I often get asked why the wireline home networking market has failed (at least so far) to take off in the same way that the wireless home networking so obviously has done. I believe that there are three fundamental reasons for this.

First and most importantly, the applications driving home networking technologies have until recently been data-driven; shared printers, web surfing and so on. Such applications place relatively low demands on network quality of service. If a data packet fails to reach its destination, it is simply retransmitted and any associated delay does not have a measurable impact on the user experience. Now, however, as we move into an era of entertainment networks supporting multimedia services such as HD-IPTV and collaborative gaming, the demands on the network quality are significantly higher. A dropped packet producing jitter in a high definition TV display is unacceptable to the viewer. That’s where wireline networking comes in. While wireless technologies provide a great solution for data networking, wireline technologies deliver the robustness and reliability required to support new multimedia services.

Secondly, wireline products have not quite managed to reach cost parity with wireless equivalents. Part of the reason for this is a simple economy-of-scale vicious circle; wireline solutions have not yet achieved significant volumes, therefore cannot drive cost, therefore do not achieve volume, and so on. Another part of the reason is flexibility. With relatively low volumes, technology vendors have typically attempted to meet the requirements of as many applications as possible, leading to highly flexible but consequently expensive solutions. With the entry of new technology vendors with tightly focused solutions, the wireline industry is set to break this cycle and finally achieve cost parity.

Thirdly (and now we get to the real subject of this note) is the historical propensity of some participants in the wireline industry to shoot themselves in the foot when it comes to establishing true global standards. Protecting a small pond has been the strategy of record for many, but with recent developments in IEEE and ITU, are we finally ready to jump the final hurdle to mass-market adoption by agreeing on a true global standard? At this point I must declare my credentials. I am a co-founder and VP of Corporate Business Development for Gigle Semiconductor. Gigle is a member of the HomePlug Powerline Alliance and is 100% in favor of standards. I believe that incompatible proprietary solutions inhibit adoption and that true global standards are a necessary prerequisite for the mass market we all aspire to cultivate. Gigle is an active participant in both the IEEE and ITU initiatives.

Much has been made about the architecture of the proposed Homeplug/Panasonic proposal, which is built on two PHYs and a part-merged MAC with a common interface. To be a P1901 device requires implementation of at least one of the PHYs. This approach allows for flexibility to implement either single- or dual-PHY devices. Silicon and product developers can produce a range of devices to suit the needs of specific application requirements. The two PHYs (one based on OFDM modulation and the other based on wavelet modulation) are not interoperable, but both coexist on the same wire without interference through the adoption of a mandatory IPP (inter-PHY protocol). For full flexibility and interoperability both PHYs can optionally be implemented (either in silicon or through system bridging), enabling the market to decide which configuration is most in demand. Is it a perfect solution? No. Is it a significant step forward on the path to consolidation? Yes, emphatically so. The one detractor of this approach is, perhaps unsurprisingly, the entity whose proposal to P1901 was rejected in favor of the Homeplug/Panasonic merged proposal.

P1901 working group meetings are held under the Roberts Rules of Order, which can facilitate moments of drama, tension and comedy (to the insider). The Madrid meeting proved to be no exception. This time around the Homeplug/Panasonic proposal again narrowly failed to achieve the 75 percent threshold, despite the support of every consumer electronics company and service provider in the room together with a majority of powerline access companies (several of which were previously unwilling to support it). Following the agreed process, the group then proceeded to conduct a new elimination round in which previously rejected proposals were brought back for consideration. In the 'In-home cluster', the group once again selected the Homeplug/Panasonic proposal, eliminating an alternative proposal presented by DS2 for the UPA. In the 'Access cluster', the proposal presented by DS2 for UPA-Opera was also eliminated, after receiving fewer votes than separate proposals from Mitsubishi and Homeplug/Panasonic. And at the end of the meeting, Homeplug/Panasonic and Mitsubishi announced that they had decided to merge proposals.

So where does that leave us? In a nutshell, it leaves us in exactly the same position as we were before the Madrid meeting. At the next meeting, which will take place in San Francisco next week, the Homeplug/Panasonic proposal will again be offered up for confirmation. Can the majority of the working group, who want to make progress, succeed in securing the necessary 75% approval? Or will the opponents, who seem determined to block the development of the standard at any cost, once again prevail? We shall have to wait and see.

Meanwhile, over at ITU-T Study Group 15, G.hn proponents are attempting to develop a draft recommendation for a standard covering home networking over coax, phoneline, powerline and CAT5 cable. The group has made good progress this year and has adopted an accelerated meeting schedule. Although there is still much to be done before the specification is finished, the group is aiming to complete the draft PHY section before the end of this year (by itself a significant challenge). The next meeting will take place towards the end of October, and an additional meeting has just recently been scheduled for the first week of November. Careful readers will note that this is at the same time as the next IEEE P1901 meeting.

So, are the wireline industry participants loading up again, ready to take aim at their collective feet by establishing an environment in which two incompatible standards are pushing for adoption? Surely the veterans who have seen this all happening before will step in. Common sense is required. A common protocol (or, at the very least, a common coexistence mechanism) is needed to avoid a situation where these two standards organizations go head to head on the powerline channel.

Mike Wilson
Gigle Semiconductor