Cabling for the future: which way?

One of my functions as a networking consultant is to direct investment in communications infrastructure. The communications industry is one of the most exciting, and I feel privileged to be engaged directly in the development of emerging technologies such as Gigabit Ethernet and next-generation structured cabling. Unfortunately, it is difficult to forecast applications beyond three years and therefore to predict infrastructure life. One thing we can be sure of is the relentless growth in bandwid

Apr 1st, 1999

Alan Flatman, lan Technologies

One of my functions as a networking consultant is to direct investment in communications infrastructure. The communications industry is one of the most exciting, and I feel privileged to be engaged directly in the development of emerging technologies such as Gigabit Ethernet and next-generation structured cabling. Unfortunately, it is difficult to forecast applications beyond three years and therefore to predict infrastructure life. One thing we can be sure of is the relentless growth in bandwidth.

So how do we cable our buildings for maximum flexibility and life, while minimizing cost and risk? A strategy of multimode fiber in the backbone and Category 5 in the horizontal was established in 1992, and we soon witnessed its global rollout, despite the absence of standards, applications, installation practices, and accurate field testers; these all followed some three years later, by which time at least 100 million Category 5 outlets had been installed worldwide. Clearly, this was a risk that suppliers and end-users jointly accepted and managed.

New applications continue to exploit the Category 5 installed base but have since consumed every ounce of its available performance. We never imagined that one day we would be running gigabit-speed local area networks over our Category 5 cables but very soon we will be doing just that. 1000Base-T Gigabit Ethernet will use a million-transistor digital-signal-processor to squeeze gigabit-per-second data into a 100-megahertz Category 5 channel: impressive technology, indeed, but not cheap. It was necessary to recharacterize Category 5 installations to support 1000Base-T, and we have since created an enhanced cabling specification based on this exercise. Category 5 had a planned life expectancy of 15 years and has become exhausted after only six years. Category 5 is therefore strategically dead, and it is time to move on.

Cabling for the immediate future now becomes more clear, as Enhanced Category 5 provides breathing space, but for how long? It is therefore prudent to consider the longer-term options: specifically Category 6, Category 7, and optical fiber. Industry specifications for Category 6 and 7 are still being developed, while optical fiber is here today.

Optical fiber has long been considered the ultimate option. It offers high bandwidth and immunity to EMI but requires more costly electronics and does not support conventional telephones.

Proposed by Germany, Category 7 represents state-of-the-art shielded twisted-pair technology with an end-to-end bandwidth of 600 MHz. This phenomenal increase in bandwidth is accomplished by individually screened pairs; however, the level of attenuation at higher frequencies is beyond the capability of transmission electronics and cannot be used for the foreseeable future. A new 8-pin connector is required for Category 7, and no fewer than eight candidates have entered this race. In addition, field testing Category 7 is a considerable challenge and will not help this already expensive option. Judged to be more costly and less attractive than optical fiber, Category 7 is expected to remain a uniquely German phenomenon.

Category 6 represents state-of-the-art unshielded or overall-shielded twisted-pair cabling and offers twice the bandwidth of Category 5. Category 6 still uses the RJ-45 type connector and therefore provides backward-compatibility, although multivendor operation and performance with lower-category plugs and sockets still requires validation. Unlike Category 7, the additional performance offered by Category 6 is fully exploitable using current transmission electronics and easily verified using today`s field-tester technology. The incremental cost of Category 6 will make it an attractive investment despite its current immature state of standardization.

So how does the end-user evaluate the benefits of next-generation copper cabling? In the absence of any planned applications, this is difficult. But let`s remind ourselves that Category 5 cabling emerged in the absence of applications. We generally lacked cabled infrastructure at that stage, so we installed the best available. Category 5 penetration levels are now running high and in a slowing market. So, while the additional bandwidth is currently unjustified, what you do get from higher-performance cabling right now is extra margin which, in turn, translates to improved equipment tolerance, enhanced EMI immunity, and increased forgiveness to cabling installation and lifetime degradation effects.

What you get is insurance, and this is something with which we all have working experience.

Alan Flatman is principal consultant at LAN Technologies (Congloeton, Cheshire, UK). He can be reached at a_flatman@compuserve.com.

More in Home
Sponsored
AV over Twisted-Pair