How We See CE

Readers of this blog are probably familiar with previous posts (here and here) that discussed some of the problems of the wired home networking industry, especially from the point of view of technical standards (or lack thereof). For many years, 3 incompatible technologies have competed for the powerline networking market, causing significant customer confusion and severely limiting the growth potential for the whole industry.

December 2008 brought excellent news on the standards front: after a lot of effort (and two weeks of late night meetings), ITU (an international standards organization which in the past has been responsible for all DSL/ADSL/VDSL standards) adopted the new G.hn standard for high-speed networking over existing home wiring (power lines, phone lines and coaxial cables). According to the ITU press release, the G.hn standard (which is now officially called G.9960) will provide "up to 20 times the throughput of existing wireless technologies and three times that of existing wired technologies."

The G.hn/G.9960 standard is a significant milestone for several reasons:

• It unifies the powerline networking industry for the first time ever (I have been part of the powerline industry for 9 years and I can say that this is Big News!)
• It unifies the powerline networking industry with the phoneline and coaxial networking industries, to create a single market that will drive volume to the same level that the Wi-Fi/802.11 industry enjoys today.
• It's a truly next generation standard: it brings performance levels that are significantly higher than what is available today.
• It has broad industry support: after the standard was adopted by ITU, four silicon vendors (including DS2) immediately announced plans to support it. HomeGrid Forum (an industry group that has Intel, Panasonic, Infineon and Texas Instruments as board members) is also actively supporting it and plans to develop a comprehensive Compliance and Interoperability (C&I) program to ensure that G.hn-based products interoperate.

The architecture of G.hn is based on a single PHY/MAC specification, which is common across all three types of media, and a Physical Medium Dependent (PMD) sub-layer which is specific for each medium. The rationale for the PMD is that the optimum parameters (for example: frequency bands or power levels) are not always the same for all media. For example, in coaxial cable sometimes it's desirable to operate in the 800-1600 MHz band, while for power lines the most attractive frequencies are usually those well below 100 MHz.

Common parts of PHY/MAC include the modulation scheme (Orthogonal Frequency Division Multiplexing - OFDM), Forward Error Correction scheme (Low-Density Parity Check - LDPC), encryption algorith (AES), MAC frame format, QoS architecture, retransmission schemes, etc.The G.hn MAC will provide support for both guaranteed bandwidth reservation and best effort service, which will accommodate the requirements of a wide range of video, audio and data application.

G.hn has been designed specifically to make it easy for silicon vendors to implement multi-wire products: because the majority of PHY/MAC elements are common for all three wires, the extra silicon complexity of simultaneously supporting powerline, phoneline and coaxial cable (as opposed to supporting powerline only) is negligible. This will allow silicon vendors to address multiple markets with one single product, which will significantly accelerate cost reduction and technical innovation.

G.hn is already sending a positive signal to manufacturers of Consumer Electronics devices that were reluctant to embed powerline/phoneline/coaxial networking interfaces due to lack of international standards. My conversations with many device manufacturers at CES this week confirm that they are really excited about the possibilities that G.hn provides, and that we can expect a wide range of G.hn-compatible networked-CE devices (Set-Top Boxes, Routers, TVs, Blue-Ray players, etc) to reach the market in 2010.

Chano Gomez
DS2