The Fiber Optic Association’s president explains a unique view of how a standard can be developed.
By Jim Hayes, Fiber Optic Association
Why is the Fiber Optic Association (FOA) creating its own standards for fiber optics? What is the purpose of these standards? And what are standards, anyway?
Here at the FOA, we get calls all the time with technical questions. Most questions concern problems with fiber-optic installations. Some are about how to design or install networks. But the majority are about testing.
Many of the questions can be answered by referring the caller to an industry standard. Most people assume standards exist but do not have copies of them for reference; or if they have a copy, they often do not really understand what it means. Because FOA members have participated in various standards activities for almost 30 years, we are usually able to answer these technical questions and often lead those asking questions to our website, where we have explanations of many of these standards. However, many callers ask why standards are so expensive and hard to get.
Recognizing a need, FOA started discussing creating our own standards, recognizing and interpreting current standards for network contractors, designers, installers and users. Part of that discussion involved defining what standards are, who can write them and how they benefit the industry.
What are standards?
Standards are confusing to many people. Some people think they are mandatory, but in fact they are just guidelines for good practice. In our investigations into standards, we found this definition from the International Organization for Standardization/International Electrotechnical Commission (ISO/IEC), the international “godfathers” of all standards that clearly define that role for standards.
ISO/IEC Guide 2:1996, definition 3.2 defines a standard as: “A document established by consensus and approved by a recognized body that provides for common and repeated use, rules, guidelines or characteristics for activities or their results, aimed at the achievement of the optimum degree of order in a given context.”
The meaning is clear. Any group of users can mutually agree on a practice that makes it easier for them to conduct their activities. So industry associations like the Telecommunications Industry Association (TIA) create standards, as does Telcordia for telephone companies, along with many private companies, the government and many more organizations.
Standards affect every aspect of our lives. The length of an inch or millimeter, a mile or a kilometer, depend on international standards. So do the pound or kilogram, seconds or hours, degrees Celsius or Fahrenheit, the octane rating of gasoline or petrol, and many more measures we use in our daily lives. Even the Internet depends on the standards for Internet Protocol (IP) data communication and Ethernet for the connection of our computer to the Internet. Somewhere, there are groups that set the standards for all these parameters, which facilitate the daily functions of everything we do.
In the communications cabling business, most U.S. users are more familiar with TIA standards or the international equivalents from ISO/IEC. These standards, which cover various types of cabling components and systems, ensure that networks may be designed to work on a standardized cabling architecture with known performance parameters. Likewise, groups like the Institute of Electrical and Electronics Engineers (IEEE) create standards for networks to operate over these standardized cabling architectures. These standards must be interdependent for networks to be built, especially when looking to the future.
Furthermore, these standards groups are continually looking to the future. Their work is to develop equipment and cabling capable of handling the next generation of networks. While users are installing systems today based on current and past networks, standards are being written for the next generation of networks that may not be available for years.
As standards move forward, legacy systems can be left in limbo. For example, the latest, fastest networks are not being specified for use on Om1 or Om2 fibers, which is the majority of the installed base and still represents a large amount of fiber sales. A test standard recently was published that was aimed at testing these future networks, but created controversy over its applicability to legacy systems.
While these standards affect all of us, they are written by manufacturers for manufacturers, expressly to allow the design, manufacture and sale of products that offer interoperability. What about the users of these products? I once asked the current head of the TIA TR-42 Committee why TIA did not write books interpreting these standards for users, the way the National Fire Protection Association (NFPA) writes books interpreting the National Electrical Code. I was told it was the responsibility of the manufacturers to interpret the standards for their customers, not the responsibility of the TIA.
Another group that writes standards affecting cabling is the National Electrical Contractors Association (NECA). NECA standards are aimed at designers and installers of electrical-power and low-voltage systems. Because about 80 percent of all NECA members install low-voltage and fiber-optic cabling, they are very concerned with standards for installing those systems and some of their standards—covering topics like installing conduit—are widely used.
FOA and NECA wrote a standard in the 1990s for fiber-optic installation (ANSI/NECA/FOA-301) that is in its third updated edition already. Recently, FOA and NECA agreed to make ANSI/NECA/FOA-301 available as a free download to those involved in fiber-optic installation. Why give it away? Users need it and it reduces our workload at FOA because it answers many of the routine questions we get on installation.
Lots more people are writing standards. According to ANSI (the American National Standards Institute), that organization has more than 125,000 members and documents more than 100,000 standards. With all those standards, why does the FOA need to create more?
As ANSI and ISO/IEC state, standards are created by consensus so that a group can engage in orderly business. IEEE writes network standards; TIA writes cabling standards—but who writes standards for the fiber-optic technicians who have to install and test the fiber-optic systems covered by the IEEE and TIA standards?
The FOA seems to be a logical choice. We already have FOA standards. We have internationally recognized standards for certifying technicians for work in fiber and premises cabling, standards for curriculum for teaching fiber-optic and premises cabling, and even standards for instructors who teach fiber and premises cabling.
We have the constituency that wants and needs these standards—38,000 certified technicians worldwide who build many of these networks. Plus we have the internal expertise to create them, especially considering our 27 FOA master instructors who average more than 20 years’ experience in various aspects of the industry, but are primarily focused on installing and testing fiber-optic networks.
The goal of the FOA standards work is different. Our standards must be derivative of other standards for networks and cabling so the use of our standards will facilitate the application of the systems those other standards define. But while their standards tend to look only to applications for the future, we must recognize all uses for fiber optics and always remember the past; so our standards are relevant for the legacy systems still in use.
The FOA also had these two goals.
1) Make the standards readily accessible, including free whenever possible. Making them available for free is simple; we just put them on our website where they can be downloaded easily, in a format compatible with PCs, tablets and smartphones.
2) Make the standards understandable. Making standards understandable is more difficult, and became the biggest hurdle in our project. The standards associated with fiber-optic testing have evolved over the last 25 years and have more “tacked-on technology” than you can imagine.
Tribulations of testing
When I co-wrote the first standard for testing fiber-optic cable plants more than two decades ago, it was split into two versions: OFSTP-14 covered multimode fiber and OFSTP-7 covered singlemode fiber. While the methodology was the same, the test conditions were different, so the two were separated. In retrospect that may not have been a good idea. And in all the revisions, changes were made mostly by adding material without rearranging, so the most recent version of OFSTP-14 spends most of the early part of the document discussing new technology; the diagram of the actual test method is relegated to the appendices.
The FOA decided to take a fresh approach. First we asked ourselves, “What should a fiber-optic testing standard cover?” We came up with these six questions that cover practically every type of test.
- What needs to be tested?
- What equipment is needed to perform the tests?
- What are the procedures for making the tests?
- What are the options required to implement the tests under various conditions?
- What are the sources of error in the measurements?
- What documentation is required for the tests?
Of these questions, all are obvious except 4 and 5. On question 4, testing an installed fiber-optic cable plant, there is only one way to conduct the test, but three ways to set a zero-dB reference, depending on the types of connectors on the cable plant and the test equipment. These are the well-known 1-, 2-, or 3-cable reference methods that were written into the original OFSTP-14/7 standards. They are included in the current standard, but as one of our master instructors who works for a major fiber-optic manufacturer pointed out, it does not specify why there are three methods and when each should be used. We decided that information was vitally important to include in our standard.
Question 5 is usually addressed indirectly in a standard. For example, it has been well known since OFSTP-14 was first written that modal conditions in multimode fiber were a major source of measurement variability. This issue was addressed by references to other standards that could be used as test conditions. By the 1990s, a method of controlling mode distribution—wrapping fiber around a small-diameter mandrel—was added to the standard.
Recently a more-sophisticated method of specifying modal conditions—encircled flux (EF)—was included in the latest version of OFSTP-14. But it has become the source of some controversy in the industry. Some argue that EF is confusing because it is not a way to control modal distribution, but only a way to measure it. That is a major change from the previous standard, which gave directions for implementing modal control. No directions are given in the new standard on how to control modal conditions, although another international standard implies the previous method (mandrel wrapping) meets the required conditions specified by encircled flux. Complaints have been voiced by some that EF was included in a standard even though it is not a proven technology, and with round-robin testing as to its capability still underway, it probably will not be proven for another year. Others contend that EF does not apply to networks except those using Om3/Om4 fiber being designed for networks at 10-Gbits/sec or faster, or even to field measurements at all.
Although the new OFSTP-14 focuses on mode power distribution as measured by EF to minimize measurement uncertainty, it does not tighten requirements for other sources of error like source wavelength and fiber size, which might be bigger factors in measurement uncertainty than mode power distribution.
The FOA decided instead to list the sources of measurement error in the standard itself, and put detailed information on controlling measurement error into background material. Thus the users of the standard could decide what was appropriate for the situation. A 10-Gbit/sec network operating over only 2-dB dynamic range is a lot different than an old digital system that could work over 10 dB or more, and should require different test conditions. In addition, the test methods for multimode and singlemode fiber are the same, so our viewpoint allowed covering both in one standard.
Finally, recognizing the preference of users for visual information rather than written descriptions (the old “a picture is worth a thousand words” adage), we decided to use diagrams as much as possible.
The next step was to create the standard for testing the installed cable plant. Then we found that our approach was so much simpler than the conventional approach to standards, we could summarize the standard in one page of diagrams and directions. That surprised even us.
What about consensus? That’s easy for the FOA, with our worldwide network of master instructors who have been helping us with our certification programs, curriculum and website for years. And they were great—helping us edit the standards and fine tune the content in real-time, online discussions. Recommendations from this group led to the development of several more standards in a similar manner. Doing all the collaborative work only by email and teleconference made the process much quicker than if it were done by physical meetings; of course, it was much cheaper for the participants as well.
To increase access to standards, the FOA put the standards on our website and invited users to download them free of charge, as we do with all our technical materials.
And there you have it—a unique view of how a standard gets developed.
- Realization of a need
- Understanding its usage
- Creating the technical content
- Getting input from those who know how it is used to ensure its correctness and applicability
Feedback from our members and the industry has been great. Having these documents gives users a simple, easy-to-use interpretation of the standards created by the network and cabling standards groups. FOA standards can be incorporated in project documents knowing that contractors, designers, installers and users will have a common “language” to define testing of their fiber-optic projects.
Jim Hayes is president of the Fiber Optic Association (www.thefoa.org). His work in standards development began in the early 1980s with NBS (now NIST)creating a transfer standard for basic optical power measurements for fiber optics. It continues with TIA and other standards groups today.