Is WiFi part of your product strategy? Let Blue Mug handle the details: integration, interop, roaming, power, performance.
Blow your mind wide open at Supernova 2002, a 2-day conference Dec. 9-10 on decentralization, and the fundamental questions facing software, communications, and media.
The above are paid, sponsored links. Contact us for more information.
Subscribe to essays from this site via email. Email to subscribe, or sign up via your Yahoo account.
EtherLinx made a big and unexpected public unveiling in June 2002 when New York Times veteran staff writer John Markoff put them on page A1. In the article, Markoff described how the two founders in a garage a few blocks away from the one in which Apple's Steve Jobs and Steve Wozniak built early computers had a revolutionarily priced wireless product that combined some aspects of mesh networking with long range point to point and in-home Wi-Fi access.
At the time, I was dubious about the combination of price and features: for $150, they said, they could wholesale a piece of customer premises equipment that could handle the back-haul and local service using off-the-shelf products that already had FCC approval.
Forums on wireless raged with questions and debate, while the founders of the company remained most mum about the details. They hadn't been exactly ready to launch their company when Markoff came calling, but couldn't resist the opportunity.
Several weeks ago, I posted a series of questions about EtherLinx that I couldn't answer based on their material, and have been exchanging email with John Furrier, one of the founders, intermittently over the last few months. We finally found a point when he could answer questions via email as well as provide some follow-up replies for more detail.
I asked Furrier, to start with, whether Markoff was accurate when he wrote: The device, which the Etherlinx executives said they believe can be built in quantity for less than $150 each, would communicate with a central antenna and then convert the signals into the industry-standard Wi-Fi, or wireless fidelity, signal for reception inside the home.
Furrier: "Yes, this statement is true. We have a working product that converts our signal over to Wi-Fi. It is actually trivial for us, since we just bridge over to an AP verses the Ethernet cable. Yes, we can do quantities of this combo functionality for less than $150."
I followed up on this question in subsequent emails, asking exactly what he meant by bridging to an AP.
Furrier: "Our product that we are deploying today is a CPE that has Ethernet out (CAT 5)." (CPE is customer premises equipment, a common telecommunications industry term.) "We have working a dual-mode COE" (Central Office Equipment used in a switch room) "which is an EtherLinx CPE that receives the long-haul signal, then bridges it over to a standard Wi-Fi radio, and then broadcasts 802.11b . This is for customers who want to have an 802.11b in their house without having an access point. It would be good for public hotspots. Think of it as a DSL modem with an embedded access point."
Finally, does this mean two radios in one device?
Furrier: "Yes, that is correct. Two different units in one device. We are transparent to the end node whether it's Ethernet Cat 5 or 802.11b broadcast. All of our software is L2 we make no modifications to the PHY layer. It's kind of like a multifunction peripheral." (L2 is network model layer 2, or the part of the networking protocol in which frames are exchanged over a physical medium, layer 1.)
Back to the original email interview. Fleishman: "How does the SDR [Software-Defined Radio] benefit what you're doing given that 2.4 GHz radios are so inexpensive? Is the current silicon not flexible enough for your needs? Most of the focus on SDR is dynamic reconfiguration across larger swaths of bandwidth."
Furrier: "The inexpensive nature of the current 2.4 is a good thing for us since we leverage those devices to load our software on. The 802.11b spec however isn't flexible for us but the PHY layer is."
Fleishman: "How do you differentiate yourself from, say, Alvarion's CPE/CO style offerings that have been out on the market for more than five years (with a high price point still) that have similar specs, and from, say, Nokia RoofTop? The literature makes it sound as though you're combining bridge, repeater, and mesh (although not much of that last point)."
Furrier: "Our big advantage is price and the ability to repeat to create low-cost saturation pools within a given area. Nokia is a pure mesh and very expensive. We are point to multipoint with software modifications to handle bridging and repeating--we also have added additional features to maximize channel reuse, etc."
Fleishman: "How can you reconcile the Fresnel zone with your statements about non-line-of-sight service? Is this entirely based on using passive and CPE repeaters, or have you built a system which offers enough signal diversity to overcome the Fresnel limits that restrict, say, DSSS?"
Furrier: "Some of both, but most of it is in the repeating capability."
Fleishman: "What's the status of FCC approval? In the original NY Times article, Markoff said something about provisional approval, but that kind of approval is hard to get and doesn't last forever. Based on the information you sent, it seems like it would be relatively easily (just costly) to get your CO and CPE device certified.
Furrier: "Already approved by the FCC since 1997...type approved."
The community and industry continues to try to understand how EtherLinx's products will affect the near-term final-mile (or perhaps final-dozens-of-mile) market for WISPs. Feedback is, of course, welcome, and I'll publish (with permission) parts of comments people want to share.