard drive size. My family recently purchased a Dell, upgraded to the 50gbps storage option and pushed the “back up now” button. Three days later, the backup completed. This is an application that could have greatly improved utility with a FTTH connection, and in fact is one of the booming applications in Asia where people have those faster connections. (Dell also would like to make money on upgrades to larger storage spaces.)

Applications like backup that run during idle time have resulted in a change in the overall data traffic pattern in Japan, as tracked by the Japanese government. Whereas 10 years ago the Internet traffic had a distinct peak around 8 p.m. (video watching time), the peak has now spread out to all hours of the day and night have risen. Idle processes, remote applications like video monitoring, and an increasing number of useful applications built with the expectation of high speed bandwidth to the home have made Internet traffic start to look more like a traditional utility like electricity or water.

Certainly video is an application that can benefit from faster download speeds. The promise of downloading high-quality movies to a home device has only recently been viable in some U.S. locations, despite a demonstrable market for such services (Netflix is called “Netflix” and not “Mailboxflix” for a reason). In Japan, the standard instructions for hooking up your TV/monitor to the 100mbps FTTH connection include information on how to best view streamed videos on your monitor. Near real-time video is considered a standard feature, one that is only available to a select few in the United States.

In general, equipment providers rely too much on the 3D, high-def video boogeyman to sell FTTH, but there is some truth to the scary stories. Unicast video, where everyone watches their own unique content, is going to increase, which reduces any sharing benefits that are currently enjoyed in the majority broadcast world. Downloaded video quality is now expected to be as good as a rented Blu-ray disk (or at least a DVD). Multiple devices in the home can now download video at the same time, doubling or tripling the demand. Customers are less and less willing to put up with long delays or pauses in their video experience. Video is indeed an application that can benefit from higher speeds at the home.

When faced with all of the applications for high-speed bandwidth, it is easy for carriers with an inferior plant to pick them apart and “prove” that any one application can easily be served with their chosen FTT (something other than) H architecture. However, anyone with children knows there is never just one bandwidth-hogging application running in a modern technical household. A single teenager can be simultaneously streaming video, downloading music, IMing, video conferencing, researching homework topics, and gaming. Multiply that demand by 2.5 for the average nuclear family and it is not difficult to quickly approach 100mbps or higher in both directions.

And that’s before adding any new “killer app” to the mix. When software designers are given the green light to assume high-speed bandwidth exists to every home, they can innovate in ways that stodgy network operators or equipment vendors can’t imagine, resulting in increased demand for high-speed bandwidth in a very profitable feedback loop. In fact, that’s happening in Asia where 100mbps is common. But that’s another topic for another day.

Scott T. Wilkinson, PhD, is vice president – product management and system engineering for Hitachi Telecom (USA) Inc., and is in charge of the product management and system engineering department for Hitachi’s North American FTTP product line. Dr. Wilkinson’s department is responsible for the specification of Hitachi’s North American products. Prior to Hitachi Telecom, he was the executive director of system engineering and product management for Parama Networks, a developer of system-on-a-chip solutions for SONET/SDH networks. He was previously director of applications engineering for Kestrel Solutions, a company developing frequency division multiplexing equipment. Dr. Wilkinson started his career with Fujitsu Network Communications in their SONET and DWDM product areas.