The Last Mile: Streaming’s Bottleneck

Streaming via dialup can be done, but what use is blurry, jerky, thumbnail-sized video? Broadband is a step up, but still doesn’t provide the bandwidth needed to deliver high-quality, full-screen streaming video. For that, you need a fat pipeline, and, thanks to the cable-laying fever of the Dot Com Boom, there’s plenty of fiber optic backbone to support large-scale "ultrabroadband," say fiber optic advocates. Those advocates have a "chicken in every pot" vision of universal ultrabroadband implementation in the U.S., and the beginnings of a roadmap to getting there. But even those fiber optics proponents most closely allied with the technology itself agree that traversing the "last mile"—getting from the backbone into users’ homes—has proven to be a challenge without a quick and easy solution. "The backbone was terribly overbuilt," says Fiber Optic Association (FOA) president Jim Hayes. "Ninety-three percent of all the fiber that’s been installed is still unused."

The term "ultrabroadband" doesn’t have a definite meaning—except perhaps among the fiber optic set—nor is there a standard agreement as to how much bandwidth the average home user needs to experience what Internet streaming can do. In a 2003 BusinessWeek interview, tech entrepreneur Mark Cuban was quoted as saying, "If there's a disrupting technology I could envision 20 years out into the future it would be ultrabroadband to the home. Once we hit 500Mbps, 1Gbps, then we'll see our lives change." Others say such speeds aren’t necessary, suggesting that to stream standard-resolution video to a television on a shared household Internet connection would require 20Mbps, while hi-def could need 50Mbps or more.

Successful business models that take advantage of broadband’s capabilities have only begun to take hold in the last year or two, which is why much of the technological innovation in solving the last-mile bottleneck of Internet connectivity has been focused on methods that compete with DSL or cable. These include ADSL, BPL, satellite, and WiMAX (see sidebar, "Broad but not Ultrabroad" ). Another technology, free space optics (FSO), offers a wireless laser-based system for data transfer and may supplement fiber optic delivery, but it’s still expensive and has limitations of its own (see sidebar, "Free Space Optics" ).

To achieve even the lower ultrabroadband speeds (up to 50Mbps), the only currently viable technological solution is laying fiber optic cable. Fiber-to-the-home advocates envision a fiber optic cable going directly into homes and offices. But even that’s easier said than done due to a potent political cocktail of land rights issues, cable ownership, territorial cable and phone companies, and just generally too much "lawyering," as Hayes puts it. Some of the carriers actually laying the cable hope instead to just get fiber close to the buildings and then allow VDSL to provide the 20 to 50Mbps into the home. VDSL, known as very high bit-rate DSL, enables 52Mbps downstream and 16Mbps upstream over copper cables, but its range is limited to only about 4,000 feet. Fiber-to-the-curb proponents envision fiber optic cables running near, but not to, homes and offices, with VDSL bridging the last hundred yards rather than the last mile.

The future viability of full-screen, high-quality streaming media hinges upon the widespread availability of sufficient bandwidth, as well as the business models necessary to make it all profitable.