BBH Central IconBBH Report Home PageSandy and Dave
  CENTRAL home  |   REPORT home About/Contact Us  |   Subscribe  |   Index by Topic  
The May 15, 2008 Issue Provided by System Dynamics Inc.
Table of Contents Print this article Email this article to a friend

The Everywire Standard: G.hn and HomeGrid Forum

We recently received an email announcing HomeGrid Forum, a new industry alliance for next-generation home networking formed by Infineon, Intel, Panasonic and TI. There already seem to be too many home networking alliances, so we wanted to understand why they were leading the effort to form yet another. To learn more about the new alliance, we talked on the phone with two of its leaders. We also talked with other leading players in "existing wiring" home networking.

HomeGrid Forum "aims to promote and influence a single, next-generation worldwide standard for networking digital content, such as movies, music and pictures, over home wiring." G.hn, a relatively-new standards effort operating under the auspices of the ITU, is currently working to create a single global standard for networking over all existing wiring. HomeGrid Forum sees itself as playing the same role for G.hn that the Wi-Fi Alliance plays for IEEE 802.11: compatibility and interoperability testing, branding, and promotion. It also aims to facilitate the standards development process.

We think there really is an unfilled need. While it will take some time to play out, G.hn and HomeGrid Forum are likely to play a key role in resolving the current confusion in "existing wiring" networking.


"Existing Wiring" Networks--The Unfilled Need

We're strong believers in "whole home" networking -- integrated networking designed to move audio, video and data around the home. Category 5/6 wiring is still the best way to do this, but everyone in the industry believes that most families will not pay to install new wiring throughout the home. "No new wires" has been the industry mantra for many years.

Some have thought wireless networking would be sufficient for "whole home" networking. We have long believed that wireless would often fall short--especially for multiple channels of high-definition video. Wireless signals are degraded by distance, walls, floors and air conditioning ducting, and are subject to interference from other wireless networks and other sources. While 802.11n (the latest generation of Wi-Fi) may be sufficient for smaller homes and apartments, some mixture of wireless and wired networking will probably be required to provide high-quality networking in every home.

Logos for five "existing wires" technologies --> Click for larger pictureExisting wiring -- coaxial, telephone, and electrical -- has great potential to carry digital signals throughout the home. Many companies have developed chipsets and products to use these wires. Industry alliances have been formed to promote them, including:

Several generations of "existing wiring" technologies have appeared in products. Some telephone service providers and cable operators have deployed products based on MoCA, HomePNA, HomePlug and UPA. Products based on the latest versions of HomePlug, UPA, and HD-PLC are widely available at retail.

Wireless networking is based on a single family of IEEE standards--802.11n is the latest version. A single industry alliance--the Wi-Fi Alliance--handles certification and promotion on a global basis. The alliance website says "the Wi-Fi industry shipped 300 million chipsets in 2007 and is growing at a compound annual growth rate of 40 percent."

The total penetration of all the "existing wiring" products is minuscule compared with Wi-Fi. The shipments of each of the "existing wiring" chipsets is in the few millions -- a tiny fraction of the global broadband market.

There are far too many standards for networking over existing wiring. It's bad enough that different standards are used for each type of wiring. But there are also two competing standards for coaxial wiring, and three for electrical wiring. Service providers have to pick sides in deciding which wiring to use, and which standards to use over that wiring. Consumers walk into stores trying to solve a problem, and see nearly identical boxes with competing flavors of powerline networking.

Many companies working on one type of existing wiring are already looking for opportunities on other wiring. HomePNA, originally designed for phone wiring, now supports both phone and coaxial. UPA has been proposed for coaxial as well as electrical. But even if they each could be made to operate well over all types of existing wiring, five mutually-incompatible technologies that often interfere with each other would be intolerable.

There's clearly an unfilled need for a unified approach to existing wiring. Such an approach would permit service providers to build a single chipset into every device and use the appropriate wiring suitable to the particular situation in the home. It would eliminate the confusion which almost certainly inhibits most consumers from adopting these technologies.

A unified approach is not as far-fetched as it might seem. All of the current "existing wiring" technologies are based on similar underlying techniques. Nearly all use OFDM (as does Wi-Fi). They employ similar techniques for security and for sharing the available bandwidth among many users. That's what led to G.hn.


ITU-T G.hn

ITU-T G.hn has been working for more than two years "to specify next generation home networking transceivers (PHY and MAC) capable of operating over premises wiring including inside telephone wiring, coaxial cable, power line wiring, data grade (e.g., CAT5) cable, and combinations of these." This work is being done in the ITU-T Telecommunication Standardization Sector, Study Group 15, Question 4 (ITU-T Q4/15). (For those unfamiliar with ITU nomenclature, ITU-T is the global organization of telephone companies; its Telecommunication Standardization Sector establishes global standards in the form of ITU Recommendations; Study Group 15 is responsible for DSL, PON and other technologies used for digital transmission; and Question 4 is its group responsible for "Transceivers for customer access and in-premises networking systems on metallic conductors".)

More than thirty companies are now working on G.hn. Most of the "existing wiring" semiconductor companies are active participants. Participants say a unified standard is well under way and should be completed during 2009. Many companies expect products will start becoming available in 2010.

Who will test these products to prove that they conform with the G.hn standard and interoperate with each other? That's where HomeGrid Forum comes in.


HomeGrid Forum

We talked on the phone with two leaders of HomeGrid Forum: Matt Theall of Intel, and Mike Bourton of Texas Instruments. Matt and Mike also represent their companies in the HomePlug Alliance--until recently Matt was President of HomePlug. So we started by asking why the world needs yet another networking alliance. Pointing to recent work by DS2, Gigle, and some of the UWB companies, they observed that "most technology leaders are moving towards an 'any wire' technology". As each company and alliance moves to embrace multiple wires, they said "the risk of incompatible technologies running on multiple wires is high" and saw "fragmentation in the industry".

They said they like G.hn, since it brought together "seven or eight service providers" (including AT&T, Verizon, France Telecom and Telenor) with "most leading silicon providers". They said G.hn already has "the critical mass for the ecosystem" and was making good progress toward a unified standard.

But they saw a gap -- they felt "the odds of fragmentation are high" and there needed to be "a companion SIG to G.hn." As with most formal standards organizations, G.hn's efforts are finished when the standard is published. But getting a standard into the market requires many years of additional work--to test devices for compliance with the standard and interoperability with each other; to promote the standard; and to provide recognizable branding so service providers and consumers are assured that devices will work together. Pointing to the role the Wi-Fi Alliance plays for 802.11, and WiMAX Forum plays for 802.16, they formed HomeGrid Forum to play these roles for G.hn.

Since the G.hn standard is more than a year away, we asked Matt and Mike why they were launching HomeGrid Forum now. They said most people underestimate how much work and time are required to establish certification testing methodologies, choose test labs, and get the process under way. They felt that creating the alliance now would provide enough time to get ready for testing while the standards are being completed.

HomeGrid Forum is also helping its members develop common positions for presentation at G.hn meetings. Mike said much of the time at these meetings is taken up by individual companies presenting position papers that often have many things in common. Some G.hn members already work together on some issues and present joint position papers. The HomeGrid Forum's Contribution Work Group is holding weekly meetings for members to hammer out common positions, encouraging a larger number of companies to air and resolve their differences prior to G.hn meetings. Forum members don't have to agree with these positions: "If you agree with the position, put your name on the presentation."

We asked why Intel and TI were taking a leadership role, since neither company has been much involved with existing wires networking. Mike said that TI "sells many products for the digital home - with lots of TI chips in those products". But he said that TI "needs one home networking strategy." He said "home networking won't be in a large market unless we solve this set of problems." Matt said Intel wanted to include existing wiring in its products, but the current fragmentation makes it impossible to choose one over another.

Finally, we asked why so many of the leading players seemed to be missing from the announced Forum membership. Some "existing wiring" semiconductor makers--DS-2, Gigle, and Pulse~LINK--are listed as members. But others--including Coppergate, Entropic, and Intellon (founders respectively of HomePNA, MoCA and HomePlug)--are conspicuous by their absence. Matt and Mike said HomeGrid was still in its early stages: "We're in a startup situation now -- in time, people will join." Nearly everyone working on existing wiring is already an active member of G.hn. "Of the 35 companies participating in G.hn, 11 are already members of HomeGrid, and another 11 are planning to join."

Other Views on HomeGrid

We talked off the record with some of the companies who are not members of HomeGrid. All said they'd been approached by HomeGrid, but had decided not to join "at this time". Most felt the effort was premature, and several expressed apprehension about Intel's and TI's motivations. Some think the big guys are trying to take over the effort, while others think they're trying to subvert it; one said "All of Intel's public commitments are to wireless. They're the leaders in Wi-Fi and UWB and they don't want existing wiring to succeed."

We can understand why the leading companies in existing wiring networking would be apprehensive. All are startups--they're minnows compared with Intel and TI. All of their assets are invested in their own technologies. They've all worked hard to create industry alliances to promote their specific technologies, and feel they're finally starting to get some market penetration. Millions of shipped chipsets might not be much for Intel and TI, but represents good progress for the startups.

That could all be imperiled by G.hn. They're participating in G.hn because they see the handwriting on the wall: whether or not they'd admit it, their current technologies don't fill all the needs--most critically for a single standard that works over all existing wires anywhere in the world. They're working energetically to create a unified standard, and hope to be in the market early with chips for the new standard.

But admitting that openly could diffuse their existing marketing message. They all have chipsets in the market now, and their continued existence depends on companies building products around those chipsets and consumers buying those products. They'd rather keep G.hn far off the radar screen, and don't appreciate HomeGrid Forum calling attention to it.

They probably do have some time before G.hn becomes a factor. We've learned painfully that things never happen as fast as enthusiasts would like, and many good ideas never happen at all.

We'll keep watching G.hn and HomeGrid Forum, and keep track of its progress.

For further reference:

( www.homegridforum.org ) ( www.itu.int ) ( www.wi-fi.org ) ( www.mocalliance.org ) ( www.homepna.org ) ( www.homeplug.org ) ( www.upaplc.org ) ( www.hd-plc.org ) ( www.itu.int/ITU-T/lighthouse/sg15.phtml ) ( www.intel.com ) ( www.ti.com ) ( www.wimaxforum.org )