Receive new posts as email.
This site operates as an independent editorial operation. Advertising, sponsorships, and other non-editorial materials represent the opinions and messages of their respective origins, and not of the site operator. Part of the FM Tech advertising network.
Entire site and all contents except otherwise noted © Copyright 2001-2010 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
Analyzing traffic since the start of extensive tracking of this site shows that readers like knowing about security flaws, improvements: It's a slow news day, and I've been looking at traffic analysis of Wi-Fi Networking News to see what our top stories have been since we started using Omniture reporting last fall. Over the last 10 months, trends are clear, driven by Slashdot and other sites that refer traffic: the most popular single stories on the site focus on security -- six out of 10 stories. Two were about Wi-Fi detectors, and the other two on unrelated topics.
As security remains a hot issue in the industry in general, you can expect that we'll continue to follow it. The top 10 stories of the last 10 months are:
Details are sketchy, but New York City may allow six telecom firms to pay up to $25 million per year to install wireless transmitters on 18,000 lamp posts: The article is full of sturm und drang about health effects, but the real story is that the city is trying to counter its dead zones without tearing up the streets. It's unclear precisely what kind of transmitters these will be, but you can bet your boppy that the goal will be wireless backhaul for the majority of the points using mesh or simple point-to-point. This endeavor could bring massively improved voice, 2.5G/3G cell data, and Wi-Fi into a city without ripping up all the roads once again or putting giant cell antennas on every last building.
The companies include well-known and never-heard-of-'em: the New York Post says they are two cellular providers, Nextel and T-Mobile, three non-cellular companies, ClearLinx Network Corp., Crown Castle Solutions, and Dianet Communications. The sixth, IDT Business Services, will provide telephone service via the Internet. [link via GigaOm]
Sprint will include a Wi-Fi user interface on its Novatel wireless data cards: The feature just offers a single interface--users must still use a separate Wi-Fi card or their laptop's built-in Wi-Fi capability to connect. Having a single interface means that customers will have the same experience whether they are using a Sprint-owned hotspot, a hotspot from an operator that Sprint has a roaming agreement with, or the PCS network, said a Sprint PCS spokeswoman.
A single interface like this is surely helpful but it seems that at this stage the market is ready for tighter integration that includes the access on a single card as well.
Posted by Nancy Gohring at 2:33 PM | Permanent Link | Categories:
Grand Haven, Mich., makes splash with full-city Wi-Fi coverage: This seems like yet another city announcement, but it might be the first city with this scale of access that 100-percent live and commercially available. (Dissenters, please write in.) While there are plans for Cerritos, Santa Clara, and Chaska (Minnesota) to have full coverage, Grand Haven may have beaten them to full deployment. (Take a look at this excellent page designed by Ottawa Wireless's PR firm: it's exactly the kind of clueful PR that should be encouraged: he's putting his money--or maps--where his mouth is.)
The folks at Ottawa Wireless sent out a press release full of the technical details, such as their support for 802.11a, b, and g, and the fact that their service extends 15 miles into Lake Michigan, providing access for boaters and marinas. The coverage extends six square miles across the town, and it's optimized to handle VoIP; a beta test is in progress right now that will cost $30 per month for unlimited calling nationwide.
The service has 300 subscribers at its formal launch out of a local population of 12,000. However, th town sees two million--yes, million--visitors a year. Customers include the city, and public safety and health groups will eventually use the network.
Ottawa Wireless cleverly layers applications and specific performance statements on top of its network, providing VPN, wireless video monitoring, and what they call 55 mph access--even while you're speeding across the lake, you can still use Wi-Fi.
Service costs $20 per month for 256 Kbps, $45 for 512 Kbps, and $80 for 1 Mbps access. Day rates run from $5 for a single day to $17 for seven days (less than $2.50 per day), obviously aimed at the vacationer who spends a week or several weeks in town or on the lake.
For $25 per month, you can get unlimited in-town roaming as a plan, or, if you spend $200 upfront, you get 256 Kbps service in your home, plus a home wireless network, and unlimited roaming. That's a good bundle. The 512 Kbps and 1 Mbps business services include one account for unlimited in-town roaming.
My friends and colleagues Richard and Angie recently passed through Grand Haven as part of their 8,000-mile RV trip and vouched for the service.
AUT will rely on Reach Wireless to install a campus-wide, commercial hotzone: In the U.S., we're used to universities and other educational institutions installing Wi-Fi as part of their basic infrastructure, including the cost of operation out of their physical plant or IT budget, alum fund, students fees (partly hidden), or other endowments. Auckland, New Zealand's technology institute has turned to commercial operator Reach Wireless, which employs RoamAD's quasi-mesh/cell Wi-Fi architecture, to deploy a hotzone across campus and handle the finances.
AUT will own the infrastructure and carry the hotzone's traffic over its own Internet feed. But Reach Wireless will operate the service, collect fees, and provided customer and billing support. Students and faculty will be charged NZ$16.95 (US$10.50) per month. Existing Reach customers and the general public will also be able to use the campus network. Current subscribers won't pay extra for access; vice-versa, AUT affiliated subscribers will have Reach within reach at a small additional cost off campus.
Here's the kicker from a press release: The network was designed in one day and deployed in less than a week.
Nigel Ballard of Personal Telco reports that Portland (Oregon) International Airport will have free Wi-Fi: Ballard told an audience at a meeting of the community wireless group this evening that the Port of Portland will turn on 25 access points by Oct. 1 to offer free service at gates and check-in areas. They're committed to covering the cost of operation for the first year, and then re-evaluating whether fees would be added. Ballard is part of the Portland Telecommunications Steering Committee, and an active community networker and commercial infrastructure builder.
The NY Times follows some of those wires (and wireless signals) around the DNC's FleetCenter home: This thorough report by Seth Schiesel follows some of the complexity managing wired and frequencies at an event of the scale of the DNC. After reading this article, I'm amazed that anything manages to work. Other stories in this vein indicate that thousands of miles of wire were pulled for this week, while the RNC venue in New York might top 40,000 miles because of some slightly longer distances involved in two spaces being used.
The Wi-Fi problem is clearly explained, and it appears that the planners did hope to reserve space for Wi-Fi. I'm guessing that the wireless equipment used by camera operators is incredibly noisy, spewing out far more than is legal out-of-band (slop-over) signal. Because Wi-Fi has such a low amount of legal signal, it's very likely that the electronic newsgathering (ENG) is treading all over its neighbor's space. There should be at least a few clear Wi-Fi channels.
The network is apparently geared to handle the equivalent of 3,000 T-1 lines--but tell that to my buddy who not only didn't get his paid-for T-1 line drop, but was told there was nothing that Verizon could do about it.
A number of outlets are critical of the Motorola Wi-Fi/cell phone: TechDirt and TheFeature emphasize how limited the phone's use will be because it only operates on 802.11a. Also, as we noted yesterday, it sounds like this platform would make a tough sell because it requires APs from Avaya and Proxim or upgrades to existing Proxim APs.
The solution isn't ideal and neither is the HP/T-Mobile device, which doesn't include voice capabilities over Wi-Fi. These are pretty typical first attempts and they'll certainly improve with future iterations. However, the enterprise solutions like Motorola's will have a tough road ahead of them. Cellular operators are typically very slow to embrace change, especially anything that may be perceived as threatening their voice business, which voice over Wi-Fi may. I'll be interested to see which operator Motorola actually launches this with and which enterprises actually use it.
Motorola, Avaya, and Proxim, today introduced an enterprise platform that enables voice roaming between enterprise WLANs and the wide area cellular networks: The solution includes a new handset from Motorola that looks like a typical cellular flip phone but can support voice over WLAN as well as voice over a GSM network. As part of the solution, enterprises must implement APs built by Proxim and Avaya, a call manager gateway from Motorola that enables the handoff between the networks, and an Avaya IP-PBX.
The phone automatically reverts to the WLAN when it's available and can seamlessly hand off calls from the WLAN to a GSM network as a user moves between them. While it looks like a cell phone, it features a lot of the capabilities of a desk phone such as buttons for mute, hold, and speakerphone. It runs Win CE so can support Microsoft .Net applications and it includes a VPN. The gateway enables push-to-talk while users are covered by the WLAN.
The platform offers some features that aren't available on the PDA introduced yesterday by HP and T-Mobile, namely voice over WLAN. "It's exciting to see the HP/T-Mobile solution, but it's an iPaq that has GSM voice on it," said Chris White, director of business development for enterprise seamless mobility with Motorola. "It doesn't do VOIP except with a softphone."
In addition, because of that WLAN voice capability, the Motorola solution supports a single phone number that rings for users regardless of the network they are connected to. Users can also use a single mailbox and access many of the same PBX-type features both inside the office over the WLAN and outside on a GSM network.
I originally thought, as reported here earlier, that one major difference between the HP handheld and the Motorola phone was GPRS but it turns out that the Motorola phone does also support GPRS. Motorola doesn't mention GPRS in its press release and didn't mention GPRS during its hour-long conference call this morning. It's surprising that Motorola would want to bury that fact.
Another interesting difference between the HP device and the Motorola phone is that the Motorola phone operates on 802.11a. Motorola and Avaya spokespeople said they chose 802.11a because the networks can handle more capacity than the other flavors of 802.11.
Sales, which will be handled at least initially by Avaya and not a cell phone operator, might be challenging because the decision to implement such a solution is complicated. An enterprise would have to decide to potentially switch existing cell phone users in the company to whichever GSM operators may support the phone--the companies haven't said which may support it. It also presumably means that a company might grapple with feeling the need to sign up additional cellular users in an effort to standardize on the phone.
The decision also may be more complicated for enterprises that already have WLANs. This solution requires companies to either upgrade existing Proxim APs or deploy new APs, which will support both the voice calls and existing laptop or other handheld based data access. However, as part of this announcement, Motorola said it is forming an industry group to standardize such platforms so that end users might have the option of using different vendors. It will remain to be seen if other companies will be interested in following Motorola's lead.
The solution will be available some time later this year and it will be introduced with a carrier which has not yet been named.
Reports from the field indicate that the Democratic National Convention Wi-Fi coverage is abysmal, other connectivity hard to find: Contrary to pre-event reports, there is Wi-Fi at the DNC. However, they haven't done much to make it robust, and the ENG (electronic news gathering) spectrum overlap is probably killing the signal, too. (ENG is a secondary licensed user across some of the Wi-Fi band.) Some reports indicate that wise consumers came equipped with cell data cards. Slow, but apparently reliable.
My colleague Dave Sifry called earlier today from the floor, bemoaning the lack of consistent service. He and colleagues called around Boston to try to find 802.11a or a/g access points, which use the 5 GHz band--which would have no overlap in the convention center. None to be had.
Blogging and Wi-Fi go hand in hand like ham and rye, mom and apple pie...Kerry and Edwards? Take away wireless access from a blogger, and it's just a guy or gal with a laptop.
Dave Winer, an ur-blogger covering the convention, quotes an email from a DNC coordinator who notes that Wi-Fi (with the delightful name of Corky1221) will be in the blogger area. Here's a blog by the guy who appears responsible for the Wi-Fi network with a lot of detail on what he's contending with. Oddly, he complains about 31 users on a single base station, which is a trivial number of users even for most consumer models to handle, and he's working with Ciscos: the giveaway is that he talks about using a WLSE, which is Cisco's aggregation product that works only with their own access points. The fellow sounds remarkably technical adept--but you can only cope with ENG overlap to some degree.
The DNC should have gone 802.11a or a/g, and simply said, look, if you want reliable connectivity, you're going to have to spend $80 on this card or something like it. Even Mac users could get an Atheros-based card from NetGear or D-Link and use OrangeWare's driver for 802.11a. Otherwise, plug in.
Posted by Glenn Fleishman at 9:40 PM | Permanent Link | Categories:
Airports are finally getting the Wi-Fi religion, but only 19 of top 50 airports have service; 6 are deciding on vendors: I know regular readers of this site are used to seeing my byline on any Wi-Fi obsessive article in the New York Times, but I can't be everywhere. Jane Levere writes an extremely tight wrap-up on the current and future state of Wi-Fi in airports, focusing on pricing, utility, and the lack of roaming plans.
As noted earlier today, SBC has signed roaming agreements with a number of airport providing hotspot operators: if it added Sprint PCS and AT&T Wireless's locations, it would have the hat trick. But until SBC actually adds it's newest partners into their network and signs additional locations, we've got a pretty scattered set of airports under any given plan.
One item not mentioned that affects Wi-Fi in airports is the airlines and concessionaires sudden ability to put in their own networks per a recent FCC clarification that only they can regulate unlicensed spectrum, not landlords or airport authorities. This report is just a month old, and I'm sure that airlines, retail shops, and other tenants of airports are still digesting the decision and beating airport authorities over the head about it. We'll very likely see a number of interesting airport options that crop up with that artificial landlord restriction removed. The authorities may try to keep a lid on through other methods, but the FCC handed tenants a blunt instrument.
(Now, don't get on me about spectrum etiquette and coordination. It's a good idea. If you have 20 networks operating in a small space all on the same channel--very bad. Tragedy of the commons. And so forth. But etiquette is enforced when no one gets utility unless everyone cooperates. Prisoner's Dilemma and all that. Thus, spectrum coordination for unlicensed use can happen ante or post facto.)
They're late to the game, but they're ready to party: It's a funny thing. When SBC Communications first announced their FreedomLink plans last year with plans build 6,000 hotspots over a couple of years, it seemed like yet another announcement of large numbers with no track record. Cometa was still on its 20,000 hotspots prediction and had only a handful. McDonald's hadn't decided its partner and was in limited trials. Wayport seemed stuck on hotels. And T-Mobile stayed focused--as it still does--on a few ubiquitous chains.
In the space of a few months, SBC has moved from last man in, to practically first mover. Let's review:
SBC has not so quietly assembled what will be the largest roaming network with a flat-rate price in about two to three months from now and then grow much larger than any other network in the U.S. They've set a good price. They have bundling deals yet to come. And they cover plenty of business territory with hotels, airports, and other venues.
Could it be that SBC drives other telcos into reselling and building Wi-Fi on a scale that's only been hyped before now? Or will SBC's drive be a strategy they follow alone, leaving T-Mobile with its private network (resold only to iPass), and aggregators like Boingo selling all the pieces outside of Wayport's retail Wi-Fi World locations and SBC's FreedomLink exclusive venues?
The test of whether a build out will happen is whether the money and commitment is there. Cometa had the commitment, not the money. Ditto, MobileStar. T-Mobile had both and has built out everything they promised when they formally launched Starbucks networks. We'll see if SBC follows suit, but there's no denying that they throw the cash around when it has this many positive strategic implications for them to retain customers, receive incremental revenue from existing customers, and provide a long-term data and voice game plan for them.
T-Mobile and HP said today that they'd start selling a new handheld that includes GSM, GPRS, Wi-Fi, and Bluetooth later this summer: There are some combined cellular/Wi-Fi PDAs being sold in Japan and probably one or two here too, but this will be one of the first major launches by an operator.
One of the more interesting aspects of this announcement is the type of pricing deal T-Mobile will introduce when the device becomes available. It will offer a plan that includes a bucket of voice minutes plus unlimited use of GPRS and Wi-Fi. The whole plan will cost under $100. Some of the cellular operators have been charging around $80 a month for unlimited cell data using data cards, though those prices have been dropping recently. Data plans on cell phones have been less expensive, but presumably users aren't downloading PowerPoint presentations on their cell phones, unless they are connecting to their laptops, which the operators frown upon. So while the $100 figure from T-Mobile seems like a good deal by comparison and because it includes voice, Wi-Fi, and cell data, it's still pretty steep when you compare it to what people pay for high-speed access in their homes. Users will likely expect to pay a premium for the benefit of using mobile data, but this might be too much of a premium.
While the analysts I spoke to seemed pretty excited about the idea of a device that includes cell data and Wi-Fi, I suspect that the audience for this device will be relatively small. The PDA market in general has been shrinking, and the market for this device will be a subset of the overall PDA market. Still, the analysts I talked to were bullish on future versions of such devices that allow more seamless roaming and voice over Wi-Fi to reach a far larger market.
Tank McNamara tackles the difficult topic of Wi-Fi in ball parks: One of my surprise favorite comic strips (by Jeff Millar and Bill Hinds) puts a good spin on a what-if for ballpark fans with laptops. Mayhem results.
Drazen Pantic used off-the-shelf, inexpensive hardware and software combined with a community Wi-Fi network to broadcast to live television: Citizen videobloggers take note. Pantic describes the system he used (drawn as a simple schematic not much more complicated than the actual installation) to perform a live, public-access television broadcast managed by him and two colleagues. The topic? How they were doing what they were doing, of course.
Pantic's essay walks through the drop in price, increase in quality, and proliferation of open-source tools and patent-free/license-free standards that can allow practically anyone to produce streaming, broadcast quality television or recorded digital video for later editing and airing.
We established a wireless connection through a local, public WiFi network maintained by the non-profit NYC Wireless, and broadcast from that spot to a computer at MNN studios. The video and audio was captured by the camcorder and fed into the laptop, where it was encoded as MPEG4/AAC streams, then sent out as a unicast stream via the WiFi connection. At MNN they played the stream through a scan converter -- which converts the stream on a computer into a video signal -- then broadcast it live on the air.
It's not just a sign of things to come. It's a sign that things have changed.
Cleveland airport gets Wi-Fi from SBC: I should really have left this item to senior editor Nancy Gohring, who hails from within spittin' distance of Cleveland, but SBC FreedomLink is now the operator of Wi-Fi service at that town's airport. The interesting twist in this installation is that there will be Internet kiosks in the airport as well as the SBC service.
I've had mixed feelings lately about posting every single airport announcement, especially as smaller markets have become unwired: at some point, just another airport is just another airport, unless you live or work near it. It's a fundamental problem with covering wireless in general and Wi-Fi in particular as it matures. There's more and more news, and it starts to blur into much of the same. I see sometimes a dozen stories a week about a small town that gets its first Wi-Fi-enabled cafe, or another community project that unwires a park or public area.
This is just to say that the proof of the success of the top-down and bottom-up movements in wireless communication are starting to permeate all media. When I appeared on a local NPR affiliate to talk about Voice over IP (VoIP) and Internet telephony on their morning program recently, the host needed a technical primer in VoIP, which is quite new to most people, but threw around Wi-Fi and wireless terms with great abandon and fluency.
FCC rule allows end-users to change out antennas on their Wi-Fi and other gear legally if the manufacturer has performed the right tests and the antennas conform to certain guidelines: Jim Thompson alerted me to FCC rule 04-165 issued July 12, 2004, which has some substantial changes for devices that use unlicensed spectrum, most significantly Wi-Fi. The rules affect devices that operate under Part 2 and Part 15 rules, and we're most concerned with Part 15, which governs 900 MHz, 2.4 GHz, and several bands in 5 GHz.
The most notable general applicable of this rule change is in section 2, which starts with point 18: "Replacement Antennas for Unlicensed Devices." Until now, the FCC has required that any antenna to be used with a device operating under Part 15 rules had to be tested and certified as part of a system. There was no mix and match proviso. Further, the FCC required unique connectors for each manufacturer, and required new connectors to be designed as the existing ones became commonplace.
"Wait," you may ask--"I can go to HyperLink Technologies or other companies and buy antennas with the right connectors and attach them to my Wi-Fi gateway. If it's illegal, how can I buy this gear?" Simple. It's legal to sell antennas; it's illegal to use them. It's the same logic that guides the sale of bongs and switchblade kits. It's opposite to the logic that underlies the Digital Millennium Copyright Act. The antenna/switchblade law essentially says that the seller isn't responsible for all the uses to which a purchaser may put a product. The purchaser is obliged to know local and federal rules and conform to them.
Thus, attach the antenna, and you're a pirate. Assemble the switchblade, and you're a criminal. Forget that there have been approximately zero prosecutions for the use of these antennas on home or business systems. But no one wants to be in de facto violation of a law, especially businesses that may considering building out Wi-Fi as part of their operations. The lawyers might look askance, and the companies might have to pay enormously higher fees to purchase legal antennas--if they're available. Those fees help cover the companies' cost in certifying the antennas as part of a system, but also represent their lock-in market for legal use.
The FCC rule doesn't suddenly make all antennas legal for all systems. Instead, they have chosen a clever middle ground. For new devices--or, presumably for recertification of old devices--manufacturers will be allowed to test the system with high-gain antennas of each major type, like omni, patch, yagi, and so forth. Once the device is certified, the manufacturer can release the characteristics of the antennas they tested for both their in-band and out-of-band signal patterns and strengths. (Out-of-band transmissions are the inevitable but not intentional frequencies that are broadcast on at typically very low levels due to harmonics and other technical radio issues.)
Thus, if Linksys certifies its WRT54G with a very high-gain yagi antenna within the Part 15 rules, then a user can add a lower-gain yagi that has all its parameters within those levels and be perfectly legal.
Jim Thompson provided a longer, detailed explanation via email:
"Let's say you have a (warning, plug alert!) HS3000 that has been tested with a 9dBi omni. Were you to find an 8dBi omni from a different manufacturer, with similar out of band gain (i.e. it doesn't generate more gain in the restricted bands, thereby causing a system that would otherwise comply with the restricted band limits to 'go illegal'), you could use it.
"You could also attach a 2.2dBi omni, as long as it didn't have more gain out of band than the antennas with which the device is certified. You can repeat the above paragraph substituting 'yagi' or 'patch' or 'grid' antenna everywhere 'omni' occurs.
"What you can't do is certify (let's say) with a single 2.2dBi omni, and then have your customer attach a 13dBi yagi, (without recertification), nor could you say, certify with a 13dBi yagi and have your customer attach a 13dBi patch (or omni)."
I have some suspicion that the recent array of Linksys add-on antennas were certified under this new rule and delayed for release until such point that the rule could go into effect. This rule would dramatically reduce the cost of re-certifying gear for more antennas, and it makes it possible for Linksys to sell a huge matrix of their own antennas at no additional testing cost beyond the initial certification. On the other hand, it also makes it easier for third parties to sell antennas legally for Linksys's devices, but sales of legal antennas for illegal uses has seemed to curtail sales before this.
Remember that until a device is retested under these rules--and who knows if manufacturers will pay to retest the current generation of equipment--you're still technically violating the law by mixing and matching antennas. Watch for more news on this front, as devices are certified until these new rules.
The FCC decided to leave the connector rules intact, even though manufacturers argued that it's so easy to get their "proprietary" connectors from third parties, that the current rules just added cost and complexity. The FCC demurred, noting that it wanted to make it just hard enough to make adding an antenna an intentional act. It gets lost in the furor over unscientific concerns about the risk of Wi-Fi and 2.4 GHz electromagnetic radiation that microwaves can injure humans at sufficiently high gain--far, far higher than the Part 15 rules allow.
The FCC said in regards to the connector issue, "...our concern that removing this requirement might have the unintended consequence of allowing uninformed consumers to inadvertently attach an antenna which causes the device to emit at levels in excess of the limits for human exposure to radio emissions." Good enough. It preserves the market for pigtails, that's for sure.
This section also revises the rules about integral antennas for 5 GHz (802.11a and other uses) devices. The FCC will now allow externally detachable antennas for these devices, which adds flexibility.
The FCC tweaked a number of other rules, including one that appears to reflect a change in thinking from its one-off approval of Vivato's beam-focused Wi-Fi--I'm trying to better understand that section. In another part, they tweak measurement standards. Part 5 covers changing rules for frequency hopping to make it possible for future Bluetooth flavors to work legally in the U.S.
Interestingly, spectrum etiquette in unlicensed bands was discussed in part 6, but the FCC declined to take any action. They gave props to Microsoft for a proposal they might implement in the future that would reduce the noise of devices that aren't actively transmitting. Some of these principles are already embodied in 802.11h, which was a required extension for 802.11a to operate in the 5 GHz band in Europe.
Apple's AirPort Express may quickly become one of the most reviewed pieces of new wireless technology: It will receive many reviews for several reasons, including the fact that it's the smallest Wi-Fi gateway (when you include its built-in power supply); it's the only one to stream audio in the particular way it does; it includes several interesting features in one wrapper; it's relatively cheap for any two of its four unique set features*. It's also from Apple and had 80,000 pre-orders, so it's a natural. (Amazon.com now shows it not first arriving until August 1, and other sources indicate a three-week backorder. But the Apple Store in Seattle says they should have another supply any day now.)
I've been working with an AirPort Express for a few days, and it's just about as easy to setup and use as Apple promises. There are no obscure settings. Joining an existing AirPort Extreme network was a snap. So was reconfiguring it as a base station and assigning it a WPA encryption key. So was playing music through its attached speakers from any copy of iTunes anywhere in our wired/Wi-Fi office. My officemates threatened to play strange music into the speakers in my office, as any copy of iTunes can use any set of AirPort Express speakers on a network unless you password protect access to the speakers.
Three reviews check in today from well-respected sources. David Pogue walks through the pros and cons of the device in The New York Times, and comes down reasonably heavy on the pro side. He misses having a remote control and notes that it's odd you can't play through several sets of speakers at once each connected to their own AirPort Express as you can with other devices. The total of the parts in one well-designed package adds to a winner for him, however. He notes that you'd need two or three other devices to come close to the Express--and in that comparison, he leaves out the Express's client mode (to connect for streaming/printer sharing to any Wi-Fi network) and its USB printer sharing, which is an expensive stand-alone add-on for 802.11g networks.
Walt Mossberg's take was substantially more negative because of a few flaws he felt were significant. He finds the lack of a remote-control a total showstopper, and I admit that that was one of my reactions on first hearing about AirTunes streaming music. Most people who purchase an AirPort Express will have their computer in another room--perhaps far distant--from their home stereo system. A remote control is a no-brainer, and Apple has signaled that something along those lines might be in the works. I imagine a Wi-Fi-enabled iPod which control iTunes on a Mac, play music directly from itself, or carry out other network and stereo functions. Another review pictures something like Salling Clicker combining Bluetooth, a cell phone, and the AirPort Express.
Mossberg also doesn't like the fact that he had to resort to AirPort Admin Utility, a fairly technical configuration tool, instead of the streamlined AirPort Express Assistant to connect his AirPort Express to an existing network that employs what must be WEP encryption. You can enter WEP keys in the assistant, so there must have been a more complicated issue.
Finally, the technology review site Ars Technica offers an in-depth look at the intricacies of the unit. In a long and interesting examination of the AirPort Express, Eric Bangeman is generally positive, giving the device a rank of 8 out of 10. As with Mossberg and Pogue, Bangeman wants a remote control--but he also wants a simple, $5 cable (mini-stereo to RCA) to be bundled with the unit instead of as part of a $40 accessory pack. Bangeman points out that you can't attach AirPort Express to a WPA encrypted network using the wireless bridging feature, which is a good proviso. If you have an Ethernet backbone, however, it's a non-issue, but it does remove one of the interesting selling points.
(*Its unique set of features: It combines USB printer sharing and iTunes music streaming with its small form factor and weight along with the ability to act as a streaming/sharing client on any Wi-Fi network.)
Newbury Networks wardrove through Boston for a few hours and found lots of open access points and promiscuous Wi-Fi cards: They'd like to piggyback some publicity on the upcoming Democratic National Convention by noting that it's easy to find unsecured access point and easy for a Wi-Fi adapter to associate with an unknown point without user intervention.
Still, the real issue is whether DNC operatives and even press will be wandering around outside the Wi-Fi-less convention center with laptops trying to connect to random networks or with their cards trying on their own. It's much more likely that those attending the DNC won't take basic security precautions like using a VPN tunnel or even turning off their wireless card when there's no network they're trying to use.
Still, with this knowledge in hand--what do you do with it? Newbury Networks says in their press release, this potentially sets up a dangerous security scenario based on the level of open Wi-Fi networks in range of the FleetCenter. Sure. And?
Two stories converge on Connexion by Boeing and Airbus's future in-flight data plans: Tenzing is folding itself into a firm jointly owned by itself, Airbus, and SITA, a Dutch-based air services integrator and air-phone operator. Tenzing will keep employees in Seattle, and SITA will be the majority shareholder. Tenzing will have the advantage now of this direct relationship with Airbus, which has begun to build Wi-Fi into planes as a standard option. The new company will also focus on in-flight wireless calls, the third firm to announce its intentions to offer or to test out this capability in the last few weeks. The new firm expects to offer that service in 2006 starting on intra-Europe flights where SITA already offers its more conventional in-flight phone service among other data integration offerings, while their high-speed data offering is still on track starting in 2005.
Meanwhile, Connexion by Boeing signed Siemens as a customer, which means that Siemens employees will have easier, and presumably cheaper, access to the service. Connexion is installed in just one Lufthansa plane so far, which flies from Munich to Los Angeles non-stop, but Lufthansa and several other airlines continue their commitment to build out their long-haul fleets.
Japan might tax the use of unlicensed spectrum for wireless LANs to protect market for licensed spectrum holders: The logic cited is fairly bizarre--that charging for unlicensed use is "fair," which I presume means "fair to companies that paid large amounts of money for cellular telephone spectrum" but not "fair for citizens who own the airwaves and can now not see fit to use them for free." Japan perhaps has a different regulatory framework than the U.S.
Imagine a bill hitting the House of Representatives suddenly that proposed a consumer tax on Wi-Fi? I think the recall petition would be filled with signatures before the bill reached its second reading. The bill in Japan won't reach Parliament until 2005, at which point the legislators involved will probably have been voted out of office, or buried under tens of thousands of letters.
AT&T Wireless announced today--as was leaked last week--that they have 3G wireless available in four U.S. cities: Detroit, Phoenix, San Francisco, and Seattle are up and running at 200 to 300 Kbps right now with San Diego (take that, Qualcomm home town?) and Dallas coming by the end of 2004. Upload speeds aren't listed.
AT&T will charge $25 per month for phone-based 3G service, with streaming video costing an extra $5 per month. Data plans are $80 per month. No word yet on whether the Motorola phone that works with this W-CDMA network will allow the same kind of sidestepping to connect a laptop via Bluetooth or USB to get the lower price. It does have Bluetooth built in, which is a good sign.
Daily Wireless has a round-up of the announcement with details about the phone; they link to this Motorola press release with more phone details. The handset will cost about $300 before rebates or other deals. It has an integrated MP3 player, a large screen, three band (900/1800/1900) support, and handles GSM, GPRS, and W-CDMA. Laptop users will require a Merlin U530 UMTS card.
SBC Communications announced a host of Wi-Fi partnerships and deals today to extend the reach of its network and maximize traffic passing across locations they operate: SBC will resell access to its FreedomLink network, which will include 300 Caribou Coffee stores (Midwest and South-central and Southeast coast) as well as the thousands of The UPS Store locations, to iPass, GoRemote, and Syniverse. This reselling agreement will be live by the end of 2004.
SBC FreedomLink subscribers will have access to Concourse Communications hotspots (JFK, LaGuardia, Newark, Detroit, and Minneapolis/St. Paul), Telmex (400 Mexican locations), and Wise Technologies (an array of retail and airport venues). Access to those locations won't be included in a monthly fee from SBC, but will cost just $4 per day, which is often less than half the cost of the retail prices for those venues. This roaming arrangement will also be in place before year's end.
In the same press release, SBC notes that it is selling 3,000 home Wi-Fi gateways each day to SBC Yahoo! DSL customers. They reiterated that later this year, FreedomLink will be offered to its DSL subscribers at a "significant discount."
SBC is deploying several tactics at once here, increasing the size of its self-operated network to make the total number of hotspots more appealing to subscribers while also making the network more appealing to resellers. Providing discounted access to other hotspot network is useful for business travelers who may have no compunction about paying $4 a few times a month for airport access.
By emphasizing the scale of home Wi-Fi adoption and its coming discounted plan for its millions of DSL subscribers, SBC also makes the network more appealing to new venues considering which hotspot operator to work with.
It's not quite the safe as Wi-Fi in your iPod, but Bay Area Free Wi-Fi Wireless Hotspots has a downloadable listings of San Francisco Bay locations: wiPod is a little database that you can download to an iPod to be able to scroll through and see free locations throughout that region. The site looks new: just a single post from July 12. Nonetheless, regional guides to free access should help windshield warriors and neighborhood officeless workers.
The Wall Street Journal reports on research showing hands-free cell phone talking while driving is unsafe at any speed: This is a purely public service announcement I make today: if you'd like to dramatically reduce your odds of finding yourself in an accident, don't talk on a cell phone while driving regardless of whether you're holding it or using a hands-free system.
The cellular industry doesn't like it, sure, but the growing body of research reviewed in this article makes it clear that talking on a cell phone while driving has a significantly different impact on your ability to stay focused on the road than any other activity, including listening to the radio or talking to passengers. As the Car Talk guys bumper sticker says (see upper right), Drive Now, Talk Later!
Studies to the contrary turn out not to be to the contrary, and have had the usual PR spin put on them. You know why we have anti-bacterial soap all across the U.S.? Because of a single study done on the cleanliness of street vendors in Mexico. It all traces back to that. A public-health crisis, producing more and more resistant strains of bacteria, results from companies choosing to use limited information to promote a message instead of performing more science to determine appropriate ways to market their products.
The same is true here: cell companies could promote safer uses of phones, but their money is in pushing minutes (at the moment). When cell operators finally switch to unlimited monthly plans, they'll want people to talk less, just as AOL did when they switched from hourly billing to unmetered monthly service. When that switch happens, you can bet we'll see a 100-percent full-court press on restricting talking at all while driving.
As a regular bike commuter, I am often in a position to see cars clearly perform precision bad driving maneuvers. About 95 percent of the time, the driver is talking on a cell phone. When I see someone driving erratically -- bad lane position, no signal, strange turns, sudden acceleration -- I start biking even more defensively. When I pass them or them me, I see them talking on a cell phone.
This is anecdotal on my part, and I'm glad the science backs me up. It also tells me that when I see people driving poorly and they look like they're not on a cell phone, they probably are--using a hands-free system.
Seattle-area firm focuses on towns under 50,000 to bring data at affordable price: By partnering with local utilities in these smaller towns, Maverick can reduce its capital risk and focus entirely on service. The company charges $25 to $60 per month for rates from about 128 Kpbs to 1 Mbps. They may be in 23 markets by fall 2005, including their current deployments being built out in Kennewick, Silverdale, and Poulsbo.
Although this brief company overview doesn't mention it, an important application for having these kinds of robust Internet connections is for Internet telephony. If you're in a rural area, "local calling" is quite local. A VoIP line from Vonage or Packet8 could dramatically decrease the charges for calling within the state as well as for long distance.
This press release sounds obscure, but it's not: you can't call yourself Wi-Fi if your proprietary add-on interferes with Wi-Fi receivers: The Wi-Fi Alliance has taken action on the alleged interference caused by Atheros Super G Turbo mode. Broadcom has stated on many occasions and presented test results that they claim demonstrate that the Turbo mode (which binds two radio channels to produce a gross throughput of 108 Mbps an a net throughput in the mid-30 Mbps) can dramatically drop the speed of nearby Wi-Fi networks running 802.11b or 802.11g flavors.
Atheros has consistently denied that those tests reflect reality: they have said their own tests fail to demonstrate this problem. Testing by Tim Higgins of Tom's Networking showed most clearly that Broadcom's equipment was most heavily affected by Atheros's Turbo mode, but only under limited, close-range circumstances. Anecdotally, I've heard more and more stories from home users finding their networks performing erratically, but I have been unable to track down in those cases whether new nearby networks--in apartment buildings or residential neighborhoods--were the culprit, whether Atheros-based or otherwise.
The Wi-Fi Alliance had promised months ago to take up this issue, which is a touchy one, given that Atheros and Broadcom are both members, and that most chipmakers and consumer manufacturers are deploying a range of proprietary and/or pre-standards extensions to Wi-Fi.
Starting today, if any extension impairs other Wi-Fi devices, the alliance can use its trademark control to pull its branding. The alliance's director, Frank Hanzlik, says in this press release, "If a product extension significantly impacts the ability of other Wi-Fi CERTIFIED equipment to operate as intended, the Alliance may withhold or revoke certification."
Interestingly, more devices are heading to market without the Wi-Fi stamp, and with this latest tool, it might mean that Wi-Fi becomes Balkanized--Atheros could opt out and potentially see no problems with that decision. Some of its OEM partners already ship Wi-Fi-trademark-free equipment.
The plan costs £70 for unlimited cell and Wi-Fi, with the first three months free: T-Mobile says the 3G side will run just 128 Kbps right now, but eventually ramp up to true 3G speeds of 384 Kbps. The press release talks about a 3G card, and doesn't specifically address whether Wi-Fi is incorporate. The card comes as part of the £199 setup price, and will seamlessly switch between 2.5G (GPRS) and 3G. Seamless switching to Wi-Fi isn't mentioned. T-Mobile UK has 500 locations in its network, while T-Mobile as a whole has 7,000, which must be counting the 4,000-plus in the U.S. It's not mentioned, but roaming to non-T-Mobile UK locations must not be included in the unlimited pricing.
Gateway offers $299 7001 802.11g access point free with $399 server: T. Bryce Yehl wrote in to note that Gateway has a remarkable rebate deal on their superb access points series that I reviewed just a few days ago. If you purchase a server for as little as $399 (before tax and shipping and not including a separate $50 rebate), they throw in the 802.11g AP for free or apply $299 towards their a/g access point, which is $399 retail.
To buy their least expensive server, go to Gateway's home page, click Small and Mid-Sized Businesses, click Products, click Servers, click the 920 Servers. Click Details, and then choose Customize. Remove the operating system, extra RAM, and support package. You can buy your memory elsewhere: this unit only comes with 128 Mb in the $399 configuration. Remember after purchasing to send in the rebate form to get the additional $50 back.
If you were thinking about the Gateway 7001 AP anyway, this is a great way to get a computer for $100 more.
Qualcomm lofts media, officials to show that plane stays in air while cell phones used: Qualcomm would certainly enjoy having more people using more advanced phones more of the time, and their proof-of-concept with a commercial American Airlines aircraft--but obviously not a commercial flight--out of Dallas/Ft. Worth on Thursday was intended to gather data and allay fears.
The fear of cell phones used in the air is certainly a bogeyman: there are a few known, unexplained instances of a cell phone in use that may have affected some airplane equipment, but there's no widely known, widely reproduced instances of cell phones affecting systems. The FAA is following a better safe than sorry approach. A Boeing engineering magazine article from 2000 has a fascinating article that doesn't entirely disprove this, but shows how hard it is to reproduce problems even with the equipment that was said to cause the problem in the first place. It's a good read, as it's sensible and conservative without being one-sided. Boeing has interests on both sides of this regulatory and spectrum fence, to be sure.
The FCC bans cell phones in flight because an airborne cell phone can wind up using the same frequency on dozens of terrestrial base stations at the same time, which has the potential to wreak some havoc on the ground by tying up spectrum and abusing handoff.
Qualcomm was testing a picocell that was onboard the aircraft: that cell captures all of the on-board cellular communications and relays it to a special ground station. It's like a cellular network proxy instead of just a shunt to the cell network. This is the third of these kinds of tests from different firms that I'm aware of from just the last few weeks. The friendly skies will be even more full of chatter soon. While the picocell was 3G, according to the press release, only voice and text messages were used over the system--no pure data. An American Airlines EVP is quoted in the article saying that in-air cell use is about two years away, which conforms with what's being said elsewhere.
Industrial-grade earplugs or stereo earpieces--which can be custom-molded for your canals--will probably be a good growth industry for the flying market. Expect to see in-flight magazines half full of advertisements on reducing irritation of adjacent loud talkers.
If you're curious about what happened at the IEEE 802 meeting this last week, Daily Wireless rounds it up: From continuing lack of agreement on 802.15.3a (despite the withdrawal of everyone but Motorola, we thought) to new groups being formed and new 802.11 subletters, the IEEE moved mountains of paperwork and specifications around. Daily Wireless provides a cogent summary with links for more information about each subject.
Sprint PCS quietly sloughs off its own building plans?: This article notes that Sprint PCS said it would have access via 2,100 access points within a year. True. But they told me that 1,300 of those would be locations that they built out themselves. In this Wi-Fi Planet article linked above, the general manager of wireless LAN service for Sprint PCS -- not the person in this role last summer -- said, We are not deploying those ourselves. But that's not entirely accurate: they're building Truckstop.net locations via their managed services division, and then Truckstop.net is operating those locations and allowing Sprint PCS to roam on them. So they're not deploying, but they are building and supporting.
InterLink wants every office to have affordable wireless security: InterLink introduced LucidLink Wireless Security several weeks ago to connect smaller offices with enterprise-grade wireless encryption. The company knows how to build RADIUS and other authentication servers, and it saw the large gap between pre-shared secrets--subject to social engineering and potential cracking--and servers that require IT expertise and cost tens of thousands of dollars.
In an interview with InterLink's president and CEO Mike Klein and vice president of marketing Wayne Burkan, they laid out how this new offering might appeal to the hundreds of thousands of businesses without information technology departments but who would benefit from a secure, login-based wireless network.
LucidLink is a small RADIUS server that has a simple front end, less complicated than, say, setting up calendar entries in Outlook. It requires a special client to be installed on computers. The client requires a reboot after installation--in the current version--and that's the most onerous task in gaining access to the network. An administrator uses a simple program to approve users as they request access via the client software. It's an option-free package that meets a high bar of wireless network security through unique WPA master session keys per user.
Klein explained that InterLink's primary business remains on the high-end. One ISP they work with, for instance, has 6.5 million users that authenticates using 13 AAA (authenticiation, authorization, accounting) servers from InterLink. The firm works with AT&T, Swisscom, and other giants, and has their software incorporated by firms like HP and Siemens and repackaged by Infobox into hardware appliances.
"Security itself is fairly complex today in the enterprise marketplace," Klein said. The idea behind LucidLink is that it "brings a consumer-oriented ease of use model." He likened it to a garage-door opener, which masks a fair amount of wireless technology with a single button.
"We're bringing enterprise-level security down into the small-to-medium business market" with the ease of use experience required by the small business, Klein said, calling LucidLink "RADIUS for the rest of us." LucidLink takes the management of this network away from an IT manager and puts it "on the desktop of an administrator, like an office manager or a receptionist," or the same person who handles cell phone numbers, access codes and badges, or building access.
Klein acknowledges that WPA (Wi-Fi Protected Access) provides a higher bar of security, but says that pre-shared keys aren't useful for small businesses because they don't reflect the realities of temporary guests or workers or people leaving their jobs or being fired. To reconfigure the key, every user has to be off the network briefly and simultaneously, and then enter the new key to get back on. The coordination is a problem for even smaller networks.
Even if you have an 802.1X system with WPA running, Klein notes that Windows XP has 60 separate combinations of options if you dig through all of the associated checkboxes and dialog boxes for 802.1X connections.
The server side isn't any better with Windows Server 2003. InterLink brought in a customer and had them get a RADIUS server with 802.1X authentication up and running. It required 177 decisions and 12 hours to configure a single user.
Klein said the company charged its engineers with a mission: it's all about suportability. "Don't give the user all the options that come in a standard RADIUS server," he said.
To use LucidLink, it does require the installation of their client, which is used to create a certificate-based EAP session using the Diffie-Hellman key exchange. The client handles the exchange of credentials for trustworthiness. The server side of LucidLink allows an administrator to confirm the client by checking a cryptographic fingerprint before they authorize them in the system.
Klein said that the biggest problem they've faced after releasing the software is that access points can be difficult to configure correctly for 802.1X passthrough. They have profiles for several access points in their initial release, and will add more common ones in the future. An administrator could then just plug in an access point, point LucidLink's server package at it, and the AP would be configured correctly with no additional work needed.
Burkan ran me through an online demo of the system to show how simple it was. The server side runs under Windows XP or 2000, while the client works just on Windows XP but they'll expand that over time. The server application requires direct access to the machine that's running it, or can be accessed via Windows Remote Access. Burkan said, "When we talk about [the] server, it's the same machine that the office manager is using to do email and Word." The computation requirements are extremely low.
After installing and configuring the server package, each client for the network installs the LucidLink package. After rebooting, the client software can be run and try to connect to the network. As each client connects, the administrator sees these attempts in the server application, and then approves them after confirming their fingerprints with the connecting user. (This removes rogue user attempts even if they can get the LucidLink client software installed.)
The user control is granular, allowing unlimited use or timed use. Users can be suspended, removed, or re-enabled. A regular visitor can have their access turned on just while they're known to be in the building for instance, but disabled at other times.
InterLink sells its software via value-added retailers, but it's also talking with Wi-Fi gateway makers. Klein said, "We bring a very happy medium and allow both ends of the market to deploy their access solutions into the middle of the market." The software costs $449 for 10 users, $895 for 25 users, and $1,595 for 50 users direct from InterLink, with support and upgrade plans costs $150, $295, and $495, respectively. The software allows unlimited registered users; the license limits active users only.
It's true that the next big leap in small business routers will be to jump from WPA-PSK, a shared key with associated risks and complexity, to secured 802.1X with WPA--a solution that InterLink seems poised to bring to a broader market.
Swiss Federal Railways is adding Wi-Fi to seven stations initially; more to follow, along with trains: Starting later this year, another 26 stations will have Internet access, too. By fall 2005, a number of east-west trains will have en route service added.
Reuters reports that AT&T Wireless is poised to roll out its W-CDMA 3G cellular data offering in Detroit, Phoenix, San Francisco, and Seattle next week: AT&T Wireless had been saying "by the end of the year," and this leak of a near-term launch is apparently surprising. The company is only promising 200 to 300 Kbps, but it's likely that W-CDMA will offer much higher burst rates, slightly slower than 1x-EvDO, which Verizon Wireless plans to spend $1 billion in the next several months to expand throughout major markets in the U.S.
I expected a rollout in Seattle because AT&T Wireless is headquartered in this region--for the time being. The merger with Cingular will certainly result in an almost total shutdown in this area as there's no logical, only historical reasons for this geographical location. The service will require one of two phones from Nokia and Motorola to access.
As TechDirt notes, AT&T Wireless has to launch service in four cities before the end of the year to avoid a requirement to repay $6 billion to investor DoCoMo. Elsewhere, I've seen this figure and deadline rejected by certain executives involved in the transaction, but I don't believe there's been a definitive refutation. [link via TechDirt]
Gateway's 7000 service access points include all the security and authentication that an office of 5 to 25 users need: Gateway released its two 7000 series models a few months ago and have received very little press. The unit comes in 802.11g ($299) and 802.11a/g ($399) configurations. It sports support for all the popular standards, including 802.1X passthrough, WPA-PSK, WDS (Wireless Distribution System), and others.
It even includes two separate wide-area-networking (WAN) interfaces: if you have a separately segmented guest wired network, you can physically connect the device through the guest interface to your public WAN and the private interface to your private WAN. The device can segregate all wireless traffic. It can also do this using Virtual LAN (VLAN) tagging by providing multiple SSIDs that are logically separate for guest and authorized user: the traffic is routed by your VLAN switch or system.
But the real winning feature in this unit is its built-in RADIUS server that supports both WEP and WPA over 802.1X using PEAP (Protected EAP). Because Windows XP and Mac OS X 10.3 both include PEAP, and it's possible to use PEAP affordably on other platforms including Linux and Solaris, PEAP is currently the broadest standard to support. Effectively, using a $299 or $399 device, you can have enterprise-level security without a separate RADIUS server or the administrative overhead. But it only supports a handful of users.
I was interested immediately in this built-in server aspect, and received a review unit some time ago, but was unable to get past some initial confusion in configuration. I started from scratch today, using the factory defaults reset button, and had no problems whatsoever with the same instructions, software, and firmware, so let's attribute my earlier issues to user error.
The device is extremely easy to configure. I followed the directions and powered up the unit, plugging my wired LAN into the LAN1 port. I ran the auto-configuration software that finds the access point. It assumes that you have a DHCP server running and this software handles the discovery of the 7000 series IP address. You click a link in the software to open a browser and enter the default passwords to connect.
Once connected, there's very little to configure for an ordinary network. I went to the User Management section and created an account for myself. I then clicked on the Security tab under the Advanced section and selected WPA with RADIUS from the Security Mode menu. By default, the settings are for TKIP as a cipher suite and Built-in as the Authentication Server. However, you can set it to AES (CCMP) and an external RADIUS server here as well. (The IEEE 802.1X option in security mode is really 802.1X plus WEP instead of 802.1X plus WPA.)
For a lower-bar, the device can handle non-WPA 802.1X clients if you check a box. This allows WEP to be used for those clients, but this compromises the security of that session, of course.
Once I clicked update, I switched to my Mac running OS X 10.3.4, which includes robust EAP type support. I ran Internet Connect, selected a new 802.1X connection, confirmed in settings that it would only accept PEAP, EAP-TTLS, and LEAP, and click Connect. I was asked to accept a certificate from Instant802 Networks. I did, and the device immediately authenticated via PEAP.
I switched to Windows XP, expecting a potentially more complicated process--I wasn't disappointed. I chose View Available Wireless Networks from the System Tray's icon for my wireless adapter. When I selected my Gateway network in the menu, it grayed out the Network Key options and left Enable IEEE 802.1X Authentication for This Network checked. I clicked Connect. Nothing happened.
I chose View Available Wireless Networks again, and click the Advanced button. I selected the Gateway network from the lower half of the screen under Preferred Networks and clicked Properties. I was now able to click the Authentication tab in the Properties dialog box and choose Protected EAP (PEAP) from the EAP type menu. If my Gateway user name and password had been the same as my Windows logon, I could have avoided a later step.
I clicked the Properties button beneath the EAP type menu. At the top of the screen, I uncheck Validate Server Certificate -- this step is critical because the certificate used in the PEAP transaction is unsigned as you are out of band when you make the connection.
At the bottom of the Protected EAP Properties dialog box, I selected the Configure button next to the Select Authentication Method popup menu (which was preset to the correct value: EAP-MSCHAP v2, which Gateway's devices support). I unchecked Automatically Use My Windows Logon Name and Password in the dialog box that appears, and click OK, OK, OK, OK. (A colleague calls this the Joe Pesci maneuver.)
After a few moments, a popup balloon appeared above the wireless icon in the System Tray reading, "Click here to select a certificate or other credentials for connection to the network Gateway Greenwood." I clicked and was presented with an Enter Credentials dialog box. I entered my user name and password, but no domain as my Windows network does not have one.
The first time through, I had difficulties because I'd failed to uncheck Validate Server Certificate. After some excellent feedback from the folks at Instant802 Networks, I was able to connect and authentication jst fine.
I highly recommend the Gateway 7000 series. Any small office hoping to avoid security woes and wanting to bypass the issues of using a single, preshared WPA key can spend a very small amount of money for robust piece of mind.
AirMagnet is introducing a new version of its portable analysis products, designed for laptop or PDA users: The upgrade is free to existing customers. The new version extends the new look and functionality of AirMagnet's distributed product that it released earlier in the year. The tool includes 19 new alarms, including one that detects a denial of service attack recently identified by researchers in Queensland. That attack made headlines when it was first uncovered but AirMagnet hasn't seen or heard of any such attacks in the market. Still, customers asked for protection against it. "One reason we hopped on this so quickly was that the press picked up on this aggressively and our customers called to say they'd like an alarm for it," said Rich Mironov, vice president of marketing for AirMagnet.
The new product also supports various authentication and encryption schemes. The software appears to the network like a client that can be running WEP, Cisco's LEAP, Microsoft's PEAP, TLS, or Cisco's latest, EAP Fast. Mironov said it's important for network administrators to have a tool that can support the method their network uses because often other tools will shut off security to do the diagnosis but problems could stem from the security mechanisms themselves.
The tool also offers some network monitoring capabilities for customers who may be running voice over WLAN. It depicts the signal strength of each packet, so a flat line indicates that a call will be delivering good voice quality but a jittery line indicates interference. Other tools can then help administrators figure out where the interference is coming from to correct the problem. The tool can also monitor how long handoff takes between cells, which will indicate if calls are being dropped as users move. Additional security diagnosis capabilities specifically for voice over WLAN will be available in future upgrades, Mironov said.
Apple ships its streaming audio, printer sharing, portable base station: Apple announced today that it is shipping its $129 AirPort Express, a 7-ounce base station that has no external power cord in its default configuration, and which combines a single Ethernet port, an audio output jack for streaming music from iTunes (Mac and Windows), and a USB port for sharing printers across Mac OS X, Windows XP, and Windows 2000. This is Apple's most Windows-friendly hardware, ever.
The company said it has 80,000 pre-orders for the unit, which is listed as shipping in 3 to 4 weeks from the Apple Store online. Amazon.com says their stock will be available July 20.
TheFeature.com discusses how DoCoMo's 3G/WLAN phone protects cell voice business model: Carlo Longino writes that DoCoMo's N900iL handset works on 3G cell networks and office Wi-Fi networks, but that the company has somehow disabled it from working in hotspots. Is it disabled, or can it just not handle the authentication? Gateway pages aren't good for handheld devices, especially cell phones, but this is a great opportunity for 802.1X to enter the scene and handle the authentication.
Longino expects that DoCoMo, T-Mobile, and other carriers won't make it easy to use voice over just any old Wi-Fi network because it could erode voice revenue.
Richard and Angela Hoy file their second weekly report on their several thousand mile trip around the US, and the hunt for Wi-Fi: This week finds our intrepid explorers, my friends, in Ohio and Michigan. They tried the city-wide Wi-Fi experiment in Grand Haven, Mich., which worked terrifically for them wherever they tried it. They're on their way to grandmother's house -- no, they're not kidding -- and stopping wherever they can find access en route. The picture at right is priceless: at Panera, the whole family got into the action.
Wi-LAN continues to stress that WiMax will encroach on patents that it asserts cover OFDM, other basic wireless technologies -- but offers RAND: Wi-LAN is trying to make it easier for companies that it expects will violate its patents, if upheld in lawsuits, that it states cover a variety of wireless standards, including implementations of 802.16, like WiMax. They're offering assurances to the IEEE that they will provide what is known as RAND -- reasonable and non-discriminatory -- licensing. This means that they have to charge a reasonable price, defined by decades of discussion over what reasonable means, and that they have to offer it to everyone, hence non-discriminatory.
Meru repackaged its software to allow users to easily configure their networks for a variety of situations: The software allows IT managers to select different configurations based on the type of use of the network, which consists of access points, a controller, and embedded software. "If you want to deploy voice, then you press optimize for voice so the controller will determine the best way for devices to connect and send packets and handoff is seamless," said Sarah Kim, senior marketing manager for Meru. "It's all the pieces of our technology that already existed but it's packaged in a way that the IT manager doesn't have to think about which set of parameters to turn on to get the best setting." Users can also choose to configure their networks for the best coverage and can opt for different configurations based on separate VLANs.
Kim says that Meru's approach to network optimization is unique. WLAN switch providers tend to rely on power adjustment to optimize network performance, adjusting each access point's power output to achieve the best coverage. But Meru networks, which operate on a single channel, optimize on a per packet basis.
WestCoast Hotels, Red Lion says come on in and use the Wi-Fi: In an interesting amenity, the WestCoast Hotel organization, which operates 70 WestCoast and Red Lion properties, wants its customers to use Wi-Fi even when they're not guests. If you're a member of their free Net4Guests program -- sign up online -- the press release encourages you to use their network whenever you want. The service is free. Guests who aren't members can still get free access when staying in the hotel. The Wi-Fi service is available in the majority of the properties; about a quarter have just high-speed wired access.
Parks Associates senior analyst Michael Cai's recent report on fixed broadband wireless technology may offer a more realistic view of the future than that painted by some vendors: He studied developed and developing countries around the globe and while he found that each region is on a slightly different path, he doesn't expect to see volume commercial deployments of WiMax until 2006. Those deployments will be mainly in Europe and Asia.
That timeframe is slightly behind the second half of 2005 timeframe that the WiMax Forum and some vendors are hoping for. While the standardization process is on schedule, the processes for certification and interoperability are likely to slow down progress. "Who knows where conflicts emerge as they go down the process," Cai said. Even if commercial products come out in the second half of 2005 on schedule, carriers will likely want to test products for three to six months before rolling out a commercial network, he said.
Cai doesn't expect large scale demand for WiMax networks in the United States until 2008 or 2009, though he cautions that there are so many uncertainties here that it's difficult to predict. The deployment of WiMax here may depend on changes the FCC may make to its spectrum policy for the 2.5 Ghz bands.
But in the meantime, Cai expects existing wireless ISPs to migrate to WiMax using the unlicensed bands. "All they care about is cost," he said. Such operators don't have the resources to test equipment or support trials so they're interested in standards-based products that essentially guarantee good performance.
Cai also has some interesting theories on which areas of the world will have the most WiMax users in the near future. While many observers suggest that the developing nations that don't yet have strong telecom infrastructure will make up the largest markets for WiMax, Cai says that developed markets will at least initially account for the highest number of subscribers. "Even if [WiMax reaches] 20 percent of the underserved market in the U.S., that's way larger than the total market in a lot of developing markets," he said. He believes that over the next few years, most of the developing countries will continue to lack demand, lack PC penetration, and lack the disposable income to support WiMax. "At least until 2009, we'll probably have more WiMax subscribers in the underserved markets in the developed countries compared to emerging markets in developing countries," he said.
A talk with AirWave's COO: In a bit of coincidental timing, AirWave's $7 million funding announcement today dovetails with a recent interview I had with Greg Murphy, the founder of AirWave's two incarnations and currently the chief operations officer.
Those with reasonably long memories will recall that AirWave started as a hotspot company and was setting up restaurants in the San Francisco Bay Area in 2001. Faced with the dotcom downturn, the company sold its Wi-Fi locations to a startup operator -- in turn acquired by Ikano and operated now as Hotspotzz -- but turned its expertise in managing remote access points into its main business.
AirWave is on the verge of releasing its third version of the AirWave Management Platform, which allows management through a central software server of dozens to thousands of heterogeneous access points (APs) from vendors like Cisco, 3Com, Avaya, Proxim, and others. AirWave's basic philosophy is that companies can be free to choose whatever access points they need, but also that most companies already have APs from multiple vendors that they aren't interested in replacing.
With a central console, the network can be monitored for new APs, which can be automatically configured; traffic statistics can be collected in one location for an entire worldwide network; reconfiguring 10 or 1,000 routers involves the same amount of work; and rogue access points can be detected when they're plugged into a network. Similar hardware and software is available from Cisco, Proxim, and others to manage just their own equipment with some support for outside integration with other APs.
I spoke to Murphy last month just after AirWave had announced that their system was being used in 10 major universities around the U.S. Murphy noted his alma mater, Amherst, had put in Ethernet everywhere as part of an early wave of making Internet access available to all students. The current wave is, of course, Wi-Fi.
"It's so much more affordable than punching a hole in every dorm," Murphy said. Every time a college opens a wall, they might find construction problems, code issues, or even asbestos. "If I don't open the wall, I don't have to know what's in there," Murphy said.
The AirWave system works well in colleges because they're among the most likely candidates to have patchworks of equipment from many companies, including commodity consumer gear, and then overlay a more comprehensive management approach on top of that existing, already-paid-for infrastructure.
In general, Murphy said, "The larger the networks grow, and the longer they're in place, the more likely they are to be heterogeneous." He said that in most environments, the reality isn't that a company has just 600 Cisco APs or 250 Symbol APs. More commonly, it might be 300 of one and 200 of another and then 10 different vendors' APs of which a company has 3 or 4 of each, often in remote offices. "Who knows who bought them," Murphy said, but they're in active use.
Standardization hasn't helped management any because the standards don't define management control across AP platforms. Even standards that are found in every product don't span the entire product feature set. "The more you standardize the more you run the risk of your product becoming a commodity," Murphy noted. He said that "everyone supports the standards," but they also add their own proprietary features.
Murphy said that the latest AirWave platform release has improved support for what the company terms a generic AP, like a consumer device that hasn't been profiled in their system and which might have limited configuration options. The platform supports the basics, like channel selection, transmission power, and security settings. The system is designed to allow system integrators to build in support for APs that AirWave might never have seen.
This support for generic APs is particularly important in an area that many companies have told AirWave is a problem: remote offices that would require an expensive service call to swap equipment or configure. The office might have a $250 Circuit City-purchased AP, but replacing it with a centrally managed homogeneous device could cost hundreds or even thousands.
Even with more wireless LAN (WLAN) devices starting to find their way into corporations, Murphy is confident AirWave has a place because the WLAN installations tend to complement existing infrastructure. Over time, AirWave will offer WLAN switch management features as well as AP management.
Murphy has noticed that WLAN networks are growing enormously in size. For instance, the Fairfax County Public Schools system is using AirWave to manage thousands of APs across more than 250 schools, and they continue to add more devices. This is not an atypical installation.
In a few years, Murphy said, all large organizations will have at least a few thousand APs to handle data, voice, and other functions. Radio frequency (RF) management will become as important as network management with the density of devices Murphy predicts. The 3.0 platform includes client software so that certain devices on the network can also read RF information and report it back. "The more ears out there, the better," Murphy said. The company has partnered with network monitoring and analysis firm AirMagnet to complement their management system.
The 3.0 release of the software can alert network managers if, for instance, a nearby access point is on an overlapping channels. In later micro-release of version 3, more automation will likely be introduced to allow the management system to make choices. "The goal is to enable the organization to determine which of those variables they want to manage automatically," Murphy said. Some companies will want alerts so they can make the decisions; others will want the software to follow rules and make changes without intervention. (The Cisco Wireless LAN Solution Engine or WLSE has extensive automated RF behavior for its IOS-based and -upgraded APs.)
The current AirWave platform is built around a few steps. A $5,000 base license supports 25 APs, and is designed for pilot projects. One of their clients uses this license to support their Network+Interop conference setup.
A professional edition includes a single-server license with no limits on APs for $30,000. Murphy said the number of APs it can handle is based entirely on the hardware limits, but that thousands could be managed from a single location. The more locations being managed, the less real-time the statistics; for very large installations, the company sells multiple-server licenses that can allow segments of a network to be monitored more effectively.
EarthLink will build Boingo's Wi-Fi software into its TotalAccess package: TotalAccess has been EarthLink's key to user happiness from its introduction about a decade ago. Originally, it handled the nasty bits of finding the latest phone numbers and then dialing and connecting. Now, it's broadened to include anti-spam, anti-popup, and anti-scam software--and Wi-Fi. EarthLink's founder and chairman, Sky Dayton, also founded and runs Boingo Wireless, and the two companies have had a partnership for some time. The integration of Boingo's Wi-Fi management software into TotalAccess makes it that much easier for EarthLink's five million subscribers to choose to use Wi-Fi.
Boingo, by the way, used to note over 6,000 access points under contract with about 3,300 live; in this press release they say 9,000 under contract. We're on the verge of seeing a large bump in network sizes from aggregators to judge by this number and details provided recently by GoRemote and iPass.
As the IEEE considers 802.11n proposals this week, Agere says 500 Mbps should be speed goal: Agere is combining MIMO (multiple in, multiple out antennas) with single and double-wide channel allotments of 20 and 40 MHz. With 40 MHz of spectrum and eight antennas (four transmit, four receive), they expect to reach 500 Mbps of raw throughput. In 5 GHz, Agere says, they could have 11 double-wide channels, while two are possible side-by-side in 2.4 GHz. Agere's proposal appears entirely backwards compatible with the single-wide, 802.11b/g channel and 2.4 GHz compatibility as a baseline.
Continental bucks the for-fee airport/airline trend and offers free Wi-Fi in all of its U.S. Presidents Clubs: Continental is almost certainly taking advantage of the recent FCC clarification that allows the unfettered use of unlicensed spectrum without a landlord's oversight. Delta, American, and United all use T-Mobile's HotSpot service in their club lounges. Chicago is Continental's one exception: it shares a lounge with Northwest.
Eliyahu Ben-Haim wrote in from Israel to note, Vienna Airport's Austrian and Lufthansa (Star Alliance) Lounges have been WiFi for about a year and remain free. SAS has Wi-Fi in all of its international and domestic lounges, but charges for access via TeliaSonera's service.
Intel Research Seattle wants you: Intel Research would like Seattle residents of 18 years of age or older who use Wi-Fi to fill out a survey. Responses will be kept anonymous.
Sascha Meinrath of CUWiN offers his follow-up on previous posts about mesh networking's scalability and utility: Continuing a conversation that began back here, and continued here, open-source and world-wide community mesh networking developer Sascha Meinrath replies and elaborates on those posts.
Chari is right on the mark with his clarifications on network performance degradation rates. The case I had made purposefully oversimplified the throughput degeneration rate. However, in real-world deployments, the actual throughput of a network probably degrades at somewhere between 1/n and (1/2)^n -- where n is the number of hops. Think of these two equations as two limits of the probable degradation rate; as anyone graphing these functions can see, they map an increasingly wide area of probable degradation rates as the number of hops increases -- representing an increasingly large "unknown". The point is that exact throughput degradation rates are fairly impossible to pin down because the variables that need to be taken into account differ by locale. As anyone who has done numerous real-world implementations will attest, bizarre confluences of factors can sometimes cause unanticipated outcomes and disruptions.
One of the major problems facing wireless deployers is that almost all research has been conducted either via computer simulations or in "in-vivo" deployments that are highly contrived (often within science buildings or even within single laboratories). This research provides extremely useful guidelines for anticipating problems; but often fails to capture the complexity of deployments in the community. A closer-to-life example of "real-world" usage is MIT's roofnet project, whose deployment is being used to help proof the ETX route prioritization metric that is being integrated into CUWiN's software. However, this network is utilized mainly by computer science students, who are not exactly representative of the population at-large.
Nitin Vaidya's work has made tremendous strides in our understanding of ad-hoc and multi-hop networks (which Chari does well to point out); but what is really needed is a truly community-based network (with all the attendant messiness) that can be utilized to explore the real-world limits of wireless networks. It is with this goal in mind that Nitin, David Young (CUWiN's technical lead), and I co-wrote an NSF grant proposal entitled, "Engineering Community Wireless Networks" earlier this year. For companies and entrepreneurs working on wireless networking solutions, the possibility of gaining real-world data is extremely valuable. Likewise, for those of us working on Community Wireless Networking solutions, these data will provide an opportunity for better understanding the constraints for deploying robust networks and create more precise parameters for degradation rates.
Chari was also right on the mark in pointing out that scalability and performance are not necessarily directly linked:
"Scalability speaks to how large of a network you can build. Scale is unrelated to the number of hops. A mesh network with a small number of nodes but few wired backhaul points and/or an in-line topographic layout may have a large number of hops. Conversely, a mesh network with a large number of nodes but many wired backhaul points and/or a lattice-style topographic layout may have a small number of hops throughout."
However, I would argue that they are still significantly correlated. In either case, an ideal networking system would be able to handle any of the topographies Chari alludes to as well as allow for multihoming (the use of bandwidth from multiple internet connection points for a single download or upload). Multihoming, however, brings us back to the same problem of scalability and hops being highly correlated. Most importantly, as Chari states, "The real limit to scalability in most mesh networks is routing overhead." Both TBRPF and OLSR protocols share the problem of not scaling to extremely large networks. They're built with a cell-phone, tower-based mesh topography in mind (one need look no further than the cute OLSR MPR flooding demo to see this). A-HSLS will scale to thousands of nodes arranged in a truly non-hierarchical fashion -- it's the difference between two protocols that are most useful by major telecoms, and one that is useful for community wireless networks. A-HSLS is more useful for Community Wireless Networking purposes, while TBRPF and OLSR are more useful for major telecoms.
As regards the 500mW limit I propose as a proactive solution in my manuscript, it is important to remember that one can today legally transmit at up to 1W; and with a exemption (which is pretty much a rubber-stamp process) and an amplifier, one can go up to 10W. The problem is not what power level one can transmit at with today's wireless card technology; but what will be rolled out in the future. I'm especially concerned about the WiMax technologies being proposed that allow for higher transmit powers within the same frequencies as today's Wi-Fi systems -- see this link.
The take-home message is simply that there are multiple uncertainties within wireless technologies -- and it is not that these unknowns should be viewed as a barrier -- but that there are very few universally correct answers. Whether one is looking at throughput degradation or "the best" routing protocol, wireless is a nascent technology with an incredibly diverse set of possible implementations. It is impossible for those of us working with wireless technologies to fully understand all of the different available options, much less have answers to all the questions that are asked of us. But in debating the pros and cons of different solutions, I am hopeful that we'll increase our collective understanding of wireless technologies and creat additional opportunities for making the right choice for particular implementations.
According to messages from the plaintiffs, the Oak Park, Illinois, lawsuit over the health effects of Wi-Fi signals on children has been withdrawn: As recently as this last week, the Safe Technology for Oak Park (STOP) parents' group was planning on pushing ahead. Their suit was an effort not to win punitive damages, the group said in the past, but to force the school district to address the health concerns over microwave radiation that STOP had raised. The school district's response was consistently that with federal standards guiding safety, which the district had conformed with, there was no necessity to open the matter at the district.
The school district's statement on their Web site reads, After two years of examining this issue and hearing expert testimony, the board affirmed in a resolution last spring that it would continue to use wireless technology as appropriate, would monitor all research and literature, and would respond to any changes in governed regulations and standards, of which we are in total compliance.
STOP parents cited many cellular phone and cellular tower studies, but despite claims throughout the lawsuit of having hundreds of studies that were germane, neither I nor other reporters were able to obtain this list from court filings or from STOP. In an email to STOP supporters, these studies were described as "the hundreds of peer-reviewed scientific studies that show biological effects from wireless radiation (many at levels many times lower than that emitted by the in-class devises)."
Without seeing the individual studies, I have a hard time specifically refuting the contention. However, the electromagnetic radiation studies I've read and am familiar with -- from ELF/VLF studies conducted a decade ago to more recent cellular phone research -- don't show the clear results that this statement would indicate. In the studies I've read, small correlative effects in small populations of test animals or people were found in frequencies not used by Wi-Fi and at signal strengths that were at least a few orders of magnitude higher. There were no smoking guns, nor 100 percent correlations.
One study mentioned in articles quoting STOP's leaders was specifically about cell phone effects on rats. STOP's founder, Ron Baiman -- who is well known for his academic work and activism in trying to obtain living wages for workers in the U.S. -- wrote in an email to STOP members, Dr. Leif Salford is, by the way is not just "a doctor" but rather is a Professor of Neurosurgery at Lund University and specialist in this kind of research. Though his paper is available on his website -- it is published in a peer reviewed (U.S.) National Institutes of Health journal. His study was a "double blind" controlled study that used 32 laboratory rats. There is little doubt that a new drug that resulted in similar test outcomes (extensive neural damage only on exposed rats that correlated with dose levels) would be immediately withdrawn from the market pending further testing. The rats had a life time cumulative exposure of only two hours.
While the study has disturbing results -- I read it and analyzed it here -- it had a small sample size and has not, to my knowledge been repeated elsewhere, nor have similar outcomes been seen in human populations. The paper also doesn't demonstrate that tumors were formed, but rather lesions were created. It's a frightening read, but given that we're not facing the kind of short-term public health crisis that the study would predict -- massive brain lesions among millions of heavy cell phone users -- more work is needed. Finally, it uses frequencies and signal strengths that aren't involved in Part 15 devices, including Wi-Fi.
These points I raise don't mean that there is no potential health effect from the widespread use of extremely low-intensity microwave radiation used by Wi-Fi, Bluetooth, cordless phones, hospital medical equipment, and other devices. What it means is that even studies involving hundreds to thousands of times the amount of exposure show limited correlative effects and that the physical likelihood of Wi-Fi being an agent of ill health are infinitesimally low. Cell phones may be another matter, but the scientific jury is still way out on that, too.
Apple Computer petitioned the court as one of the defendants, the article cited above notes, about the company's contention that the court didn't have jurisdiction over the matter. The FCC recently reaffirmed in an unrelated decision that only the FCC can control the use of unlicensed Part 15 devices, and that landlords and other groups cannot separately regulate unlicensed airspace. While the issues are different, the basis of that decision is the same. The parents would likely have to sue the FCC over safety guidelines and win to achieve their goals.
In a mailing to a list of STOP supporters and journalists Friday morning, Baiman wrote:
On the advice of our Attorney we have withdrawn our complaint. Recent court decisions brought to our attention by the Apple filings have made it clear that we cannot win the jurisdictional argument. Needless to say we continue to believe that the substantive health risks are serious regardless of these legal arguments and continue to believe that a responsible School Board would (at a minimum) have an official policy allowing concerned parents not to expose their children to unnecessary microwave radiation.
The court document history is available here, but it doesn't include filings, so sheds little light on the matter. The original suit's text is available with students' names removed from the Oak Park School District page on the lawsuit.
At another point in the email that Baiman wrote, in a section that I believe was written several months ago and quoted, he writes: On a related matter, evidence has recently surfaced of 14 deaths, mostly of persons in their 40's and 50's, many from brain tumors, in a two block area around the "cell tower park" at Madison and Austin. Activists in Austin and Oak Park STOP members are trying to get media attention and an official investigation into this.
Epidemiological studies have to eliminate all variables that could contribute to this kinds of patterns, including statistical clustering. Given that Baiman is an economist, and one that I admire for demonstrating the clear decline in real wages in this country, his lack of scientific rigor is surprising.
When children's health is at stake, the highest standards for safety should be used. If there's a potential, unproven risk that's been demonstrated causally and requires more study, children shouldn't be exposed to the risk until it's clearly disproven. In this case, however, the preponderance of scientific evidence coupled with pure analysis of the physics of radio wave energy levels show that children's well being isn't being gambled with.
The founder and chief architect of mesh networking firm Tropos adds his two cents to the "ugly truths" about mesh networking thread: Narasimha Chari has added his comments that clarify many of the problems with mesh networking that Francis daCosta raised at Daily Wireless several days ago. I linked to daCosta's original post and posted and linked to two replies on the topic already in this archived item.
Chari says that daCosta has his primary calculation wrong: instead of a bandwidth reduction in a single-radio mesh system where n is the number of hops of (1/2)^n, it's 1/n. This calculation results in an enormously higher throughput than daCosta (click slide from Nitin Vaidya that Chari references at upper right for the dropoff by hops in throughput).
Chari makes a number of additional useful points on this topic that are worth considering. Tropos has deployed through its integrator and resellers the largest number of mesh points in commercial use in the U.S., and potentially worldwide, that I'm aware of. If they don't know what they're talking about, then a lot of networks are operating by hypnosis instead of science.
A conceptual contest winner envisions Bluetooth luggage tags: The tags on your luggage and that you carry signal when your luggage is near. Conceivably, this concept could be expanded so that you would register your luggage and your tag: without both parts, you wouldn't be able to leave with the bags. [link via Gizmodo]
The Register, a British tabloid style technology news site, accuses Your Humble Editor of misdeeds: You have to love Andrew Orlowski's style. I do. In this brief piece commenting on my New York Times article about commuter Internet access -- in which Wi-Fi isn't the point, just a mechanism -- he manages to climb on his hobbyhorse about for-fee Wi-Fi hotspots failing. Andrew, you and I can race hobbyhorses: I say Wi-Fi will spread and you say...wait, you agree!
The commercial proposition for hotspots remains in my mind as perilous as it is in Orlowski's; I'm more optimistic that bundling with cell voice and cell data plans will be the ultimate way in which pay-for-use Wi-Fi will persist. T-Mobile's pricing, for instance, is pretty reasonable and I know a number of people who have T-Mobile voice, GPRS data, and Wi-Fi because it's simple and ubiquitous--it's the cheapest solution to a problem of connectivity.
It's possible, and many folks including me admit this, that for-fee Wi-Fi won't actually be a long-term strategy if bundling doesn't succeed. Wi-Fi could be available for free as a necessary amenity at coffeeshops and hotels. (I continue to believe, too, that airports and conference centers will charge for Wi-Fi even if no one uses it because that's how those institutions think about amenities.)
That's a little beside the point, though, because my article was focusing on whether commuter Internet access was viable and feasible. All of the current experiments use Wi-Fi as a means to distribute access, but none of the experiments are charging for service. It's not a question of "Will commuters pay?" yet, but rather "Are commuters even interested?" If they are, pricing will be figured out. Using Wi-Fi on a boat that you're on for an hour or 90 minutes a day is an entirely different matter than a traveler who might need to use hundreds of different hotspots, and has the choice to pick free ones. Commuters aren't road warriors: they're captives to convenience and conveyence. (I like that phrase: perhaps I have a career in tabloids, too.)
More seriously, Orlowski accuses me indirectly of drinking Intel's blue soup, as it were, because I quote an Altamont Commuter Express (ACE) rider who works for Intel. Orlowski writes:
The Times discloses that the name of this Wi-Fi user was provided by the trial operator, PointShot. It doesn't mention that PointShot's experiment is funded by Intel, who we learn in the article is also Dickson's employer. Intel's capital fund has helped to sponsor a number of Wi-Fi trials in North America and Canada.
I had looked into this before quoting Terry Dickman--Andrew, check that typewriter, his name's not Dickson--to ensure that he had no ties to ACE or PointShot. I had also made sure via ACE and PointShot that I knew who was funding what. University of Phoenix is paying for the service on ACE, not Intel. Further, PointShot's rounds of private investment are publicly available through press releases, and Intel Capital is not an investor.
Intel did fund a marketing study in Canada with PointShot on the VIA railway line to determine passenger attitudes about having Internet access en route. The VIA line is experimenting with this offering between Toronto and Montreal.
The CEO and president of PointShot, Shawn Griffin, read Orlowski's article as well and confirmed via email all of the statements made above. PointShot has no ongoing or financial relationship with Intel or Intel Capital beyond that one survey.
Because commuter Internet access is so new and so sporadic, PointShot offered customers to talk with. I asked PointShot and the individuals I spoke with to confirm that they had no connection with ACE or PointShot and were not family or friends of people involved with either group. Mr. Dickman was a wonderful person to interview, because having Internet access on his train has given him back hours of his day.
We disclosed this in the story to make sure that readers knew from whence we obtained this perfect user. He sounds perfect because he's self selected: he liked the service so much that he contacted PointShot and then agreed to be interviewed. A spokesperson for ACE told me that they have 45 to 60 users on an average day out of about 1,500 weekday riders. Virtually all of these people are commuters.
Another error is that Orlowski states Intel Capital has sponsored Wi-Fi trials. In fact, Intel Capital has invested in companies that produce wireless LAN switches, cellular/Wi-Fi technology, and run hotspot networks like STSN or provide back-end technology to hotspot networks. Intel itself put massive co-marketing dollars into promoting Centrino's Wi-Fi service, which allowed hotspot networks to advertise like crazy in the last year. I am unaware of Intel Capital specifically funding experiments like PointShot's ACE run, the ferry system, or even hotspot networks qua experiments (as opposed to commercial deployments that parallel existing networks).
I wrote to Orlowski asking him by what means he determined that Intel was funding ACE, but I don't expect to hear back from him, as I have, tabloid like, called him names on my personal blog for, ah, his unique styling of facts and circumstance when it affects the reputation and careers of others.
Update: Orlowski sent me email and updated the article noting that Intel's investment was in Broadreach (via Intel Capital), not PointShot. Broadreach is working with PointShot in the UK (with Virgin, it appears from recent reporting) to install PointShot equipment on trains. Broadreach would operate the network. The rest of the points remain as they stand.
Political blogger 601am.com has been told there will be no Wi-Fi on the Democratic National Convention and Republican National Conventions floors: We don't have this confirmed yet, but Aaron Bailey of 601am.com, a staffer for a publication covering the conventions, was told by several authorities that Wi-FI interferred with "broadcast television signals" and was thus banned.
This rang a bell, and I contacted Tim Pozar, a microwave expert who recently joined the Electronic Frontier Foundation's staff. Tim writes extensively about the Part 15 FCC rules which govern use of unlicensed bands, such a the 2.4 gigahertz (GHz) band in which 802.11b and 802.11g (the two most popular Wi-Fi flavors) perform their magic.
Tim noted that Electronic News Gathering (ENG) is a licensed -- and thus priority -- use of a part of the 2.4 GHz band. ENG is used to send signals from television cameras to remote trucks or studios. The Part 74 ENG use is licensed, which means that unlicensed Wi-Fi users can suck eggs if they don't like being banned. Tim suggests that the frequency coordinator for the venues made a big stink about this as likely every ENG channel will be in use in the conferences. They don't need unlicensed "mucking" up the band. Tim has written this paper about licensed uses that overlap unlicensed bands.
Tim notes that 802.11a, which uses the 5 GHz band, could have been deployed successfully, but few people have 802.11a or 802.11a/g cards. It's still not common technology because of its lack of backwards compatibility and shorter range at the same power outputs.
More on this story as we receive information from the field. I'm especially curious to see if newsgatherers who create their own software base stations or plop Linksys access points onto the expensive Ethernet drops they pay Verizon for will be hunted down and shut off.
Take Control of Your AirPort Network is an ebook aimed at Mac users who want to build, secure, and extend a wireless network: My latest book -- electronic only, $5, 89 pages -- covers the ins and outs of setting up a wireless network for Macintosh users. I spend a lot of time covering Apple AirPort, the dominant method by which Mac users cut the cord, but I also present the less expensive alternatives and associated tradeoffs.
The book covers how to choose a base station, solving basic configuration problems, and has an 11-page section on setting up your own dynamic addressing using DHCP and NAT through a variety of techniques. I also go deep on extending range and securing a setup, with appendixes that include configuring AirPort Express (shipping any day now), and picking alternative Wi-Fi adapters for both older and newer Macs.
The idea behind this series of books is to provide highly focused, short titles at a low price. For five bucks, how can you go wrong?
Reuters reports that the FCC commissioners voted 5-0 in favor of what sounded like a Nextel plan to reorganize an ugly band: Nextel owns tiny slices of a very ugly chopped up range in the 700 and 800 MHz, and their proposal was to reorganize it so that the public-safety purposes to which other slices are put can be consolidated and provided with more reliability.
For their part, Nextel would receive a contiguous range in the 1.9 GHz band, and be less likely to cause interference in public-safety bands--which many blame not on Nextel but on the equipment used for fire, police, and emergencies--and have an easier time in providing cellular coverage at those higher frequencies.
The FCC values the deal at $4.8 billion, but will give Nextel credit for $1.6 billion in returning the lower-band frequencies. Nextel has to put up a line of credit for $2.5 billion to cover relocation costs, which are allowed to exceed that amount. If Nextel's final bill is less than $3.2 billion, they have to pay the government the difference.
Nextel's response was muted; they had wanted to pay $1.36 billion. Other cellular operators were complaining that the deal gave Nextel too much; it looks like the FCC chose this approach to make it worthwhile to public safety without letting Nextel off the hook.
The Washington State Ferries are leading the rush to bring wireless Internet service to regular commuters: I contributed a couple of stories to the Thursday New York Times Circuits section about commuter-Fi: Internet access available by Wi-Fi relayed from cellular, satellite, and fixed or mobile land services. The Washington ferries will soon have the largest service by orders by magnitude that serves commuters: potentially, they could have 10,000 users a day by fall. The next biggest in the U.S. with regular service is Altamount Commuter Express (ACE), which I write about in the article as well, but which has just 45 to 60 users on an average day.
Commuter Internet access is potentially the single largest concentrated market for data users. While six million people commute to work via public transportation on an average day in the U.S., only a tiny fraction have access to any live Internet signal at any stage in their journey at any price. I don't have the percentage of those commuters who carry laptops or handhelds, but specific routes already have a significant percentage of passengers already using their laptops en route. (More staggeringly, by the way: 75 million trips are taken each average weekday on some form of public transportation on an average weekday, with 66.5 million of those trips by bus.)
This is a market that cellular data operators would love to tap into, but there are two limiting factors: cell data isn't ideally equipped for people riding at high speeds inside metal containers; and many of the commuters who might use a free service (in exchange for looking at advertising) or pay $20 to $30 per month for high-speed access wouldn't pony up $350 for a PC Card and $80 per month for the comparable cellular service if they don't need it outside their commute. That is, you have plenty of commuters who don't meet the road warrior profile, but have every bit as much need for data access during their daily shuttling.
The main article focuses on the various efforts that are well underway in the U.S., and mentions a few international projects, such as GNER's recent rail launch, the Linx train service in Sweden/Denmark, and Paris's bus project. The meat of the story looks at the Washington States Ferries use of a government grant to test the ways by which high-speed service could be offered continuously. The Port Townsend company that's building out the test system says they could offer well over 10 Mbps with their configuration if it's called for. (That firm actually specializes in military applications; the ferry system project is a rare case in which civilian work will pay back in military projects.) I was stunned to learn that the Washington ferries represent nearly half of all ferry trips in the U.S. each year.
The ferry system's test uses fixed antennas to provide bandwidth. The other transportation operators I spoke to are either trying or planning to try cellular data or cellular and satellite. The illustration (thumbnailed above) that accompanies the article shows how PointShot, a rail Internet access builder, uses cell, satellite, and caching servers to provide continuous service.
In a sidebar, I cover the options in the Northeast, which are scant. You can get low-speed but good service on a shuttle from Manhattan to Boston or Long Island, but the MTA (which operates the rail lines around the city) has no plans, and Amtrak doesn't think the technology is ready, although it did allow AT&T Wireless to install Wi-Fi in six Northeast train stations.
I took a day trip to Port Townsend to test out the currently only unwired ferry in the fleet; you can see some of my photos here.
AirCell says it successfully showed broadband speeds on an airplane uses ground-to-air communication: AirCell reports average speeds of 300 to 500 Kpbs, and peaks of 2.4 Mbps. Their plan is to use a network of ground-based towers to offer cellular voice and data in planes with passengers using their own, existing cellular equipment. The test was with a business jet, but they say they looked at many parameters. The company says it is in discussions to offer the service commercially by next summer.
Connexion by Boeing and Tenzing on focused on satellites to bring bandwidth to planes, but AirCell and Sky Way (which has a more Homeland Security oriented business) are using ground towers pointing at the air. It could wind up being substantially less expensive if they can get the frequencies and the sites, especially if they can backhaul bandwidth to the ground towers wirelessly. [link via Daily Wireless]
JiWire collapses providers into locations, offering a more distinct look at the world: JiWire has just finished a massive update to their site which features some redesign, but most significantly changes the way in which they count hotspots. Previously, each provider at each location was counted once. Now, they count providers at locations separately from unique locations. Their current count is 46,799 hotspots (providers for each location) and 35,695 unique locations, which are physical addresses with one or more provider offering service.
This method extends the principles of counting uniquely to better understand the scope of networks that I've written about recently. JiWire continues to make it simple to find free hotspots. Click Hotspots at the top of any page or Advanced Search near any hotspot search form and you can check the Free button under Access Fee to restrict results to free locations.
You can also read the first of an excellent service of features written by my friends Richard and Angela as they travel with their four kids of all ages from Maine to Michigan down to Texas and back to the lobster state by RV. They're running an Internet business as they spend a month traveling, and need relatively constant access to stay on top of their operations.
Linksys releases legal antenna add-ons for several of its Wi-Fi gateways: We've said it in this space before, and we'll say it again: in the U.S., you cannot legally mix and match antennas with Wi-Fi gateways. The FCC's Part 15 rules require that you only use antennas that were certified with the gateway that they were designed for. In response to a recent query made by a colleague of mine, the FCC wrote back a very clear response:
When using Part 15 (2.4GHz) devices, mixing and matching components is a violation of the rules. Only systems can be certified. Please see http://wireless.fcc.gov/outreach/2004broadbandforum/comments/ YDI_externalamps.pdf.
There you go. Oddly, it is perfectly legal to sell antennas separately from gateways without violating any laws. It's only deploying them that is technically illegal. We haven't heard of any letters or prosecutions, however.
Linksys isn't getting around this, but has rather taken several antennas and several gateways and paid the fees to get each of the combinations certified in a matrix. This cost them at least a few hundred thousand dollars in fees and associated expenses entirely separate from the development costs of the antennas themselves.
Linksys has two kinds of antennas and matching stands for each. One add-on, the TNC Connector Antenna, is designed for two-antenna systems: WRT54GS, WRT54G, WAP54G, BEFW11S4, and WAP11. The one-antenna device, the SMA Connector Antenna, works with more specialized equipment: WRV54G, WMP54GS, WMP54G, and WET54G.
Both antenna types cost $60 each (suggested retail), while the respective antenna stands are another $30. The stands allow mounting the antennas to a wall or ceiling.
D-Link meanwhile introduced two antennas, Tom's Networking reports, that appear to be freestanding: they don't discuss them being tested with systems, and thus are illegal to use as they suggest unless I'm missing the fine print. The D-LInk specs says, The D-Link ANT24-0400 Omni-Directional Indoor Antenna connects to a variety of D-Link Wireless PCI Adapters, Routers, and Access Points. A footnote reads, Works with any 802.11b/802.11g compliant devices with a SMA or TNC connector only.
A strange little announcement from Starbucks yesterday about its Wi-Fi service: It's hard to figure why Starbucks was compelled to announce they'd crossed the magic 3,100 store threshold--why was 3,100 such an important number? They added stores in five states that apparently had no or little service before, including Indiana and Wisconsin, which doesn't seem quite newsworthy either.
The release says that service rolled out in August 2002, when it actually started in May 2001 and went official later that year. The company has erased the entire May 2001 to August 2002 experience from their minds, despite the several hundred stores that had access during that time, including the doomed branch in the World Trade Center at which my friend MIke Daisey was working that morning and managed to get far away in time.
The most interest part of the Reuters story prompted by a press release was that subscribers to the service--they don't say how many--use a T-Mobile hotspot on average after 9 a.m. for an hour eight times a month. That's great market insight.
DSL providers have gained on cable Internet marketshare; Wi-Fi is one lure: The Wall Street Journal reports that the efforts by Verizon, BellSouth, Qwest, and SBC coupled with reduced prices are paying off. More new broadband customers chose DSL over cable in 2004's first quarter. Cable companies are warming to the notion of Wi-Fi for intra-home sharing, however.
Cable companies still trot out the "sharing Internet is like stealing cable" for connections shared beyond a home. It's clearly not. The law governing cable theft doesn't cover Internet sharing. It may certainly be a violation of the terms of service, and thus a reason for canceling a cable modem subscriber's account. But I sincerely doubt that a cable firm would win a prosecution under the cable theft laws which govern programming.
The head of a mesh company argues mesh networks don't scale -- and one of the folks behind an open-source mesh software project examines the argument: MeshDynamics sells a multiple-radio solution for mesh networking, and the head of the firm wrote a brief article explaining why single-radio mesh networks can't work beyond a very small deployment. I asked Sascha Meinrath of the CUWiN project for his feedback on Francis daCosta's comments.
(After this item was posted, Jim Thompson contributed his own extensive thoughts about both daCosta and Sascha's statements, hence the "third" view of the revised title.)
While I do think that Francis daCosta brings up some potential pitfalls to wireless mesh networks, the doomsday picture he presents is based on a flawed understanding of how mesh networking topographies work. I'll explain below:
1- Radio is a shared medium and forces everyone to stay silent while one person holds the stage. Wired networks, on the other hand, can and do hold multiple simultaneous conversations.
2- In a single radio ad hoc mesh network, the best you can do is (1/2)^^n at each hop. So in a multi hop mesh network, the Max available bandwidth available to you degrades at the rate of 1/2, 1/4, 1/8. By the time you are 4 hops away the max you can get is 1/16 of the total available bandwidth.
This problem exists only when all tranceivers within a mesh topography "see" each other. And herein is the flaw in the argument. Within a mesh network Request To Sends (RTSs) do silence nodes within range; however this degradation moves in waves--so if part of a mesh consisted of 7 nodes (of which G is connected to the Internet):
A-----------------> ------B-----------------> ------------C-----------------> <-----------------D-----------------> <-----------------E-----------------> <-----------------F-----------------> <-----------------G-----------------> | Internet Connection
Here's what would happen. A would pass a packet to B; when B passed a packet to C, A couldn't talk--thus the 1/2 reduction in throughput; when C passed it to D, the same problem would occur for both A & B (thus a 1/4 throughput); likewise for D to E (because D would silence A, B, & C), thus a 1/8th throughput. However, when E passes a packet to F, A is unaffected, when F passes a packet to G, both A & B are unaffected. Thus, in this solution, throughput would theoretically max out at 1/8th (which is probably still much more throughput than the average Internet connection--where the usual bottleneck resides).
What this really points to is the need for power control in radios (which is something that CUWiN wants to work on), smart antennas, and other innovations that help to create wireless topographies where as few radios as possible "overlap." I've written about some of these solutions in a paper that I'm adapting for a book chapter -- you can download this.
3- That does not sound too bad when you are putting together a wireless sensor network with limited bandwidth and latency considerations. It is DISASTROUS if you wish to provide the level of latency/throughput people are accustomed to with their wired networks. Consider the case of just 10 client stations at each node of a 4 hop mesh network. The clients at the last rung will receive -at best- 1/(16,0000) of the total bandwidth at the root.
This simply points out the need to separate inter- and intra-nodal communications architectures--a problem that CUWiN has both already identified and implemented.
4- Why has this not been noticed as yet? Because first there are not a lot of mesh networks around and second, they have not been tested under high usage situations. Browsing and email don't count. Try video -- where both latency and bandwidth matter -- or VOIP where the bandwidth is a measly 64Kbps but where latency matters. Even in a simple 4 hop ad hoc mesh network with 10 clients, VOIP phones wont work well beyond the first or second hop -- the latency and jitter caused by CSMA/CA contention windows (how wireless systems avoid collisions) will be unbearable.
I do agree that QoS problems continue to plague most mesh wireless networks. It's a problem that needs to be solved and that most deployments and commercial (and open source) solutions sidestep. I think Francis is wise to blow the whistle on this deployment problem; I think that many commercial mesh systems have been way oversold--which will only make the problem worse.
I am constantly amazed at how little most wireless companies know about the physics, software, and hardware of the networks they deploy. Most don't even realize that if they're using routing protocols that use Standard Link State they're going to crash and burn when they scale up. For a quick graphic of the problem, just check out page 29 (labeled page 26) of this link.
This is why CUWiN is creating an A-HSLS (Adaptive Hazy Sighted Link State) protocol (as far as we know, the only open source A-HSLS protocol). We believe that routing overhead will kill networks well before throughput does.
I am optimistic that solutions will be forthcoming. What we really need today are "altruistic venture capitalists"--folks who are interested in investing in the public good -- people who will sopport the development of CUWiN (or other open-source projects that are working on these solutions) so that we can build mesh wireless systems that not only work and scale, but exceed our current expectations of what we, today, believe is possible.
Sanswire demonstrates its Stratellite technology--perhaps something less risque would have been a better name: Stable at an altitude of 13 miles, the Stratellite is supposed to offer the kinds of benefits for broadband and voice service that typically would be reserved for ground-based communication. Think of the awesome line of sight! They're demonstrating the technology in July, and then plan to set a date to deploy an actual unit.
Wineries may use wireless to sense vine conditions: Sonoma County growers are testing a Wi-Fi-based monitoring system to help reduce and control inputs like water and pesticides. It may also provide more detail about the pico-climates to produce more reliable crops. Rodents chewed through some of the cables for the monitors--it's not 100-percent wireless, of course!--setting back the project a bit. [link via Agen Schmitz]
Emirates airline uses Wi-Fi to connect to Tenzing service: Tenzing had hoped to make Emirates the first Wi-Fi equipped airline to offer Internet access, but Lufthansa and Boeing beat them by a few weeks. However, Emirates is the first airline to use Wi-Fi that was built into the plane during construction. The Airbus A345 was delivered in December with Wi-Fi ready to go. The equipped planes fly from Dubai to New York, Sydney, Melbourne, Auckland, and Christchurch.
Tenzing's current generation of service is quite slow compared to Boeings: a maximum of 128 Kbps. Emirates is using their email-only service as well, which provides access just through a proxy to email accounts. Tenzing's next-generation service will increase to about 450 to 900 Kbps in both directions, or even faster depending on an airline's choice of options; it's due in 2005.
iPass say they crossed the 10,000 hotspots mark in their aggregated network and then some: iPass now offers 11,179 Wi-Fi hotspots worldwide through aggregated partner networks. The release also says that they have 12,000 more under contract for the future, which is a quite extraordinary number.
The 2,900 hotels included in this total all have Wi-Fi in at least public spaces, a company spokesperson confirmed. Many of them may have wired in-room broadband, however. An increasing number of hotels are starting with Wi-Fi from the get-go; others are converting from wired broadband to Wi-Fi for greater flexibility as costs drop, range increases, and wireless backhaul becomes simpler.
The largest hotspot operators in the U.S. have boiled down to three companies for the near future: T-Mobile (about 4,500), Wayport (several hundred hotels with a mix of Wi-Fi and wired broadband, plus thousands of McDonald's coming online this year), and SBC FreedomLink (several thousand The UPS Store outlets, managed by Wayport).
55 students unions will be unwired: SUBzone, which is building out the network, says 1.2 million students will be covered through their efforts starting in Leeds. Pricing isn't noted.
T-Systems has 10,000 hotspots in the network they're talking with: Tony Smith writes that there are plenty of aggregated networks deployed or in the works, and execution will be the factor for T-Systems just as with the other announced players. He sums it up neatly: We've heard similar bullish statements from other Wi-Fi companies with an eye on this level of partnership, but all anticipated announcement dates have come and gone without a peep.
The high throughput camps have formed to win the heart and soul of 802.11n, Wi-Fi's next speed bump: The EE Times explains that one camp led by Atheros proposes using 40 megahertz (MHz) of bandwidth (up from 22 MHz in the 802.11b/g specs); Intel is part of this proposal. They call their approach TG nSynch--the boy band pun must be intended. On the other side is multiple-in/multiple-out antenna system maker Airgo which plans to build around 20 MHz of bandwidth; Broadcom and Texas Instruments most significantly stand on their side. Their name: World Wide Spectrum Efficiency (WWiSE).
China's hotspot market grows, but users don't follow: The payment system might be hampering some of the growth of various networks throughout and near China. The cost also seems incredibly out of line: China Netcom charges US$3.62 per hour and service must be purchased in advance using a hard-to-find prepaid card. China Netcom has just about 160 locations, most of which were installed nearly a year ago.
Shanghai Telecom has several hundred hotspots, and pushed usage initially through an ADSL bundle. China Mobile has nearly 1,500 hotspots, but they allow users to pay via their cellular phone bill using SMS to confirm purchase of time.
South Korea's usage is mentioned in passing: with tens of thousands of hotspots in that country, hundreds of thousands of people are regular users.
GNER has started the commercial launch of its onboard Internet access service on the East Coast Maine Line in the UK: Swedish firm Icomera has unwired the first commercial GNER run--previously, it was all testing--and nine more runs will follow. First-class passengers have free access to the continuously available wireless service, while other riders pay £2.95 to £9.95 per trip.
A fascinating estimate lurks at the end of this BBC story: According to telecoms analysts BWCS, 625 million people worldwide will be travelling on wi-fi enabled trains within the next five years.
We've clarified a couple of points on recent posts about upgrading Alvarion BreezeMax 3500 "pre-WiMax" gear to interoperate with certified WiMax equipment: Alvarion isn't offering the upgrades for free as part of the original purchase. Once the WiMax Forum begins certifying gear and Alvarion understands what will be necessary to upgrade the products, it will negotiate the cost for upgrading the gear with customers that are interested in the upgrade. See Glenn's post and mine for the changes.
Once the company begins marketing certified equipment, it will continue to support the "pre-WiMax" equipment, but will stop selling it. Also, the upgrades to the "pre-WiMax" products won't make the products identical to the line of certified gear. Instead, the upgraded network will become something of a hybrid, able to support the original customer premise equipment (CPE) as well as certified WiMax CPEs from any vendor.
Some buyers of the "pre-WiMax" equipment aren't actually buying it because it can be upgraded, said Carlton O'Neal, vice president of marketing for Alvarion. "They don't care if we call it WiMax or a tomato, they want it," he said. This set of customers is buying products because they meet their current business cases, he said. He believes that the willingness to buy uncertified products is an indication of a new confidence in broadband wireless as a viable business.
PC World investigates whether WPA labeling matches WPA certification: Sean Captain discovers that even when a product says it has WPA, you might need to check twice. Microsoft's now-discounted 802.11g router said on the box that it had WPA certification, but that's not the case. The device slipped in under the wire for Wi-Fi certification requiring WPA late last year, and the upgrade for WPA isn't due until this month. The Wi-Fi Alliance's database of certification had the correct details, however. Captain warns that if you need WPA, make sure the device has it: just being 802.11g isn't enough.
We can't quite picture WiMax becoming as ubiquitous and cheap as Wi-Fi in two years, but Intel can: Intel stands by its expectation that WiMax technology will be embedded in its notebook computers by 2006. What will these transceivers pick up? Hard to tell. WiMax isn't poised at the moment to become a mobile or fixed public space replacement for Wi-Fi or for cellular 3G. But Intel must expect it's on track for that. But remember that Intel put 802.11b into its first-generation Centrinos and has only recently added an 802.11g option into Centrino's successor.
My hat is off to Winda Benedetti who has written the definitive article on free coffeeshop Wi-Fi for the Seattle Post-Intelligencer: This lengthy, funny, accurate article on the increasing number of coffeeshops who offer free Wi-Fi as part of their basic business model takes the cake--the coffee cake. It's the best piece written on the subject. (Disclaimer: I write regularly for a competing Seattle newspaper, and have written about coffeeshop Wi-Fi, too, so I'm crying espresso tears as I praise this piece.)
Benedetti extracts all the different facets to the free-versus-fee battle in this realm, including the expectations of the shop owners, the kind of traffic they see (more anecdotal than dollars-and-cents), and some of their righteous anger against competitors who have the temerity to charge for a connection.
There are some nifty subsections about the etiquette that coffeeshops hope their customers engage in--like buying stuff if they use the network--and whether paying for service buys you more reliability than using free networks.
The article ends with a long list of coffeeshops in the Seattle area with free Wi-Fi. I now have several within a two-mile radius of my office--two are within one block, even.