Normes Ethernet à paire unique : le câblage structuré est-il mort ?

What triggered this recollection was the advent of single-pair Ethernet standards, and discussions we had in a recent TIA standards meetings about potential use cases. Specifically, we discussed how we'll accommodate use cases within the single-pair Ethernet standards. Before we get to the potential use cases, however, it would be helpful to know what has been underway at IEEE in regard to single-pair Ethernet standards.

There has been ongoing work at IEEE in regard to Single Pair Ethernet. Most of the work was initiated with the intention of being used for M2M (machine to machine) and the industrial space. When initially exposed to this work, some thought it would not apply to our commercial spaces (at best), and worried that it would be the death knell of structured cabling (at worst).

Here are some Single Pair Ethernet standards you should be following:

• 802.3bp 1000BASE-T1 (published), which is an Ethernet standard that provides 1 Gbps Ethernet over at least 15 m and up to at least 40 m on one pair of balanced cabling
• 802.3bw 100BASE-T1 (published), which is an Ethernet standard that provides 100 Mbps Ethernet up to 15 m on one pair of balanced cabling
• 802.3cg 10BASE-T1 (in progress), which is an Ethernet standard that provides 10 Mbps Ethernet up to 1 km on one pair of balanced cabling
• 802.3bu (in progress), which is an Ethernet standard that covers power over data lines on one pair of balanced cabling
  
  

After studying the potential impacts and use cases of single-pair Ethernet, there is now broad acceptance and support of its potential in a wide range of facilities and applications, including traditional commercial spaces and, yes, even data centers.

There are a number of projects happening across all traditional standards-development organizations that serve our industry; given our audience and my personal experience, however, I’ll focus on three main TIA documents:

1. ANSI/TIA-568.5, Single Balanced Twisted-Pair Telecommunications Cabling and Components Standard, a new standalone standard currently in the initial draft stage
2. ANSI/TIA-568.0-D.2, an addendum to address the topology requirements and media choices for single-pair Ethernet; the second addenda to the Generic Cabling standard was approved as a project
3. ANSI/TIA-862-B.2, an addendum to address the use and application cases for single-pair Ethernet; the second addenda to the Intelligent Building Systems (IBS) standard was approved as a project
   
   

There are various topologies, from simple two-connector topologies to four-connector topologies for short channels, and up to 10 connectors in longer channels. There are also provisions for a multi-drop topology in the 802.3cg protocol of four drops in a 25 m channel.

Power delivery and gauge size are in a state of flux given the required study to provide accurate de-rating factors for such a wide range of media; within TR-42.7 (Copper Cabling Systems Subcommittee), a derating table is being created for everything from 18 AWG to 28 AWG. The work within TR-42.1 (Commercial Building Telecommunications Cabling Subcommittee) is looking at providing generic guidance for a short-reach topology (<15 m with two connections) and a more traditional-reach topology (100 m with four connections); I wouldn't be surprised if we also had a long-reach allowance for up to 1 km reach of 802.3cg.

When choosing a media to support telecommunications needs, there are a number of things to consider: signal delivery, immunity to other signals, power and portability/mobility. Single-pair Ethernet has a number of interesting use cases, from applications where signal delivery is important (but bandwidth requirements are moderate) to fixed positions that require power to operate and control end devices.

In the ANSI/TIA-862 IBS world, think of lighting, sensors, controllers, valves and other IoT (Internet of Things) devices. For the commercial space, imagine what using an active consolidation point (CP) could mean. With an active CP, we could provide moderately high-speed connections in a decentralized model using the same grid system we do for wireless.

Even in higher-bandwidth applications, like wireless, there may be a play for single-pair Ethernet; with the increasing use of location-based services, cell sizes are becoming smaller to provide increased granularity.

With the use of these micro-cell grids, there is a corresponding reduction in bandwidth requirements, so 1 Gbps may be a viable solution. What about the penultimate user of high-speed networks? Data centers could also make use of single-pair Ethernet networks for lighting control, PDU monitoring and control, air/moisture/smoke sensors and many more.

For a long time, some in the industry have simply put 4-pair cabling of the latest category at the outlet with some fiber in the backbone. This set-it-and-forget-it view of cabling design has always been risky from both the systems and client perspectives; this "design" was not intentionally built to suit their needs, and is becoming obsolete.

The complexity of requirements, systems potential and options available for structured cabling will harken to the heydays of the ’90s, where structured cabling was considered theoretically and practically as a profession – not merely as a commodity to be provided without active engagement.

With the advancement of networking protocols and applications, and the growth and evolution of IoT, we are seeing the end of the structured cabling world we’ve known so well for the past 30 years.

I, for one, look forward to the future of structured cabling and the ability it will have to service the new technologies available today, on the horizon for tomorrow and even technology we can’t dream of yet.

Want to learn more about what’s on the horizon for single-pair Ethernet? To learn more about specifying your next structured cabling project, watch our recent webinar: How to Specify a Cabling Infrastructure Project.