On the Bright Side, the Latest Developments in Solid-State Light Sources Promise to Bring Single-Mode Fiber Much Closer to the User. Whether a User Will Want It or Not Is An Entirely Different Story.
January 1, 2003
Pundits have long crowed about optical’s pending dominance over copper. Even that old Jewish mystical tradition, the ‘Kabbalah,’ maintains that ‘space’ was originally created from light bursting forth from the primordial ‘dot.’ Yet today, we are witnessing the shrinking of space.
And light has its role to play there as well, as our globe is quickly enmeshed in optical fiber.
Each year, Cabling Systems takes a special look at the optical sector — a 50,000-foot market survey of what’s up — a quasi Fiber Report Card. This year brings several surprises. Let’s start with some quick general numbers.
On the vendor side of the market, San Jose, Calif.-based Infonetics Research Inc. reports that worldwide optical network hardware revenue totalled US$2.26 billion during the third quarter of 2002, down 14 per cent from the previous quarter: This comes after several years of double-digit increases.
That kind of performance deserves a failing grade.
On the user side, TeleGeography Inc. reports in its 13th annual survey of international telecommunications trends, that international telephone traffic, the traditional bread and butter of the long-haul optical network industry, increased just 10 per cent in 2001 (the latest available full year figures) after two years of unprecedented growth.
This, says the report, is the slowest rate of growth in 20 years. That’s not overly surprising, perhaps — considering the popularity of alternative communications channels such as e-mail and the World Wide Web, but distressing just the same.
A separate report from the same, Washington, D.C.-based, research group, paints a similar picture for the Internet.
“The growth rate of international Internet bandwidth slowed to just under 40 per cent in 2002, says an October report, after “more than doubling” each year since the introduction of the Web. Total capacity between some major cities actually shrank, the report adds, so don’t be looking for easy marks from optical next semester.
It is from this rather bleak networking climate that three primary trends emerge: Single mode fiber is moving inside; somewhat contradictorily, interest in fiber to the desktop is slowing dramatically; and, not too surprisingly, more work is being done by fewer people.
Let’s take a brief look at each.
On the bright side, the latest developments in solid-state light sources are promising to bring single-mode fiber much closer to the user — if not all the way to the desktop — by dramatically lowering the cost of optical networking. This, some suggest, could spark yet another round of infrastructure upgrades.
Most indoor optical links are across multimode fiber today, while high bandwidth backbones and long-haul networks typically run single mode fiber, a less expensive alternative that provides much longer distances between links and much higher bandwidth.
However, with the recent introduction of Vertical Cavity Surface Emitting (VCSE) lasers — solid-state light sources similar to those used in today’s CD Players — single mode structured cabling solutions are becoming far more accessible.
The multimode fiber used to date will reach only 220 metres; and then only with specialized, mode conditioning, patch cords, explains veteran Montreal telecommunications consultant, Dunn Harvey. “But this is still short of the 300 metres called for in the (TIA) 568B structured cabling specification,” he adds, so a cable upgrade is often called for.
“The newest multimode fibers will reach that far,” he points out, “but single mode may well become the preferred solution, because it provides so much more elbow room. Fifty micron (single mode) was used by telco’s back in the 1970s,” he adds, reflecting on his long career. “Now it’s the ‘new’ fiber for customer premises installs.”
Peter Schweiger, Business Development Manager at Agilent Technologies Inc. in Mississauga, Ont. says multimode fiber is still predominant in-house, but agrees that “forward looking people will now begin installing single-mode fiber to reduce bandwidth and distance limitations while moving forward.”
But it may be a slow take-up, he adds. “It’s not a cost issue any longer. Where it used to be much more expensive for everything from connectors to transmitters, that’s not the case anymore. And we’re finally seeing standards written for single-mode within building,” he adds, so optical is becoming far more accessible.
“I just think most people still see fiber as a backbone technology,” he explains. “Users just don’t yet realise how close fiber is to them already; that the majority of fiber installations today are not between cities, but within cities — either in the WAN or on the edge between the central office and the customer. So I think you will see a switch to single-mode everywhere,” says the test specialist. “It just might take a while.”
But copper, and even most existing fiber, will not work for the new 10 Gigabit Ethernet, Harvey points out. “The bandwidth may be there, but not the distance. So people who have already installed the previous 62.5/125m fiber, can now use it as pull wire for the new fibers” he adds.
And that brings us to the second and most surprising trend of this report card, fiber to the desktop (FTTD) is losing momentum.
DESKTOP FIBER LOSING MARKS
Fiber’s pending domination of in-building systems has been ‘just around the corner’ for some 10 years now. In a world where business empires can crumble in a single calendar cycle, you know that can’t be good. Sure enough, just as advanced optical technologies begin filtering into horizontal networks and after years of FTTD hype, LAN users appear to be turning away from optical, instead giving top marks to copper-based solutions such as 100 Mbps and gigabit Ethernet.
Roberta Fox, president and senior partner of Fox Group Consulting, a technology and telecommunications management consulting firm based in Markham Ont., notes they looked at fiber to the desktop for a recent in-house upgrade and have researched the topic for several clients over the years.
“The need is not there and will not be there for the general business marketplace for some time to come,” says Fox. “Fiber is still more expensive to install and maintain and there’s still a perception that it’s fragile. And at the end of the day,” she adds, “why bother?”
We’re even told of folks designing their patching systems around gigabit Ethernet using copper cables, rather than committing to one of several incompatible optical solutions that promise similar bandwidth at far greater cost.
And though Michael Spencer, Partner at Smith Anderson Consulting in Toronto, points out that new solutions from 3M Corp. and other vendors now offer Layer 2 switching and low cost next-generation fiber products that may well help drive fiber out closer to the user, he and several other experts failed to see a need.
“Copper’s doing it,” says Fox. “And copper technology just keeps going faster and faster. You don’t need to put in fiber to get gigabit Ethernet, for instance; you can do it on copper. Fiber is still more expensive to install and maintain,” she adds “and there’s still a perception that fiber is fragile so, at the end of the day, why bother?”
DOING MORE WITH LESS
The third primary theme in this report is not new, but it does seem somewhat more urgent this semester. It revolves around surviving in an ever-shrinking pond.
Agilent’s Schweiger notes that today’s economic climate means fewer people must now accomplish more, and in less time.
And, he says new tools and techniques are making such productivity increases not only possible, but also commonplace. “There is a need for speed out there,” he says, “and not just from a bandwidth perspective. And fewer people are able to do more today, given the right tools. Until recently it would take one person one day to test 12 fibers,” he says: 12 being the maximum number of fibers an installer would run into in any one cable.
“But today that same person could just as easily have 864 fibers to test in one cable,” he adds. “Using traditional techniques
that would take about three weeks to test, but with today’s new types of test equipment and procedures; it can be done in one to two days.”
Such is life in a highly competitive environment
Many of our experts pointed out that the formal training of field personnel becomes increasingly important in a world where getting it right the first time may mean the difference between turning a profit and taking a loss.
Several other subjects have been identified as requiring greater effort this year.
There are issues with the fiber backbone becoming the bottleneck, as cheap, high-speed copper-based solutions such as gigabit Ethernet take hold.
We’re also told of a growing concern regarding performance and distance over existing optical networks. And also, that as bandwidth needs climb, some operators of older networks are finding their fiber does not live up to original spec.
Such disappointments could mean an ‘F’ grade for the installer when it comes to repeat business.
Even fiber’s marketing potential gets little better than B plus this year. Rather than evolving into simpler solutions, confusion is said to be building with the addition of several specialized options such as 50 micron and application optimized fiber.
Bright spots? If there is one, it may well be in consumer electronics, where fiber-optic leads for household and personal audio equipment — long a rare and expensive option — are appearing on retail shelves everywhere and in great variety.
A BELOW AVERAGE TERM
In a nutshell, fiber continues as the solution of choice in carrier networks. From an installation standpoint, however, an abundance of ‘dark’ (unused) fiber — laid throughout the ‘dot com’ years on the assumption demand would have them sparked up in no time — has become a glut. Where quantum leaps in bandwidth are required, Wavelength Division Multiplexing (WDM) (a major contributor to the bubble’s burst) appears the option of choice — laying multiple ‘colours’ of light on existing fiber rather than building out the network.
So although bandwidth requirements are still growing, far less new fiber is needed to meet that need.
In-building network operators — the one great hope for the optical industry — are having a hay-day with off-the-shelf copper-based networks that provide up to a gigabit at commodity prices. So even if fiber eventually pulls in a passing grade here, it might just as well climb back into that primordial dot for a while and wait for the latest achievements in copper-based technologies to mature.
All things considered, it seems fewer folks will be ‘seeking into the light’ for the foreseeable future, at least from a cabling standpoint. More to the point, those who do will expect a far bigger bang for their buck.
Chris Blythe is a broadly published writer and infocom analyst. He can be reached at firstname.lastname@example.org.