Subscribe to EDN

Home Transportation: Truth In Marketing

August 2, 2007

This blog post references my cover story ‘Home Transportation: Benchmarking Powerline, 802.11, and Ethernet‘ in EDN’s August 2, 2007 edition. It’s one of a series of web addendums to the print writeup.

I was admittedly being sarcastic (albeit, I’d argue, with good reason) when, in the main article, I pointed out of the three ‘200 Mbps’ powerline camps’ technologies:

The three contenders’ products all embed 100-Mbit Ethernet transceivers. Feel free to draw your own conclusions about the disparity between marketing claims and real-life performance capabilities.

To further elaborate on this issue, I’ll reprint a portion of an inflammatory blog post I crafted earlier this year during the CONNECTIONS conference. Consider my three-month-back words in light of this hands-on project’s benchmarking results.

I was reminded of my breakfast conversation with Linksys (who has launched products based on the HomePlug AV powerline networking standard) yesterday, and with NETGEAR (whose powerline networking products are based on DS2 silicon) this morning. Both company’s devices, as well as those of Panasonic, claim ~"200 Mbps" bandwidth. But, as my past testing of DS2, HD-PLC, HomePlug 1.0 and HomePlug 1.0 Turbo consistently reveals, the technologies deliver far lower throughput than their touted PHY rates would otherwise suggest.

Linksys and NETGEAR both gave me the answer to my "what about that disparity?" question that, from my past conversations with powerline backers, I suspected I’d hear…."we’re no different than wireless or any other networking technology". And, in a sense, they’re right; anyone who’s used 802.11b or a/g can agree that the marketed 11 Mbps and 54 Mbps data rates aren’t achievable in real-life conditions. But whereas with wireless you might get around half the promoted PHY rate, with powerline it’s more like one-fifth…..or one-tenth. That is, if it even works at all. Plus, DS2, HD-PLC and HomePlug AV are completely incompatible; not only won’t they inter-communicate, they won’t even "gracefully coexist". Adding one powerline technology to a LAN that’s already got a powerline spur based on another powerline technology will bring down both powerline segments.

If I was in DS2, the HomePlug Powerline Association and/or Panasonic’s shoes, I’d be redoubling my efforts right about now to put aside partisan differences and craft a compatibility compromise that also encompasses realistic performance claims. Otherwise, this finicky but still extremely promising (by virtue of its ubiquity) interconect technology might not ever leave the launch pad, constrained by greedy lawyers and their clients.

The high-performance powerline market frankly is a mess right now, albeit with some geographic segmentation. Analysts tell me that Panasonic’s HD-PLC technology is dominant in Asia, not surprising given that the supplier’s headquarters are in Japan. Similarly, Spain-based DS2 and its UPA approach currently has the lion’s share of Europe’s business, while Intellon-dominated HomePlug reigns supreme in the Americas. Peruse major networking equipment suppliers’ product offerings and you’ll encounter tremendous inconsistency, reflective of powerline networking’s immaturity and consequent uncertainty:

  • Belkin marketed HomePlug 1.0-based products but, to date, has not announced devices based on any faster technology
  • Buffalo sells both HD-PLC and UPA ‘200 Mbps’ gear
  • D-Link offers UPA-based equipment
  • Linksys markets HomePlug AV-based adapters
  • NETGEAR sells devices based on HomePlug 1.0 (unveiled first), HomePlug 1.0 Turbo (most recently) and UPA (unveiled second)

Remember, folks, the only cross-technology interoperability is between HomePlug 1.0 and HomePlug 1.0 Turbo, and that at lowest-common-denominator HomePlug 1.0 speeds. How a vendor with incompatible technologies within the same product category deals with the inevitable consumer confusion in a still-profitable manner is beyond me.

I’d like to particularly focus on NETGEAR with respect to this post’s ‘truth in marketing’ theme. The company’s WGX102 powerline adapter embeds a ‘54 Mbps’ 802.11g transceiver. It’s a slick idea for adding an access point to your network….but the company mates the Wi-Fi circuitry to a ‘14 Mbps’ HomePlug 1.0 transceiver. Pragmatically, I understand why NETGEAR made this choice. By the time the WGX102 entered the market, ‘11 Mbps’ 802.11b gear was passé; designing in and trumpeting the WGX102’s 802.11g capability was the only way that it’d sell. The NETGEAR representative I met with at CONNECTIONS took great pains to point out to me that the company marketed the product as a ‘wireless range extender’, not an ‘access point’. Still, I still can’t get past the fact that HomePlug 1.0 puts a performance choke on the device’s wireless potential….a bottleneck that’s not clearly explained to consumers.

Am I being too hard on the suppliers, folks?

Posted by Brian Dipert on August 2, 2007 | Comments (2)

August 16, 2007
In response to: Home Transportation: Truth In Marketing
Todd Krein commented:

Brian, I agree that the numbers can be misleading. Several technologies have standardized on reporting not only the PHY rate (which *is* relevent), but the usable payload rate ("goodput" as opposed to "throughput") as well. To first order, you can often correlate the ratio of PHY rate to throughput (not goodput) with robustness. Much of the additional PHY-level throughput is used for FEC a/o redundant encoding (like 8b10), or layer-2 acknowledgements. This sort of thing can significantly improve the quality of UDP traffic, even to the point of negating the need for TCP in some applications. Second, having a 100bT interface on a "200Mbps" network isn't as unreasonable as you might first expect. If you have four nodes on a 200Mbps goodput network, that allows two simultanious 100Mbps conversations. With the exception of dedicated media servers and FiOS/DOCSIS3.0 network spiggots, few applications require that kind of instanious, much less sustained, bandwidth. GigE is isn't yet common on the majority of CE computers and devices, so the cost is rarely justified, at this time. Most networking silicon for the protocols you describe require an external microprocessor to handle the Enet interface, as well as the networking stacks. (It might be interesting to open them up and see what processors people are using, and compare adapter costs based on micro, or lack there of.) Today, the difference between a micro with 100bT and GigE may be as much as ten dollars. That can be $30 at retail. How many people really are willing to pay that? Anyway, some thoughts to chew on.


August 7, 2007
In response to: Home Transportation: Truth In Marketing
Gary commented:

I think you miss the point. The 200 Mbps refers to what uis happening on the powerline, not the ethernet cable.

POST A COMMENT
Display Name
captcha

Before submitting this form, please type the characters displayed above. Note the letters are case sensitive:

Advertisement
Advertisement
Advertisement
About EDN   |   Site Map   |   Contact Us   |   Subscription   |   RSS
© 2012 UBM Electronics. All rights reserved.
Use of this Web site is subject to its Terms of Use | Privacy Policy

Please visit these other UBM Canon sites

UBM Canon | Design News | Test & Measurement World | Packaging Digest | EDN | Qmed | Pharmalive | Appliance Magazine | Plastics Today | Powder Bulk Solids | Canon Trade Shows