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 2001-2013 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
Our site's top stories viewed during 2005 are mostly from previous years: It's the irony of a blog that your traffic is either on your home page or in the archives. Stories that age well tend to continue to receive high traffic.
1. WPA Cracking Proof of Concept Available (Nov. 2004)
2. Weakness in Passphrase Choice in WPA Interface (Nov. 2003)
3. Coffeeshop Turns off Wi-Fi on Weekends (May 2005)
4. WPA for Free under Windows 2000 (Feb. 2004)
5. Municipal: Texas Fights; Indiana Bill Dies; NYT Covers; Philly Councilman Shills; Colorado Suppresses (Feb. 2005)
6. Change Your Linksys WRT54G Admin Password Right Now! (June 2004)
7. The Path to 802.11i (Nov. 2003)
8. GoogleWiFi (Apr. 2005)
9. Most Wireless Speakers Don't Live Up to Goal (Sept. 2004)
10. WPA's Little Secret (Nov. 2003)
You'll notice a trend here: Six of the top 10 stories are about security, four about Wi-Fi Protected Access. Only three stories in the top ten ran in 2005, while three date back to security issues first written about in 2003! And the 2nd most popular story was a white paper written by a security expert that we were given permission to post. Top story number 9 is quite curious in that it's just a one-paragraph pointer to a David Pogue column.
There's an awkward dance between access point and Wi-Fi adapters: A Wi-Fi adapter is full of 1999 and 2002 technology; Wi-Fi access points, enterprise WLAN switches, and gateways have the latest goodness. How can the industry add features to adapters that allow them to be as smart and manageable as the AP side? A new IEEE standard, of course.
802.11v will allow configuration, diagnostics, and dynamic adaptation of settings. Those will all allow enterprises to save enormous amounts of time in provisioning devices--although some end-point security tools can already handle some of this--and troubleshooting. The dynamic changes in settings would allow devices to respond to network conditions or control signals sent via an access point to an adapter.
Most importantly, perhaps, adapters could participate in load balancing among access points, eliminating kludges by WLAN vendors that push and pull adapters to the APs they want them on.
Lompoc, Calif., has cut the cord: The first stage in their two-prong municipal network will launch Jan. 3, with ubiquitous Wi-Fi coverage across the town. Some good early feedback helped them increase the density of access points by 70--from what number, it's not reported. Service will be $20 per month with a $5 per month equipment rental for bridges, if needed.
The articles doesn't mention that the Wi-Fi system isn't the ultimate goal, but rather that fiber is the next stage. The Wi-Fi network provides short-term residential bandwidth and mobility, while fiber will be part of the longer-term plan to make the time more attractive to business while increasing access by residents and business to higher-speed offerings.
Hampton Jitney runs its buses not just between Manhattan and its antipodes in Long Island: The company also runs a car-and-bus shuttle (you plus a car) down to Florida's east and west coasts, with an Ambassador option. It's $1,528 each way for two people in a 31-person bus and their car (on car trailers), and now the company also offers Wi-Fi, as it does on its New York fleet.
Spain's Renfe rail operator will add Wi-Fi: They've been testing a service they plan to roll out using satellite Internet access.
San Francisco's request for proposal for its citywide network is out: The city published a PDF of the RFP today; responses are due Feb. 21, 2006.
Has your latest Linksys caused you grief or joy? I ask this question not because I have statistical evidence, but rather anecdotal. The last two Linksys devices I purchased have operated erratically. I advised a colleague to buy two (for bridging) recently and he could not get either to stop crashing even with the latest firmware. Checking forums, I'm seeing similar stories.
Linksys has moved from Linux to VxWorks (a proprietary operating system) for its latest routers and may be gradually transitioning its devices. The firmware these gateways are running are the first releases, and may be in the process of significant revision.
If you're seeing problems with a WRT54G v5 (check the underside) or other new Linksys equipment (check the version number), post a comment below. Please post, too, if your gateways are working perfectly fine. I'll post a summary if I get many posts.
Update: On June 8, 2006, Tim Higgins of Tom's Networking posted an interesting article comparing performance of the WRT54GL (essentially version 4 of the WRT54G, still running Linux) and the WRT54G v5. Higgins focused on performance, not reliability; I'm hoping he provides a case study of a stock unit's reliability, too.
The Local Agency Formation Commission (LAFCO) grilled SF's director of technology on the muni-Fi bidding process: The oversight committee is concerned that the county board of supervisors--SF is both a city and a county--will be presented with a foregone conclusion that requires and up or down answer. Chris Vein, the tech director, said it wouldn't play out that way. This IDG article notes that a Comcast cable franchise decision was packaged in that way.
Minneapolis bidder offers cut-rate plans, more details: Two finalists remain in building a citywide fiber and wireless network: US Internet and EarthLink. The former proposes a $16/month rate for low-income residents; regular rates would still undercut current cable and DSL broadband pricing. (Until the incumbents offer 1-year, 60% rates as they have in other communities.)
The article notes Minneapolis will shift $1.5 to $2.5 million in telecom costs (voice and cellular) to the new network, while US Internet estimates $25 million in expense to build the hybrid network. EarthLink did not provide a number.
US Internet also plans to deploy what sounds like 4.9 GHz public safety wireless alongside regular Wi-Fi. EarthLink will use a VPN and a VLAN, although the article makes it sound like it's just a VPN. US Internet will use WiMax for backhaul; EarthLink will use Motorola Canopy, which ain't far from WiMax. (The article paraphrases an analyst, who says EarthLink "every third Wi-Fi hot spot would relay information back to the data center through a series of Wi-Fi interconnections." It's a paraphrase to be sure: it's not Wi-Fi at all, but unlicensed point-to-multipoint.)
The article ends with a local group's position that the city shouldn't bid out and contract out this network. The city's point person says words that should please every moderate, while probably inflaming pro- and anti-municipal advocates: "The city lacks the money, competence and ability to build and manage that kind of a network right now."
15-year-old prompts free Wi-Fi in Carlisle: The downtown business association in the Borough of Carlisle liked Thomas Blitz's notion of free Wi-Fi downtown. It's in three restaurants--free of charge from PA.net--and more may come. PA.net also put Wi-Fi in the nearby public library.
Navizon offers some competition to Skyhook: This new service from Mexens Technology uses a form of peer-to-peer mapping. The seeding must come from people with either cell radios or Wi-Fi built in who also have a GPS (Global Positioning Satellite) receiver. These seeders can upload their information regularly, while all users of the service--currently free--can receive downloads of new locations when they synchronize.
The company also ties into Google Maps via its Web site to show registered users locations that have been mapped. Registration is also free. This is an interesting variation on wardriving which requires enlightened self-interest to succeed: Volunteer enough foothours of mapping and, if others do the same, a city could quickly be well-covered.
Navizon also adds applications, like finding buddies or (soon) tagging information to locations.
The missing piece in reading this Wi-Fi Planet article is whether once a city has a basic level of coverage that someone with just a cell radio or Wi-Fi could add additional points without carrying a GPS. Skyhook extensively pre-maps a location, but the company told me months ago that their software will report back new information to be integrated from software running in the field on users' computers, too.
At one point, BPL was seen as the third broadband leg: Cable, DSL, and BPL would provide robust competition in the marketplace. Instead, it's turning out to be cable, DSL, and much slower wireless today with WiMax potentially tomorrow. But BPL just got some traction with a Texas deal.
Current Communications Group and TXU Electric Delivery will offer BPL to two million Texas customers by the end of 2006 in busy markets like Dallas-Ft. Worth. They've chosen the correct flavor, and that's been one of the battles. Broadband can't transit transformers directly, and the "final 50 feet," let's say, has been tricky.
There have been many proposals to stick Wi-Fi transmitters on the transformers to bridge that gap; other methods, which seem superior, extract the data from the line coming into the transformer and connect it to the feed to a customer's house. The latter is what Current/TXU seems to use because customers need only plug in a device into an power outlet to use the service.
TXU will pay $150 million to Current for a variety of utility-related services that the network will allow, including monitoring for line breaks and potentially remotely turning off service, which requires a truck roll now. I've heard from other utilities that remote meter reading is huge as well. Electrical meters are replaced on a regular duty cycle that can span decades, and there's always new construction. This allows a rollout of BPL-based meters to a relatively large minority of homes quickly.
Because these two firms are using the direct outlet method of BPL, devices like soda machines could report their status.
Dealmac is reporting a semi-ridiculous low price on this best-selling Linksys model: The WRT54G--almost certainly the new version that can't be modified with homebrew firmware--is $54.99 with free shipping at Amazon, but with a coupon code and a rebate coupon, it's just $37.24. Read Dealmac's instructions to get the bargain.
Phoenix activates airport: The retro-futuristically named Sky Harbor International Airport has free Wi-Fi everywhere but at check-in. The airport is a hop, skip, and a jump from Tempe, which is unwiring entirely.
The state of Maine sets goal for rural broadband: The governor wants to extend broadband to a large portion of rural areas that have none now--places with five people or fewer per square mile by his definition. He also wants ubiquitous cell coverage, which will cost about $55 million. The article oddly omits mention of Midcoast Internet Solutions, which has been using relatively inexpensive off-the-shelf equipment since 1999 to beam broadband all over. They serve a big chunk of Midcoast, which starts about an hour from Portland. (I lived in Midcoast Maine for two years in the early 90s: we still paid extra for Touchtone service back then to give you some idea of the state of telecom.) It does mention Ubiquitair, which serves the Casco Bay (Portland area).
Aurora, Ill., approves muni network: The city was first to electrify and now has $5.6 million budgeted for a municipal wireless network. They'll also spend $7.8 for fiber optic connections among city buildings, which likely replaces a lot of fixed-line recurring costs. The city will use gaming proceeds to service the network debt, but expect it to pay for itself in reduced telecom charges over time.
Farmers Branch, Texas, hires Tempe-unwirer: The 12 square miles of this town near the Dallas-Ft. Worth airport will be lit up with a municipal network by NeoReach. This project explicitly promises voice, data, and video.
Seattle Post-Intelligencer business reporter John Cook blogs about Vivato's demise with a Washington State focus: The city of Kent, a booming town near Seattle, just deployed a Vivato switch. Spokane, where Vivato had its engineering HQ, has been running a network downtown. Southeast Washington had a 1,500-square-mile network served by Vivato equipment. (Not mentioned: Portland, Oregon's VeriLAN, which launched with Vivato gear and still has a Vivato switch pictured on their home page.)
Of particular interest to existing deployments is what happens when equipment breaks, as Vivato has discontinued support. Interesting, the City of Spokane's large deployment--at one time dubbed the largest municipal wireless network in the US., Cook notes--has only 80 users per day.
The company received over $65 million in capital during its lifetime.
Update: John Cook filed a longer story with more details in Tuesday's Post-Intelligencer. He spoke to a range of Vivato users. The folks at Columbia Rural Electric, which cover 1,500 square miles with Vivato equipment, cleverly hired a Vivato engineer.
Interesting article out of the Northeastern Penn. newspaper about Wilkes-Barre's municipal Wi-Fi plan: There was a rush to get the network underway by Dec. 31 in this city in order to beat a deadline in the now-infamous law that almost scotched Philadelphia's broadband plan. But this article examines the details of what's known as Act 183. The law appears to state that any municipal broadband network that isn't in operation by the end of this year can only be built if the incumbent carrier is notified of a city or town's plan to build such a network. The carrier has two months to respond and, if they respond that they will build their own network, have 14 months to build it, the article says. (This part is tricky as the law could allow extensions of this time indefinitely.)
The article points out that the definition of municipal broadband isn't very tight, and that it's unclear whether hiring a private contractor, having a non-profit that actually signs a contract (as in Philly), or not using public tax dollars would all exempt a city from these terms.
Verizon's attempt to clamp down on municipal broadband has certainly failed on three counts. First, the publicity surrounding the submarine law created a national outcry and a loud debate with many sides involved. It spurred development of municipal broadband in cities that never considered it before. Second, Verizon was forced to sign a waiver for Philadelphia to get the law passed. Third, the actual law appears to be vague enough that Verizon would have to sue municipalities that choose to believe that it doesn't cover many kinds of networks being built. That lawsuit would almost certainly backfire as well.
The local power utility is delaying the rollout of Dunedin, Florida, Wi-Fi network: The company that the city contracted to install a Wi-Fi network has been unable to secure an agreement so far with Progress Energy, which controls the poles. The city thought it owned the poles, but had given them to the energy firm three years ago. This isn't unusual: the pole rights are governed by a variety of regulations, and sometimes utilities accept maintenance responsibilities in exchange for easier access for their own purposes.
The energy company isn't per se opposed to having Citi WiFi, the contractor, put Wi-Fi transmitters on its poles. But they don't have a streamlined process for it and they have to conform to the rules set by FCC and state rules. These rules usually require certain kinds of non-discriminatory access mediated by space on the poles and other factors.
Update: City WiFi's Frank McCarthy wrote to correct a few details. First, as the article I link to notes, City WiFi currently has approval for 8 of 10 poles they requested. Second, there is Wi-Fi service running now with transmitters on city-owned buildings and other facilities; that started up in June.
Third, Progress Energy initially stated that City WiFi would never get access to any power distribution poles; that's changed. PE is apparently now saying that City WiFi still can't have access to light poles, but McCarthy expects that position to change with negotiation, too. City official are committed to making this happen, and their political pressure is apparently helping.
City WiFi's experience should be both a cautionary tale of dealing with utilities that have their own agenda and regulations, and a point of optimism that politicians can provide a push to cut through delays and recalcitrance.
With Vivato ceasing operations yesterday, it's high time to look at three wireless data firms hailed as having revolutionary technology around the same time: My colleague John Markoff at The New York Times wrote three stories on the future of wireless technology in 2002 and 2003 that shaped the way the business community viewed the future of wireless data at home and for business use. Each story focused on one company and its context in the larger world of Wi-Fi and broadband wireless.
The first piece was in June 2002, on Etherlinx, a firm that claimed to have created a combination WLAN (wireless local area network) and broadband point-to-point router using SDR (software-defined radio), off-the-shelf equipment, and a modified version of the Wi-Fi protocol.
The second article appeared in Nov. 2002 about Vivato, which introduced multiple-antenna technology that would allow entire floors of buildings or large and distant outdoor spaces to be lit up with much greater ease than anything that existed at the time. Separately steered "beams" of Wi-Fi could be managed with separate characteristics--switched, in fact.
The third feature in Aug. 2003 focused on Airgo, which had developed chips based on MIMO (multiple-in, multiple-out) technology. They claimed that through the use of spatial multiplexing and better use of antennas they could extend Wi-Fi's range and speed.
Let's look at whether the technology these firms represented have turned into marketplace ideas. (I come not to bury Markoff, but to vindicate him...)
Etherlinx: WiMax and preceding broadband wireless standards use non-Wi-Fi technology for final-mile, point-to-multi-point connections at high speeds. This technology preceded Etherlinx, and was still very expensive at the time their coverage appeared in 2002.
Many companies now make equipment that could ultimately serve residential customers through inexpensive home gateway (customer premises equipment or CPEs). Some of the CPEs may ultimately have two radios: one for WiMax or mesh Wi-Fi reception; the other to create a local WLAN. Two radios allow coordination of channels.
Etherlinx's Web site hasn't been updated since about 2003. The last I heard from anyone at the company, they were trying to get details changed in long-ago published stories about the company related to staff.
Vivato: Smart antennas reach further. Multiple antennas produce spatial separation for increasing throughput in the same amount of spectrum. Switched WLANs produce better management control with less fuss. All of these ideas--some of which preceded Vivato--are thriving. Beamforming and spatial multiplexing using multiple antennas is the basis of 802.11n and is found in dozens of consumer gateways. Switched WLAN architecture--using access points that are coordinated centrally--has replaced using centralized equipment that relied on beams, but the switched nature has persisted.
Vivato ceased operations yesterday.
Airgo: Airgo just released its third-generation MIMO chips which have a raw maximum rate of 240 Mbps, which translates to over 100 Mbps of real throughput, exceeding widely available wired Ethernet for the first time. (The trick is dynamically expanding to use 40 megahertz of spectrum, or the equivalent of two Wi-Fi channels, along with two spatially separated data streams: this is roughly four times Wi-Fi's capacity of 20 MHz and one data stream.)
Airgo has been vindicated in the marketplace, although it's still a startup company fighting for its business among giant chipmakers and scrappy, but older chipmakers. It's holding its own through constant innovation. The incorporation of MIMO as a significant element of 802.11n through proposals that may cut Airgo out of the negotiation process may give an edge next year to other manufacturers.
Scorecard: The technology Markoff wrote about over this two-year period represented fundamental changes in the broadband and networking market, and all the ideas contained within the products--whether they originated them or were preceded by similar ideas or products--have come to market as significant forces that will dominate the industries in which they're deployed.
Vivato has ceased operations according to a company spokesperson: I confirmed via a spokesperson Thursday night that early enterprise wireless switch maker Vivato has shut down. Unstrung was reporting earlier today that the buzz on the street was a Dec. 20 halt. A reliable source told me this evening that the shutdown had already occurred, and I was able to confirm it late Thursday with the company. Update: The company posted what it calls its "wind down plans" on Saturday on its home page. They will liquidate the inventory and will have a portfolio soon of intellectual property to sell off.
Vivato made news in Nov. 2002 when John Markoff filed a major business story in The New York Times about the then-revolutionary product that Vivato was slated to introduce the next year. The company's braintrust was extolled, and the firm had prototypes to show off. In the months that followed, they offered extensive demonstrations of their technology, which involved a phased-array antenna that was intended to control and receive signals intelligently, steering Wi-Fi to users and being highly receptive to distant transmitters.
The demonstration worked terrifically. A number of San Francisco-based journalists and community wireless advocates put a demonstration switch to the test and were quite honestly amazed. But in practice, the multi-thousand-dollar gateway couldn't be put into effective production.
Delays dogged the company as, according to reports I received, the units coming off the production line were incapable of achieving the expensive, handmade prototypes' characteristics. Originally, the system was billed as delivering three simultaneous steered beams across entire floors of buildings from indoor placement or entire sides of buildings from outdoor locations, but that appears to have been impossible to achieve.
I knew that Vivato might not be able to deliver because a major PR effort to broadcast Wi-Fi across Central Park quietly failed for technical reasons--it was never announced publicly and those involved didn't want to talk about it. At the same time, normally straight-talking people within the company couldn't give me a clear and frank answer as to delays in production.
The firm reportedly created an inventory of its 802.11b switches which came on the market just after 802.11g hit ratification, making them obsolete for enterprise purposes on delivery. The inventory of that first-generation device were unsellable at retail because of performance and the 802.11g issue. A number of gateways were sold off around the country to smaller firms, colleges, and institutions at bargain prices. Several colleges I've spoken with have one or two Vivato 802.11b gateways for lighting up arenas or other outdoor spaces.
An 802.11g switch took a while to produce, but it limited claims to a single steered channel when it shipped in 2004. They paired this switch with a much cheaper bridge that would fill in niches that the main device couldn't cover. Between shipping their 802.11b and 802.11g switches and thereafter, founders, key executives, and engineers left the company. The firm refocused on outdoor markets, like ports and stadiums.
As Vivato's first and later products came to market, another firm was gaining interest: Airgo. Airgo's multiple-in, multiple-out (MIMO) technology--also first reported by Markoff in the Times 10 months after his Vivato piece--would turn out to be cheaper and simpler than Vivato's approach, and, more importantly, the first generation of the chips in products shipped in late 2004 and worked as advertised.
While a single MIMO gateway can't cover an entire floor of a business, a single Vivato gateway can't serve enough users; Vivato's monolithic approach wasn't compatible with the scale of users, purposes like VoIP over Wi-Fi, and the throughput that's now demanded in enterprises. MIMO hasn't penetrated the enterprise yet, but as part of the 802.11n standard, it's the direction to invest in by company IT departments.
Simultaneous with the growing awareness of impending MIMO shipments in 2004 was the maturation of the wireless LAN switch market. WLAN switches, unlike Vivato's beam-forming antenna, could coordinate access points located throughout an enterprise. The first devices generally required a special Layer 2 switch to which the APs had to be directly connected; that difficulty was relatively quickly eliminated in most products by 2004, which then supported Layer 2 tunneling for APs to be located anywhere on a network and controlled centrally.
Cisco's acquisition of Airespace, one of the largest revenue-producing startups, marked the acceptance of that trend into the mainstream, as did Aruba landing the Microsoft contract for their main campus and worldwide offices.
Vivato's approach turned out to be the wrong one from so many angles, although aspects of the first Vivato switch have permeated the market in a more mature, cheaper, and flexible form.
You can read Wi-Fi Networking News's extensive historical coverage.
[Several links via TechDirt, which has followed Vivato closely]
The Wi-Fi-enabled music player technology will appear at the Consumer Electronics Show: I received an in-person demo months ago of MusicGremlin's very interesting technology which keeps a music catalog resident in portable devices that license the technology. Any users near to other MusicGremlin subscribers can exchange music files freely because the files use digital rights management (DRM) to secure playback. Thus users who choose to enable sharing can let other people view their collections. The system can use ad hoc Wi-Fi networking or wireless LANs. It could also work with 3G cellular networks; it's ecumenical.
I've been waiting for the company to announce a hardware partner, as they didn't plan to manufacture their own devices. Apparently, they were tired of waiting themselves and will release their own hardware in the first quarter of 2006.
The company has licensed music via MusicNet, 2,000,000 songs in full. They'll have a variety of options, including a monthly unlimited music subscription, a la carte pricing, per-song downloads, Web-based listening, and preset music channels.
Hawking introduces USB adapter with an 8 dbi antenna: This thing looks very 1950s sci-fi retro, but produces a remarkably strong, directional signal. It has five LEDs that show signal strength, too, which should help improve directionality. It'll retail for about $70.
Now there are two approaches to reaching further with Wi-Fi: brute force and smart antennas. This is the prettiest example of brute force I've seen. But is it better than beamforming and spatial multiplexing in MIMO's finesse?
Indianapolis airport unwired: AT&T (formerly SBC) has installed its FreedomLink service through this airport. Service is included in FreedomLink plans, or $7.95 per day.
New York City looks at municipal Wi-Fi: There's a bill to create a commission that would study affordable broadband for residents. The bill will be voted on Dec. 21. The mayor hasn't said whether he'll sign it.
PC World's latest results on product reliability and service show "average" is the standard for Wi-Fi: The race to the bottom and commoditization of Wi-Fi consumer products are taking their toll. No Wi-Fi maker stood out as above average in any category of reliability that PC World put in front of its survey takers. In struggling to make equipment work in my daily reviewing, testing, and operations, I find reliability to be lacking. Gateways at home and work need to be regularly rebooted even with no configuration changes. Manuals are versions out of date (even in downloadable PDF form) and lack critical terms. Devices don't function quite as promised, but they do function.
A friend recently spent six hours wrestling one day with configuring two devices (an AP and a gateway) from a bestselling maker the name of which shall be withheld. Several hours into it, he checked the firmware. The mail-order firm, also reliable, had shipped him one unit with a 2003 firmware image. Upgrading firmware helped a bit, but the two-unit system kept crashing. Finally, he figured out one problem after a tech support call with a tired sub-continental woman who answered every question with, "Sure, why not?" Setting the AP to "wireless repeater" instead of "wireless client" resolved the problem. His investment? At least 10 hours total.
If you want to be depressed about the state of the electronics industry, read the overall scorecard: No winners in their desktop and wireless gateways category, and sometimes multiple losers on the other side of the equation.
Bluetooth, the bundle of business-card and file exchange protocols wrapped on top of slow short-range radio, is expanding: The group that controls the standard, the Bluetooth SIG, made announcements earlier this year about severing its strong ties to its specific existing radio standard, and pushing for its application stack to migrate to many other radio formats, including ultrawideband. The application stack includes elements like object push, dial-up networking (used for 3G), and faxing.
The announcement today formalizes and pulls together several ongoing agreements, including their intent to work closely with The Wi-Fi Alliance on radio interference, and the Near Field Communications Forum and UWB developers on application-layer standards.
Near-field is a particularly interesting area in which communications only work when devices are touching or in very close proximity, and is being looked at for touchless payment that would avoid card swipes. The folks at Freescale are considering near-field for one of their methods of UWB device pairing.
Back in the Bluetooth radio world, Iogear releases 2.0+EDR dongles: These $40 USB 2.0 adapters support the latest Bluetooth spec, 2.0 with Enhanced Data Rate, which boosts speed from 1 Mbps to 3 Mbps. While this seems trivial, it improves the overall throughput of individual audio and synchronization devices used simultaneously, making it easier to have many Bluetooth items in operation on the same network.
Iogear is shipping both Class 1 and Class 2 adapters. Class 2 adapters claim 20-meter (66-foot) range, which is the distance almost exclusively and erroneously attributed to all Bluetooth equipment. Class 1 adapters have higher-power output and can reach 100 meters (330 feet), according to the spec. Class 2 adapters are $40; Class 1, $50. Iogear's adapters are compatible with Windows as well as Mac OS X (10.3.9 or later).
Motorola and Burton allow business snowboarding meetings: The latest Burton jacket has Motorola technology embedded to take calls over Bluetooth through speakers and mikes built into the jacket. It's also wired to handle an iPod. The jacket's $600. Don't use it while skiing--wink, wink, the company says--but for calls and tune-listening after you've reached the bottom.
We can see that the product puns will just keep coming: Beamforming antenna chip designer, consumer electronics (CE) gateway maker, and muni-wireless customer premises equipment (CPE) developer Ruckus introduces RIOT: Ruckus Interoperability and Open Testing. Ruckus is pushing hard in the home, and this program may let them offer CE makers an alternative or supplement to 802.11e and its Wi-Fi certified form known as WMM.
The idea here is to offer a testing program designed for the home, not all environments. WMM tests for particular characteristics, such as frame bursting (reducing wasted airtime) and packet prioritization into queues tagged by media content. RIOT, on the other hand, will test for interoperability and performance in home-like environments for streaming media.
Ruckus has gained remarkable traction in a short period of time. I'd be tempted to dismiss RIOT as marketing, except for how fast Ruckus has spread into multiple segments and signed up partners.
Brookline, Mass., considers citywide Wi-Fi: Vendors are being contacted. It won't be free, but it will probably push down prices across the town, says a local wireless advocate. And, incidentally, the city would move its Internet access to the new provider. This is often a footnote in municipal plans, but it's one of the reasons telecos--Verizon in this case--are opposed to the networks. Cities and towns pay incumbents billions for telephone and data access each year; a shift in those lucrative, stable contracts could substantially undermine incumbents' revenue over a relatively short span of time. [link via JOHO the Blog]
It's a bit of a shocker when two highly competitive rivals merge: The acquisition is awfully cheap for iPass of their most vocal competitor at $76.5 million, a relatively high premium to their current stock price. iPass has focused on corporate roaming with end-point security through an increasingly full-featured program that handles connections to dial-up, Wi-Fi, Ethernet, and cell networks, but also enforces policies on using firewalls, VPNs, and anti-virus/anti-spyware programs. GoRemote, in the past, has emphasized a three-pronged approach of remote offices, teleworkers, and roaming employees. Both companies have seen stock doldrums as they transition from dial-up to Wi-Fi for access.
The deal must be approved by regulators, but in a highly competitive access business, its unlikely that the GoRemote acquisition would provoke undue scrutiny.
Oh, and iPass also says they have 35,000 live hotspots worldwide. A significant number were integrated into their aggregated network in the second half of 2005, the press release says, include nearly 10,000 from KT in South Korea and 2,500 from Japan Telecom.
The excitement grows as the 160,000-person nears launch: Two weeks ago, a story cited the Phoenix suburb with about 50,000 college students as the 'first" city to go all Wi-Fi, but it's actually the first city of any scale. The next biggest that declare full citywide coverage are focused on public safety or are fairly small, like Chaska, Minn.
This USA Today article notes that NeoReach (using Strix Systems equipment) will install 400 access points to serve the 40-square-mile city, just over a river from Phoenix. The newspaper account accurately described a key point often misinterpreted by non-technical reporters: "The network is strong enough only to be picked up outdoors or through one wall, meaning those who want service in their businesses or homes will need a box that serves as a signal booster and router." That's exactly right. CPEs (customer premise equipment) will be used to bring the signal indoor. The CPE market will explode with the growth of muni-Fi.
While MetroFi has unwired Cupertino (pop. 50,000) and Santa Clara (102,000), they haven't yet promised ubiquitous coverage. A quick check of addresses shows that some aren't served. I'll be paying some site visits in January to networks under construction--maybe including Tempe!
The FCC has moved closer to auctioning four megahertz in the 800 MHz band for air-to-ground telecommunications and data: This auction is being closely watched as it will pave the way for domestic U.S. data and cell calls in the air being a much cheaper method of relaying than via satellite. The order released today [PDF] doesn't set a date, although May 2006 is likely based on conversations I had recently with four interested parties, two or three of which will be bidders on the spectrum.
The executive summary of today's order: Realistically, it pushes back the practical deployment of cell-based voice and any pure data in domestic aircraft to terrestrial stations from mid-2007 to mid-2008, and almost certainly no earlier, unless Verizon Airfone wins all the licenses and is able to complete a transition much faster than it now says is possible. One month ago, the four operators I spoke with (AirCell, Connexion by Boeing, OnAir, and Verizon) were expecting a mid-2007 launch based on a mid-2006 auction. (Verizon Airfone said Monday in a press release that they would be able to launch data services in 2007 if they win a license, but this doesn't conform with the below details.)
The 4 MHz will be auctioned in an odd way, with three potential configurations: two sets of 3 MHz overlapping across 2 MHz in the middle; a set of 1 MHz and 3 MHz; and a set of 3 MHz and 1 MHz. The logic is that two pairs of 1.5 MHz bands are needed to provide reasonable speed. The remaining 1 MHz is left over, and there's no provision for a winning bidder of 1 MHz to build out service. Companies can bid on all kinds of pieces, and the optimum dollars for a single configuration will win out.
Verizon Airfone currently occupies all 4 MHz for its underused phone service and will have two years following the auctions to move its service down to 1 MHz which may overlap with a winning bidder. They may win the bidding, and this order today requires more monitoring if they do so. They received a five-year license renewal running until 2010 for their phone service.
AirCell, a potential bidder and existing operator of a U.S.-wide network of ground stations used for general (non-commercial) aviation, filed several requests for changes to Airfone's incumbent terms. Among them were a request to reduce the transition from 4 to 1 MHz to six months (or even one year, they suggested later) instead of a full two years. This was denied, and it's a shame, because it means that any winner of 3 MHz in any of the three configuration options will be unable to deploy for as long as two years following the auction's completion unless Verizon chooses and is capable of migrating faster. The order notes that Verizon has to touch and test every ground station and about 3,000 commercial and government planes--there's no possibly of remote upgrades.
AirCell also wanted Verizon's license to use 1 MHz dropped from five years down to two, asserting that this gives them an extra benefit in operating in the 1 MHz longer to detriment of whichever winner bidder (if not Airfone) has to share that 1 MHz of spectrum in a non-interferring manner. This was denied, too.
Space Data, a stratospheric tethered balloon broadband firm, won a decision that allows it to use the air-to-ground frequencies on its balloons, but lost the rights to use the frequencies for secondary purposes unrelated to air-to-ground transmissions.
And yet another decision went against AirCell's desire, too. The FCC commonly offers bidding credits to what they define as "small" and "very small" businesses, based on the average three years' preceding revenues. AirCell and Space Data had fought for 35 percent credit for very small and 25 percent credit for small businesses. Boeing--which operates its satellite-based Connexion service internationally--thought 25 and 15 were enough; Airfone wanted none. The order specifies 25 and 15 percent credits for very small and small businesses they define as averaging $15 million over the previous three years and $40 million, respectively. It's unclear where AirCell and Space Data fall as both are privately held and don't appear to report revenue.
Interesting, two of the three sitting FCC commissioners--the temporarily majority-wielding Democrats Copps and Adelstein--both used the opportunity to concur (which means disagree with in this context) in part with the order because they dislike the monopoly it continues to grant Verizon Airfone by implication. Adelstein wrote [PDF], "Ultimately, we could have taken a number of more specific actions to support competition in the event Airfone wins the exclusive three megahertz license. But we fail to do so today." Copps wrote [PDF], "I remain concerned that America's aviation industry and its passengers will not have the full range of choices in air-to-ground broadband that they might otherwise have enjoyed."
A lot of folks have been eyeing Palm's TX model, with Bluetooth and Wi-Fi included: The list $300 device is long on features, short on memory, but a good all-around handheld in my testing. My favorite "deals" site, dealmac.com--covering much more than Mac products--has ferreted out a Buy.com discount that puts a $270 price tag on the Palm TX including shipping, and then cuts it by $15 off for new Buy.com customers.
The satellite operator launched its Broadband Global Area Network (BGAN) yesterday, IDG News Service reports: The company has two birds in the sky that use beamforming antennas to allow focusing bandwidth in more discrete areas and relocating bandwidth on demand. The 4G satellites will eventually provide OnAir's in-flight broadband service as well.
The BGAN service will offer 492 Kbps speeds and meld voice and data simultaneously, which the company claims is unique for wireless offerings. Calls will cost about US$1 per minute; data, €4 to €7 per megabyte. These are initial prices Inmarsat expects will fall as resellers roll out plans. A terminal will run $1,500 to $3,000 and be laptop-sized for portability.
North and South America officially lack coverage at the moment; they're waiting for the third launch in spring 2006.
Will WiMedia get the edge on ultrawideband (UWB) by having a fast track to ISO standards approval?: The Ecma International group released a UWB standard based on what's called the WiMedia UWB Common Radio Platform. This version of UWB was formerly under the auspices of the Intel-led Multi-Band OFDM Alliance (MBOA), which merged its physical layer tech into the WiMedia Alliance, formerly focused just on higher-level application and network protocols.
Ecma will get their version of the standard fast-tracked within ISO, which gives it--I believe--a real leg up in the process. Ecma has fast-tracked a number of other major standards in the past. One of the stumbling blocks for UWB has been international regulation, and some regulatory bodies have been in a holding pattern awaiting more standardization on UWB. If the ISO adopts the Ecma standard, that might be a green light in 2006 to finalize approval in many parts of the world based on where I think the regulatory recommendations stand now.
It's possible that Motorola and Freescale, backers of Freescale's original and classic approach to UWB, can stall the project at the ISO standards group. But the ISO has less of the aspects of IEEE consensus building that have led to the stalemate in 802.15.3a over UWB for Personal Area Networks.
Peter Judge has some great reporting on the issue, including a strong reaction from Freescale, over at Techworld. Freescale has the only shipping UWB silicon and owns key basic patents on the space. They've been involved in the long march from idea to production over a long period. They're notably peeved that other firms have radically changed the notion of UWB and that almost all companies in and around the wireless and wired networking space are backing WiMedia. (Not all--just almost all.)
Nintendo says 200,000 users have checked in at its Wi-Fi Connection site worldwide: This in just three weeks since launching with Mario Kart DS designed for Internet and LAN gaming over Wi-Fi.
Hotspot and access point aggregated management software company Sputnik expands, updates its product line: The company specializes in providing a centralized console that allows management and reporting across a network of Wi-Fi access points, whether for academia, hotspot networks, hotzones, or companies.
Sputnik Server 110 is a 1U rack-mounted server pre-loaded with 10 AP licenses and the Control Center software for $2,699; additional licenses can be purchased. The company's new AP 210 ($279) has a 285 milliwatt transceiver and the 260 ($399) has two such radios. They're designed for extended coverage, and can handle, the company says, point-to-point links of up to two miles. They have the nice feature of keeping traffic isolated, so that users on the network can't turn on promiscuous mode to examine other users' data.
They also released a Linksys WRT54GL firmware image which allows the new Linux-based model (an old model renumbered and sold at a higher price) to run the Sputnik Agent software. The firmware works on older WRT54Gs and all models of WRT54GS. They'll sell you preflashed WRT54GLs for $99 and WRT54GSs for $109.
Qualcomm now required to smile when it says Wi-Fi stinks: That's a comic distortion intended for effect. Qualcomm uses Wi-Fi for its corporate Internet access, and merely thinks it's inappropriate given its unlicensed nature and design intent to operate on a metropolitan basis. Rather, to pick an example out of thin air, CDMA2000 1xEV-DO might be a better.
Regardless of MAN/LAN debates over Wi-Fi's use, Qualcomm joining the alliance is recognition that handset makers and cellular operators worldwide want unlicensed mobile access (UMA) in their phones, among other uses of Wi-Fi. Qualcomm's participation may make it easier for compatibility and interoperability of standards.
Broadcom has declined so far to provide any non-licensed access to its Wi-Fi chips: A project that has been working to reverse engineer access using legal means has released its first working drivers for Broadcom 4300 series chips. The project requires the use of the SoftMAC software as well to compile working drivers within Linux. The first successful use was documented in email Dec. 4 to the developer's mailing sent from a PowerBook running Linux with the project's drivers installed.
Atheros has allowed a third party to create a layer between the low-level functions of its chips and high-level drivers. The madwifi Hardware Abstraction Layer (HAL) prevents developers from having access to most of the radio functionality, which would might allow use of frequencies that aren't legal in particular countries, use of encodings that aren't allowed, and other regulatory problems.
The Economist magazine ran an article early this year critiquing the timidity of Atheros and Broadcom, noting that "if the firms are really worried, they could release most of the interface, keeping back those features that are legally sensitive." Neither Atheros nor Broadcom speak much publicly about this matter. [Link via Jim Thompson]
WiQuest keeps pushing the ultrawideband (UWB) envelope: It's latest chips can top a gigabit per second. This linked mainstream article suggests it's "turbo-charged Wi-Fi," but UWB only works at high speeds over very short distances, making it a great cable replacement for FireWire and USB 2.0, among other wired standards. The WiQuest technology is ready for sampling, the company says, but it doesn't conform to any current standard.
A technical piece at EE Times provides a bit more detail, including the fact that competitor Freescale is complimentary of WiQuest pushing the envelope. Freescale expects 660 Mbps chips by early next year. The 1 Gbps rate works just within five to six meters, good enough for some home-entertainment purposes and computer peripherals. The first version of the chips are in reference designs for USB 2.0 replacement, with both host and adapter designs for complete integration.
The Wi-Fi Alliance will certify a power-saving mode for battery-operated handhelds: The Power Save mode is an addition to Wireless Multimedia (WMM) certification, and is intended to reduce or remove power-wasting elements of the Wi-Fi standard less appropriate to VoWLAN (VoIP over wireless LAN) handsets as well as handheld devices. A number of chipsets have already been certified under this process, which can save 15 to 40 percent of battery usage.
Chips from Atheros, Conexant, Broadcom, and Cisco, among several others, will get the nod, although there will be no separate product labeling. It's more of a comparison feature for enterprise buyers than a starburst shrinkwrap box selling point. In an interview a few weeks ago with Frank Hanzlit, managing director of the Wi-Fi Alliance, he said that there would likely be more certifications over time, such as WMM itself, that would not be required but would be important points for purchasing managers in planning deployments.
Three organizations have teamed up to create a standard database of wireless threats: Every industry now seems to have its own threats database, and there are good reasons to standardize on names, behavior, and vendor responses. CERT started over 17 years ago to help assess Internet vulnerabilities; with years of wireless attacks in the wild, the Wireless Vulnerabilities and Exploits (WVE) site is overdue.
Network Chemistry, a wireless network security firm, is one of three sponsors, along with training and certification firm CWNP, and the Center for Advanced Defense studies. The editorial board includes highly credible members. The site perhaps softlaunched earlier, but the formal press release went out today.
One of the points of the WVE, like similar codifications of threats, is that a discrete number assigned to a particular exploit or vulnerability means that it can be referred to across the literature without duplication. This makes it easier for vendors and open-source projects alike to check off specific known problems from their lists, and for certification processes that involve reducing risk to have a testbed as well.
For instance, if I want to talk about weak initialization vectors (IVs), a flaw that made the broken WEP encryption standard even more broken than it was, I'd write about WVE-2005-0021. One flaw in the organization of the site is that the URL should allow a WVE number to be used after the URL; instead, the numbers in the database are arbitrary and don't match the WVE numbers.
Aurora, Illinois's aldermen discussed some of the realities of running municipal wireless networks: They looked at Indiana University, which has a similar population (across two campuses) to Aurora; I've interviewed IU's wireless IT manager, and they have an excellent infrastructure. But the aldermen note that IU spends $250,000 per year in support costs. This number is misleading because the WLAN at Indiana is just a part of the overall operations, and not the network over which most data is transferred.
Arlington, Va., wants 10 areas served at no cost: The city proposes that a private firm build out free, lower-speed access for residents and municipal purposes in 10 areas and town, and charge fees for higher speeds or sell ads. The article has the first (and possibly last) mention of a municipally focused broadband firm that's shutting down: "Leesburg-based HighSpeed America began operating municipally backed wireless Internet networks in January. The company is shutting down after less than a year." Never hoid of 'em.
Wilkes-Barre, Penn., will have competing hotzones on top of citywide Wi-Fi: PenTeleData, a regional company, will build Wi-Fi zones across two colleges and nearby housing. They bid on the citywide offering, but didn't think the city's financial terms made sense. PenTeleData has the operator attitude towards Wi-Fi, that it's a mobile technology suited for use in particular environments; the city wants to sell the idea that it's a dial-up replacement or very low-speed DSL. There's a misstatement in the article because the writer is looking at raw Wi-Fi speeds: "Wi-Fi can’t match the speed of Internet access piggybacked over cable lines, but it is generally faster than DSL access over phone lines." Wi-Fi designed for campus-wide coverage typically can beat DSL in particular places, but Wi-Fi designed for municipalities will likely be one-quarter to one-tenth as fast as prevailing ADSL speeds on the download side.
Town says, company says: The city of New Orleans says that BellSouth withdrew an offer to donate a building with 250,000 usable square feet because of the city's plan to build a municipal wireless network. BellSouth disputes this.
MetroFi abruptly changes its municipal broadband wireless model: MetroFi came before other city-wide Wi-Fi efforts, requiring little city cooperation beyond leasing pole space and no bidding. The entrance of companies into the space that will offer free access to residents and visitors--such as Google's accepted plan for Mountain View and proposal for San Francisco--has apparently sent MetroFi into a new mode. They've unwired Santa Clara and Cupertino, offering 1 Mbps down and 256 Kbps up, charging $19.95 per month.
Their Sunnyvale launch--already serving about a third of the households--will require that users keep an ad banner up at all times (a half-inch strip) to receive free service. If it's successful, they'll roll out free, advertising-supported service to their original two cities, too.
While the article cites the company's CEO saying that this is a test, a visit to MetroFi's home page reveals their home page stating that their service is free, footnoted that it's only free in Sunnyvale. Doesn't feel like a test. Feels like tectonic plates moving underfoot.
Update: Via email, CEO Chuck Haas commented on the plan, noting that "The advertiser revenue will greatly augment and could well exceed the revenue from our existing $19.95 paid subscriber service. When you look at the alternatives for local advertisers, MetroFi is an affordable and
powerful way for local businesses to increase revenue."
Google thinks so, too, with local ads representing about 40 percent of domestic ad spending.
Network Computing turns in a long feature looking at tools that help troubleshoot and monitor Wi-Fi networks' health by examining the radio frequency space they operate in: The monitoring space is very hot; I'm asked all the time to be briefed on the latest product or product revision. I've demurred because Wi-Fi Networking News lacks an RF testing environment. The folks at Network Computing show how to examine the field. They look at a dozen products from several manufacturers, only some of which directly overlap by function.
The reviews sort products into spectrum analyzers, calibration tools (just a single entry), wireless protocol analyzers, site surveying tools, and performance and security evaluators. Spectrum analyzers look at the RF patterns and are typically intended for more sophisticated users and uses; protocol analyzers deconstruct data into known patterns to see how the airspace is being used by data devices. A spectrum tool can show if a microwave oven is disrupting your network or an errant cordless phone, while a protocol tool will let you see how 15 networks on the same channel in a small space are producing throughput degradation.
A big caveat in the protocol analysis category is how well the products can decrypt packets when you have the network keys, and how they deal with authentication and dynamically generated Wi-Fi encryption keys.
To cap off the article, there's a sidebar on free tools that can perform a subset of the functions covered in the feature.
Akron will roll out 62 square miles of Wi-Fi/WiMax: They'll mesh Wi-Fi first, and then add WiMax later. The intent is to mix voice, data, and video over the network, but not be tied to Wi-Fi exclusively. There will be both free and for-fee services. The city will have NeoReach Wireless (part of MobilePro) build and run the network using Strix Systems gear, which has upgradable radio slots. That would indicate that backhaul could start with Wi-Fi and move to WiMax with a field ugprade.
Yellowstone Wi-Fi blankets Wyoming city: Cody has Wi-Fi across many parts of the city. The company providing it, TCT West, markets it as Yellowstone Wi-Fi and pays the city a utility pole fee along with power costs. The system cost $100,000 to set up, and the firm charges $6 an hour up to $30 per month. Tourists are expected to drive revenue. TCT West also provides fixed wireless Internet access to residents. Another firm in the nearby town of Powell may provide a similar service there.
IT Architect rounds up the state of mesh networking: Six vendors are in the space, Andy Dornan writes: BelAir, Cisco, Firetide, Motorola (mesh division), Tropos, and Strix. He divides them into four categories: metro-area (Tropos and BelAir), indoor with newer outdoor options (Firetide and Strix), proprietary client (Motorola), and central management (Cisco).
Dornan also divides up the companies by how many radios they use. With one radio, such as in Tropos's nodes, the number of hops to backhaul is limited, but it does keep per-unit costs lower. More radios means fewer locations for backhaul, as with units offered by Strix or Motorola. (Motorola's Canopy division, of course, specializes in backhaul, and EarthLink's municipal network design will use clusters of Tropos nodes in mesh formation backhauled via Canopy transceivers to aggregate backhaul to central points.)
The lack of a standard means no interoperability, but 802.11s is underway at the IEEE and may result in indoor interoperability. It's possible 802.11s will help just at the CPE (customer premises equipment) level rather than intranodally, too.
The Linux flavor of the WRT54G is still in production, despite newer VxWorks-based version: Good news for the community of hackers, developers, and experimenters who rely on the Linksys WRT54G to power their projects: while the product sold under this model number will no longer use the Linux operating system as its basis, Linksys has created an offshoot model that will continue down the Linux path under the name WRT54GL.
The WRT54G in versions 1 through 4 used free and open-source software components for its embedded operating system and most of its functions. After some negotiation a few years ago, Cisco's Linksys division released all the code they had modified as they were ostensibly required to under the terms of the GPL (General Public License) and other software licenses for components they had used. Wireless hackers immediately figured out how to replace the baked-in firmware with their own firmware images to take advantage of the WRT54G's low cost and high availability. This appeared to end with the November 2005 release of v5, which switched to a proprietary OS while halving the RAM in the unit.
The new Linksys WRT54GL, however, is the v4 version of WRT54G stabilized on Linux. It will list for $79 MSRP (Amazon.com has it in stock for $71) and Linksys expects to sell 10,000 of them per month. The firmware for the WRT54GL (version 4.30.0) is already up on the Linksys GPL Code Center. Some folks on Slashdot have complained about the price difference between this perpetuated Linux model and the cheaper proprietary OS WRT54G--but that's how the market economy works. With 120,000 potential sales per year or $8,500,000 in gross retail sales, a competitor could engineer a better box for that market.
In an interview with Linksys, they told me that they had this model in the works for some time, but didn't want to tip their hand until shipping product was available in the channel. Company spokespeople emphasized that while the "GL" model can be flashed, they don't support any firmware but what Linksys officially releases. Still, they've kept a Linux model because the market is there for it. Linksys confirmed what some analysis firms have been saying: to their knowledge, the WRT54G is the single bestselling model of Wi-Fi gateway in the world. I'm not sure what model gets the No. 2 nod.
Peter Rysavy wrote in Network Computing today about his disappointment with the disappearance of the Linux-based WRT54G--he didn't get feedback from Linksys before he filed his piece--and what he sees as the lower-stability of the licensed VxWorks operating system that underlies the v5 release of the gateway. (I've heard this from others, including a colleague as recently as yesterday.) VxWorks is a closed-source system that Linksys told LinuxDevices.com uses half the RAM of the embedded Linux OS. This lets them conserve costs in a commoditized market in which they sell hundreds of thousands of WRT54Gs per month--every dollar saved is millions a year recouped.
Linksys, by the way, didn't develop their embedded Linux, but in turn licensed it from Broadcom, who created the Linux-based code for its reference design. Reference designs are built by chipmakers to provide a ready-to-go product that incorporates their chips. They can often be bid out and built without much effort beyond designing a case and customizing or replacing an installation wizard.