While high-speed wireless networking is moving rapidly to fully-interoperable products, powerline networking seems to be going in the opposite direction. Three entrenched groups with comparable but non-interoperable product lines are coming to market simultaneously. The products appear attractive, but consumers will find it hard to sort through the conflicting claims, and nearly impossible to figure out what will work with what.
Three Warring Camps
We have written many articles about the evolution of home powerline networking. In tests we conducted at our house more than four years ago, we found that the original HomePlug 1.0 products worked more consistently than the early versions of Wi-Fi available at that time.
As the second generation of powerline networking was being developed, the industry split into three camps: the HomePlug Powerline Alliance (HomePlug), with Intellon as the major semiconductor provider; the Universal Powerline Association (UPA), led by chip maker DS2; and High Definition PLC (HD-PLC), led by Panasonic. All three groups now have similar products on store shelves.
We're planning to conduct tests of these new products similar to the tests we ran four years ago. Our earlier experience suggests that they will work quite well, but will probably fall well short of their touted speeds.
Unfortunately, the standards war may make their performance irrelevant. Devices from one camp do not interoperate with devices from the others. Even worse, they interfere with each other if they're connected to the same electrical lines. While the powerline camps compete against each other, other networking technologies are likely to win the market.
HomePlug is the oldest of the powerline networking industry groups. Products based on its original standard, HomePlug 1.0, reached the market more than four years ago.
At CES in early January, we visited the HomePlug booth and talked with Oleg Logvinov, President & CEO of semiconductor maker Arkados and HomePlug's Chief Strategy Officer. Many of the products on display were based on the original HomePlug 1.0 (advertised as 14 Mbps) and on Intellon's 85 Mbps HomePlug 1.0-compatible "turbo" chips. A few products were based on the new HomePlug AV, operating nominally at 100 Mbps.
Most of the products shown were various forms of HomePlug Ethernet adapters--using HomePlug to extend an Ethernet network throughout the home. We were taken by one new product from Asoka--a combination surge-protected power strip and Ethernet hub based on HomePlug 1.0.
After a long gestation period, products based on HomePlug AV have started appearing. Typical of these is the Linksys PLK200 PowerLine AV Ethernet Adapter Kit. One online retailer describes it as conforming with the "HomePlug AV PowerLine network standard" and providing "up to 100Mbps data rates". Various online retailers show prices between $200 and $235 for a kit with a pair of units; individual units are priced at about half that.
Originally focused on high-speed chips for access BPL, DS2 has had considerable success in serving the IPTV market in Europe and Asia. The UPA standard based on DS2's technology is now appearing in retail networking products.
At CES, we met with Victor Dominguez, DS2's Director of Strategy and Standardization. We thought that DS2 was a fairly new company, but Victor--one of its founders--said the company started in early 1998. He said DS2 provides a broad range of firmware and software for developers, and told us the company has about 140 people of which "90% are engineers and 65% are working on R&D". He said the company has a strong focus on IPTV, using both coax and powerline wiring to provide full-house coverage for several standard-definition or high-definition television streams.
Typical of recent DS-2 based products is the Netgear HDX101 Powerline Ethernet Adapter. One online retailer describes it as a "Powerline HD Ethernet Adapter" with "data rates up to 200 Mbps". A kit with two devices sells for about $165; a single adapter is about $84.
A recent example of home networking devices based on Panasonic's technology is the Panasonic BL-PA100KTA HD-PLC Ethernet Adaptor Starter Pack. One online retailer says it provides "speeds up to 190 Mbps" and sells a pair for $200. It cautions "This product may interfere with ... other PLC adaptors which do not use the HD-PLC standard."
How to Choose?
These devices are all powerline to Ethernet adapters operating at about the same claimed speeds. They appear next to each other on store shelves and on retail websites such as PC Connection and CompUSA. All are offered as kits of two adapters and as single add-ons, and all are close to the same price point.
How will the customer choose between them? If they choose based on the speed claims (100, 190 and 200 Mbps), they're likely to find that none will achieve more than a fraction of its claimed speed.
Even worse, unless the customer has researched these technologies, it's nearly impossible to figure out what works with what.
This is a great formula for a marketing disaster.
The End Game
Since we ran our first tests more than four years ago, we have believed that powerline networking had a good chance to become the default winner for "no new wires" high-speed home networking. Since most home devices are plugged into an electrical outlet, we think most consumers would prefer to use PLC if all it took to connect devices together in a network was to plug them in. If PLC chips were embedded in most devices, wireless communications would probably be relegated to supporting the relatively few mobile devices used in the home.
But the standards war makes it very difficult for makers of PCs, set top boxes, TV sets and the like to commit to embedding powerline chips.
HomeRF versus Wi-Fi
Wireless networking was once also divided into two warring camps. When "Wi-Fi" got its name in April 2000, it faced a formidable competitor in HomeRF. HomeRF was arguably a better standard for home networking--it had built-in quality of service mechanisms to separate time-dependent voice traffic from data--and Intel was one of its strongest proponents. Although Wi-Fi was primarily directed to--and optimized for--the enterprise data market, its proponents argued that employees would bring their laptop PCs home, and would not want to swap wireless cards each time they moved between the office and home.
Wi-Fi and HomeRF products sat right next to each other on store shelves, and neither retail employees nor consumers could sort out the conflicting claims. The lack of volume kept prices high. To nobody's surprise, very little wireless networking technology sold through to the consumer--until Intel reversed its position.
When Intel abruptly switched camps in April 2001, Wi-Fi took off quickly as chip makers focused their attention on a single standard. The Wi-Fi logo assured consumers that products would be interoperable, and prices dropped quickly as volumes increased. When Wi-Fi chips started appearing in notebook PCs, and Intel started its Centrino program to make everybody aware of Wi-Fi, the game was over and HomeRF disbanded.
"All Fall Down?"
We think the competing powerline networking camps will be fortunate if the scenario plays out like wireless, with one camp "winning" against the others. It seems more likely that all will lose.
As we wrote last year, these various powerline devices interfere with each other, competing for the bandwidth of the electrical lines. If several different technologies are used simultaneously in the same home, neither will work properly. This is unfortunate because different technologies might try to share the same electrical lines:
The CEPCA group, which includes most major consumer electronics manufacturers, has pleaded for the widespread adoption of a "coexistence" standard--thus far to no avail. Members of the HomePlug camp feel that interoperability is the only valid approach.
Many other technologies are competing for the same market niche. Cable and telephone companies have already deployed MOCA over coaxial cables, and telephone companies have deployed HomePNA 3.1 over phone lines and coax. As we discussed in the previous article, IEEE 802.11n will be on the market later this year with a "Wi-Fi" certification logo. With the powerline camps at war with each other, there's a substantial risk that they will kill each other off--clearing the path for these other approaches to home networking.
A Way Forward
In the past few days, we raised these points in telephone discussions with Oleg Logvinov of HomePlug and Chano Gomez, DS2's Vice President for Technology and Strategic Partnerships and head of its US office. Both agreed that confusion was building in the marketplace, and felt the HomePlug and UPA groups could play a role in educating retailers and consumers.
HomePlug and UPA have developed "badges" that appear on retail products that have passed their conformance and interoperability tests. Chano said the UPA badge appears on many products on retailer shelves: "Go to Best Buy and Frys, you will see our logo."
Both Chano and Oleg agreed that consumers--and retailers--could better understand what works with what by positioning these badges on retail products sold online. But both also pointed out that semiconductor companies and trade groups have little leverage over equipment makers and even less over online retailers. So it's likely to be an uphill battle.
Several retail websites claim the HDX101 is "powerline-friendly — easily coexists with NETGEAR's Wall-Plugged or HomePlug compatible products". We didn't think DS2 and HomePlug products would work properly when sharing the same powerline, and asked Chano whether Netgear's claim was true. He said Netgear has the largest share of the consumer powerline networking market, with a large installed base of products based on HomePlug 1.0. He said Netgear "took advantage of our software to customize our chip, so the software is different from our default. HDX101 automatically detects whether a HomePlug 1.0 device is there or not. It can coexist with HomePlug even with degraded performance, reducing bandwidth by half or something like that."
In our discussion, Oleg placed the priority on interoperability, and said coexistence will take just as long: "If you have to respin the silicon, you can do interoperability as quickly as coexistence. Industry analysts have to educate the public that it's important to unify the interoperability standard."
Chano reiterated the need for coexistence between home networking, access, and television products. "UPA and CEPCA have filed a joint coexistence proposal with ETSI, and most likely will also file one with IEEE." He was hopeful IEEE P1901--the IEEE Task Group on BPL standards--would resolve both issues: "We have the right forum in P1901. They have created three clusters - access, in-home, and coexistence. Proposals are to be submitted in June or July."
Major potential customers such as broadband service providers, consumer electronics manufacturers, and power utilities clearly prefer a single interoperable standard. We think all three powerline camps should be able to agree on the need for interoperability as a mutual long-term objective. If they can agree to that, perhaps they could then agree to work together on a fast track to coexistence--if it can be implemented faster.
There's not much time left.
For further reference:
( www.homeplug.org ) ( www.intellon.com ) ( www.upaplc.org ) ( www.ds2.es ) ( www.hd-plc.org ) ( www.panasonic.com ) ( www.arkados.com ) ( www.asokausa.com ) ( www.cepca.org ) ( www.mocalliance.org ) ( www.homepna.org )