Receive new posts as email.
RSS 0.91 | RSS 2.0
RDF | Atom
Podcast only feed (RSS 2.0 format)
Get an RSS reader
Get a Podcast receiver
| Sun | Mon | Tues | Wed | Thurs | Fri | Sat |
|---|---|---|---|---|---|---|
| 1 | 2 | 3 | 4 | 5 | 6 | |
| 7 | 8 | 9 | 10 | 11 | 12 | 13 |
| 14 | 15 | 16 | 17 | 18 | 19 | 20 |
| 21 | 22 | 23 | 24 | 25 | 26 | 27 |
| 28 | 29 | 30 | 31 |
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 or JiWire, Inc.
Entire site and all contents except otherwise noted © Copyright 2001-2006 by Glenn Fleishman. Some images ©2006 Jupiterimages Corporation. All rights reserved. Please contact us for reprint rights. Linking is, of course, free and encouraged.
Powered by
Movable Type
« March 2005 | Main | May 2005 »
L.A. mayor releases committee report on filling gaps in broadband coverage (PDF): The committee includes a few academics and the rest are incumbent telco and cable company folk. That’s why the report’s tone is quite extraordinary. It mostly evenhandedly lays out the scope of the problem, looks into what approaches might be taken to solve it (including all the models out there), and reaches conclusions that aren’t unreasonable.
Pro-municipal advocates won’t find the report anything to cheer about except the scope of the problem as stated, but I don’t find anything misrepresented—except the definition of Metcalfe’s Law.
Posted by Glennf at 2:44 PM | Comments (0) | TrackBack
The summit includes the city, non-profits, and Michael Oh: You can’t swing a Yagi antenna in Boston without hitting something that Michael Oh is working on. This summit is due in part to his efforts along with others interested in expanding wireless Internet access. The agenda is to figure out how wireless technology could make Boston better without, it seems, any particular hook like “free wireless for everyone.” Meeting details are on Boston Councilor John Tobin’s Web site.
Posted by Glennf at 4:24 PM | Comments (1) | TrackBack
What never seemed like a good idea has been dumped: Verizon will “phase out” its Wi-Fi hotspots in New York which were attached to payphones. Verizon never integrated these hotspots into anyone else’s network, which means that only their DSL subscribers had access. This is a sure strategy for disappointment, of course, because the idea of wandering around to find free access at phone booths was always bizarre. If Verizon looks to this Wi-Fi network as proof that Wi-Fi hotspots are a failed strategy, it explains some of their very bizarre public statements on how Wi-Fi works.
Posted by Glennf at 4:20 PM | Comments (1) | TrackBack
It’s not clear whether “open-source” means boot our code in this scenario: Aruba has released its bootloader, a method by which an access point with the right hardware can load Aruba’s AP code when detected by its central WLAN switch on a network. That’s all well and good, but it doesn’t bring much to the table—yet. Aruba promises more. In this article at Linux Pipeline, I examine the promise of open source for Aruba and the industry, and get a little into the issue of the latest proposal for WLAN switch AP interoperability.
Posted by Glennf at 8:08 AM | Comments (0) | TrackBack
Tropos announced its next generation of gear today, encompassing 802.11g speed: The new series of mesh Wi-Fi routers offers the higher speeds found in 802.11g, as well as support for a new mesh operating system and management console that offers more flexibility and support for security. (They’re using Atheros chips, by the way.)
Tropos estimates that with $68,000 of their latest gear, one typical city square mile could offer 10 to 15 Mbps per user concurrently. They estimate a sevenfold higher cost for this throughput from other vendors; I’ll be curious to see competitive math on that.
The Tropos 3210, an indoor mesh router, is shipping now, along with the 5210 outdoor router. The 4210 mobile router should be available by third quarter. The 4210 is designed for ad hoc access from non-fixed but static locations, according to an earlier press release.
The new management and operating system includes intra-router AES encryption support, which should answer one of the concerns I’ve heard raised about outdoor mesh Wi-Fi: with AES, there’s virtually no opportunity to tap into communications between mesh nodes.
Now, there’s definitely an issue of sniffing other users’ traffic at any given node, however. The new system offers supports for up to 4095 VLANs and 16 ESSIDs, along with 802.1X and WPA support. This should make it much easier for metropolitan-scale networks to standardize on using 802.1X even for residential users connections to the network.
Take a Wi-Fi bridge with 802.1X client support as a CPE and set it up as the gateway using 802.1X with WPA Enterprise to a node using an ESSID reserved for that purpose and market it as secure access. Or support individual computer access using free built-in 802.1X software or Funk or Meetinghouse’s $40-$50 clients for other platforms.
Other users can connect via insecure links but can be heavily warned about using SSL email and other security mechanisms. It can be a service differentiator.
Posted by Glennf at 11:11 AM | Comments (0) | TrackBack
Two weeks ago, the Progress and Freedom Foundation released two reports critical of Philadelphia’s plan: Adam Thierer wrote one of these reports called “Risky Business: Philadelphia’s Plan for Providing Wi-Fi Service.” Based on articles from earlier in the year, the report was initially to be release by the Cato Institute, with which Thierer was associated. It was also drafted before Philadelphia’s business plan and RFP was issued. The report was updated and released after Philly’s plan—Thierer might have gotten advance copies as some people did—but it still bears many of the marks of ideas that were distributed about Philadelphia’s plan that aren’t actually in the plan.
When it first came out, I labeled the report more work from sock puppets. But I’m wrong: there’s a fair amount of original thought in this. I don’t agree with some of the assumptions made, but Thierer does spell out those assumptions which makes it easier to engage in a dialog about whether those assumptions are valid. His colleague Patrick Ross at PFF also pointed out that PFF has a variety of funders some of whom are backers of municipal networks, like Intel, which has taken a generally favorable stance, and may eventually lobby for it.
Fundamentally, I believe I come down on the idea that reducing corporate taxation and providing regulatory relief are arithmetically similar in my political philosophy to increasing the tax burden on individual taxpayers. Perhaps some of you can use that to pin a label on me, but I’m not sure one fits.
Interestingly, many of the new plans for broadband wireless are taking the tax issue into account and are requiring that none of the funding for the networks comes from taxpayer dollars. Offsetting the risk and putting money into the private sector takes some, but certainly not all, of the wind out of the sails of those who want entirely private sector competition.
One of the key issues I agree with Thierer on is that an underreported aspect of many municipal plans—including those that aren’t disputed by incumbents or others—is that municipal facilities like light poles, conduits, and buildings are being made available at no cost to the entities that will build the networks, whether private or public. That no-cost basis actually has a price tag attached that prevents competitive entry into the market on equal terms. This is true in Philly’s plan and in Minneapolis.
A state representative asked me a few weeks ago: what would keep SBC from coming into a city in Texas that decides to build its own municipal Wi-Fi network (on whatever basis)? I said, nothing. If they’re using unlicensed spectrum, no one can really stop them. But the access to facilities for putting in access points is a killer: if SBC has to license for every private rooftop, like cell carriers do, and face opposition or stalling tactics from cities and towns on using poles and other utility facilities, then SBC can’t build a network at the same cost or perhaps even one that has the same scope.
I have lengthy comments, and thus provide them after the jump. Download Thierer’s report and then follow along. I’ve quoted small passages and provided page numbers for context.
Page 2:
"Moreover, there are serious fairness issues raised by Philadelphia's entry into the broadband marketplace. First, there is the question of how fair it is to other competitors in the field when governments gets involved in the provision of service. In particular, when government becomes a market participant, it can have a "crowding out" effect on private sector competition, innovation, and investment."
This is often stated, but I haven't seen concrete examples of long-standing networks driving out competition. In Tacoma, Washington, Tacoma Power's Click! network was established partly because of a failure of two incumbents (then TCI and US West) to offer a plan in the mid-1990s for Tacoma's infrastructure. Now, there is robust competition and choice among Comcast, Qwest, and Click! for broadband and between Comcast and Click! for cable. Click! itself sells broadband on a wholesale basis through three private residential ISPs and five private business ISPs. A Comcast spokesperson was quoted two years ago praising the competitive uplift that Click! provided to the city.
The definition of competition is extremely important in this context as well as a potential strawman. Many cities are building broadband when no robust infrastructure exists but incumbents want to preserve the future right to offer that infrastructure at a time and price they determine on a de facto monopoly basis. Competition has to be defined in reference to a body: the consumer or taxpayer, or the company and government. Later, we'll get into the taxpayer side of this equation.
"...if Philadelphia's bet doesn't work out, the costs for local taxpayers and consumers could be enormous. There is also the question of what else could be done with the money Philadelphia is proposing to spend to build its own public high-speed networks. Today's projected expenditures are also unlikely to account for the ongoing technological disruptions we can expect in the future."
Wireless technology has only become enormously cheaper year over year. WiMax is part of the Philly RFP. Philly includes $4.5 million in capital spending for year 2 to 5 to address technology change.
"In sum, in their rush to ensure that local residents are provided with affordable high-speed access, city officials may be doing them a great disservice by crowding out private sectors investments that do not require public bailouts should things go wrong. If private actors are willing to assume the risk, and expend the significant sums it takes to deploy high-speed networks, it makes little sense for local governments to intervene and force citizens to assume those risks instead."
This may have been written before Philadelphia released its plan and Minneapolis announced its approach. In both cases, risk is offset to private lenders and hands-length transactions.
"Before taking such an extreme step, lawmakers can work to reform regulations and tax policies..."
This gets to the heart of many of the anti-municipal broadband bills in state legislatures. Typically, the real issue is removing expensive universal service requirements and adding tax advantages. The viewpoint of this writer and most conservatives is that reducing taxation on companies does not equate to putting more of a burden on taxpayers because the companies plough the additional revenue they have into more investment which can result in reduced costs and improved services as well as a more robust business climate. That's a big assumption, but I don't take a particular stance. I would note that reducing fees that companies pay through regulation and taxation is the other side of taxpayer risk.
Page 3
" Also, means-tested financial assistance can be delivered directly to needy households should city officials feel those citizens cannot afford to pay the going market rate for broadband services. This has the advantage of not locking-in Philadelphia to any specific broadband technology. Instead, residents could use the equivalent of "broadband vouchers" to shop for service on their own from competing providers."
But that results in public and philathropic subsidy of private companies--transferring taxpayer money to private companies to pay for lower-income Internet access. This is precisely what I thought this report was coming out against: using taxpayer dollars for the benefit of a subset of citizens.
Page 4:
"The city would not directly manage this new utility, rather, the nonprofit corporation would oversee all operations and generate its startup funding from foundation grants, bank loans and other sources. At least in theory, therefore, the city would not be required to kick-in any of the initial funding for the project."
It's not in theory: it's stated explicitly in the plan and Mayor Street appears to have bet part of his political career on the city not providing a dollar in direct funds.
"In essence, this means the city would agree to give most (if not all) of its communications business to the non-profit corporation and grant it preferential access to publicly owned infrastructure that private firms would not have."
I agree that this is a significant problem. Minneapolis' plan has the same element although a private company will get this preferential treatment.
This is a method of creating a legal franchise for public infrastructure that's not available to all. It means that the non-profit has to act as the most non-discriminatory wholesale provider and keep up on technology to earn its right to hold that de facto monopoly franchise.
"Is Broadband a Public Good or a Natural Monopoly"
There's a whole set of economic theories here that I won't pretend to be qualified to address, and my interest is primarily in the technological and political implications. So without expressing support or disdain for this section, I leave commentary on it open to those with a richer background in this area than myself.
Page 6:
"* Of the 32.5 million "high-speed lines," 23.5 million provided "advanced services," (services with speeds exceeding 200 kbps in both directions)."
This definition has always been a sop to telcos who wanted to offer 256K DSL and have it deemed broadband. Barely, barely broadband.
"* Subscribers to high-speed services were reported in 94% of the nation's zip codes. In 81% of the nation's zip codes more than one provider reported having subscribers. And 99% of the country's population lives in the 94% of zip codes where a provider reports having at least one high-speed service subscriber."
Specious measure also used in other reports. Zip codes were originally assigned and continue to be added based on population density. They're related to carrier routes. Using Zip codes as a basis of DSL deployment doesn't say much. Texas, for instance, has dozens of counties in which DSL or cable modem service is available in a very tiny area in town centers. In some counties, it's not available at all. Hundreds of square miles might be encompassed by the least populated areas. And there's a post office in the middle of that. I'd like to see this measure dropped. It's like saying that 95 percent of the population lives within two hours of an international airport. What does that tell you about airport location versus where population lives?
Page 7:
" 'the U.S. already teems with wi-fi access,' notes Manhattan Institute telecom scholar Thomas Hazlett."
Hazlett's essay was down on municipally run Wi-Fi networks, but it was also clear that spectrum policy management failure is at the root of dissatisfaction with competition. Hazlett argue that cell carriers, denied enough spectrum to have true 3G, cannot be the robust 2nd or 3rd competitor in markets where that would actually produce real private-sector competition. Hazlett is opposed to government-run broadband, but it's also tricky to see when cell carriers will have enough spectrum to be effective national competitors for broadband.
Page 8
"Various terrestrial and satellite-delivered wireless services have also been proposed in recent years. Wi-fi is clearly the trailblazer on the terrestrial wireless side."
A weak point in this report is that this section should be much longer. One of the most effective counter-arguments to municipal broadband is that it's likely that a combination of licensed and unlicensed private sector networks will appear over the next one to three years that will offer true metropolitan-area point-to-point wireless broadband. There are still issues (see Hazlett comment earlier) about spectrum availability, but there's a lot of innovation. The biggest competitor to my eyes to a Philly-like network isn't wireline, but rather a Sprint/Nextel or Clearwire network or a company like TowerStream ramping up to be a business and residential carrier competitor.
Page 10:
"Even an investment in wi-fi along the lines of what Philadelphia is proposing, is a risky roll of the dice. And this is where the difference between private and public enterprise becomes crucial."
I find this statement fairly tenuous, but not specious. Wi-Fi will likely be replaced at some point in the future by something else. But the next generation of Wi-Fi (802.11n, due to be ratified by 2007, but with robust versions on the market by mid-2006) will extend its life by several years. There's no suggestion in the marketplace that the multi-billion-dollar investment in residential, commercial, and hotspot deployment can turn on a dime. Because Wi-Fi uses unlicensed spectrum and because its cost has dropped so far so fast (with future drops expected as it becomes integrated), Wi-Fi is not a bad bet for the five-year window of the network.
With $4.5 million in year 2 to 5 capital expense budgeted in Philly's plan, this would allow them in year 3 or 4 to replace all of the infrastructure without incurring expense above what they've budgeted. Some of the largest upfront expense is administrative and logical: setting up the network's parameters, billing, etc. And setting up power and other connections for the nodes of the network. If the raw cost of the network is $10 million today, replacing 100 percent of it should be $2 million in three to four years partly because you'll need less equipment to serve larger areas (higher throughput, better reflection absorption, lower costs) and partly because the fundamental setup was done in year 1.
Page 11
"Experience of Other Cities Does Not Bode Well for Philadelphia"
Most of the examples below have been thoroughly debunked. The references are to reports as far back as 1997 before systems were fully built. The Beacon Hill report cited relies on numbers from 2001 and earlier that didn't reflect the state of the systems in March 2004 when that report was released, and which misinterpret a number of publicly available financial details. (Beacon Hill didn't disclose in the report their funding by the New England Cable Telecommunications Association, either.)
The Free Press report from April 2005 is the most extensive; anyone who wants to cite Marietta, Tacoma, etc., needs to read that report first and respond on-point to its reporting:
http://www.freepress.net/docs/mb_telco_lies.pdf
Page 12:
"· Federal and state tax exemptions;
· Exemptions from various other taxes such as local property taxes, excise taxes,
franchising fees, or gross receipt ta xes;
· Tax-exempt debt securities;
· Low-interest government loans or loan guarantees."
Interestingly, the debate on these issues has shifted since the NMRC report was issued earlier this year. Both Philly and Minneapolis's plan would have these problems. Philly would create a non-profit that would be exempt from certain taxes like any other non-profit, but wouldn't have all the municipal tax benefits described.
Minneapolis's plan puts all risk and taxation on the shoulders of the private contractors who will build and operate their network, thus obtaining none of these advantages.
I haven't seen the anti-muni reaction to Minneapolis's plan yet.
"· Free and unfettered access to rights-of-way, easements, lampposts, water towers, and public buildings (including schools and libraries);"
Municipalities that outsource their networks may still be offering this advantage and at no cost to private or non-profit entities that build the network.
"· Exemptions from rate or price controls that private firms must live with."
Most broadband in most states isn't subject to price controls except in those states that require services to be sold on a non-predatory basis.
Page 13:
"In fact, the business plan assumes rapidly declining capital expenditures after Year 1. After the initial $10 million investment in Year 1, the plan posits that capital expenditures will plummet to just $200,000 in Year 2, $100,000 in both Years 3 and 4, and then fall to zero by Year 5."
This is confusing. I queried someone close to the plan who confirmed that the $4 million listed for gross capital reserve is part of the overall capital budget. The $500,000 figure is intended to come from initial funding while the $4 million comes from cash flow.
Page 14:
"Outsourcing the start-up costs will help insulate the city from some of the costs should things turn sour for the venture. Then again, should that occur, it seems unlikely that the city would just let the operation whither away."
This is just speculation: the city isn't just outsourcing, it's removing its control. If the network goes south, I could equally speculate that it would be political suicide for a city hall official to pick up the tab.
"The Wireless Philadelphia blueprint makes this clear when it notes many times that service rates are to be held below the prices charged by private sector competitors, and that any revenues generated by the network will be funneled to other programs of the city's choosing."
The non-profit will be charted with certain expectations, but once it's in operation, the city will have substantially less control of it than this implies. It will be a legally distinct institution.
"What this means is that, should things go wrong, the city (and, therefore, taxpayers) will be forced to pick up the pieces."
This is still speculation. My guess is that this part of the report was written before the city announced its plan, so some segue text was introduced to make the argument that what follows would make sense.
Page 15:
"After all, if the city is guaranteeing wi-fi access to the masses as the equivalent of a birthright entitlement, many citizens might come to expect that other services will be subsidized as well..."
That's a remarkable leap in logic.
"Again, capital expenditures would fall to merely $200,000 in Year 2, $100,000 in Years 3 and 4, and then zero in Year 5."
See earlier: this is incorrect, but a justifiably easy mistake to make given how the business plan was written.
Page 16:
"The blueprint does contain a short technical analysis of potential future wireless networking technologies, such as WiMax, but says very little about what sort of investments might be necessary to ensure those technologies are put into place if they come to supersede current generation wi-fi networks."
Actually, the plan states that WiMax should be considered as part of a proposal. WiMax could likely be used for backhaul and then gradually rolled out for point-to-point use as needed. Pretty simple.
"If things do not go well in the near term and the Wireless Philadelphia's initial $10 million investment (and whatever else they have plowed into it in subsequent years) turns sour, then the potential for a taxpayer bailout enters the picture. Obviously, should this occur, local residents will be left footing the bill for the failed experiment and they will lose financial resources that could be used for other public services. The municipality's bond ratings could also suffer as a result, making it more difficult and expensive to raise funds for other public projects in the future."
Again, we're getting years away and lots of speculation away from what's been proposed. It's not out of the question, of course, but this is the Malthus argument for the network: you know, population always increases indefinitely until we run out of all resources and we all die. Familiar to any debate club member.
"Thus, taxpayers could be stuck with greater tax burdens if private entities leave town and take their tax dollars with them."
Some telcos have threatened to take operations out of towns that plan to build muni networks.
Page 17
"Regardless of such concerns, many cities like Philadelphia will charge forward with wi-fi municipalization plans on the assumption that private firms will not abandon the expensive networks they have already built or planned."
See the earlier discussion on bundling. Because these current wireless plans don't offer the speed, consistency, or low-latency of most wired broadband networks, add-on and bundled services won't be offered. So as long as Comcast et al can offer cable, voice, and broadband on a single bill and promise 5 Mbps or much more in the future, that's not a hard sell against an unbundled service that runs at maybe 1 Mbps. Bundling will keep competitive wireline providers in the market.
"If the municipalization pus h is primarily about the price of service—as many local officials seem to argue it is—then it is likely that local officials will constantly apply pressure to keep the municipal broadband utility's rates lower than whatever the private competitors can offer."
I can agree with that completely. This is one of the risks that has been undercited elsewhere and one of the primary ways in which taxpayer-funded municipal networks could hit a wall.
"In antitrust terms, this referred to as "predatory pricing.""
Oregon apparently has these laws, and there is some concern that Ashland's fiber network has been pricing below cost according to a local newspaper.
Page 19:
"Such concerns partially explain why some state officials have mounted legal challenges to local municipalization efforts."
Now, please. This concern is largely funded by incumbent lobbyist. The fellow who introduced the bill in Texas, Rep. King, said he was unaware of the issue before he was told about and has consistently spoken about the scope of his bill incorrectly--which is why it's had such popular opposition. The original Texas bill would have put out of business all airport Wi-Fi providers, for instance. Further, enacting laws isn't the same as mounting a legal challenge--that would involve the courts, not the legislature.
"These state officials recognize that there are broader, long-term economic costs for the state and the nation should their local municipalities choose to substitute taxpayer-subsidized competition for market competition."
I would love this report to pieces if a reasonable explanation were in this paragraph, too, about the vested interests and the methods by which they have been involved and disguised their methods. The NMRC is one example, but the PFF co-released the other report in this pair with the NMRC, so it's unlikely they would be critical here.
Issue Dynamics, Verizon, and SBC are all deeply, deeply involved in getting these laws introduced. Some incumbents have spoken at length and openly about their super rosa role. I have writtea lot about the sub rosa role. Let's not be naive here and pretend state officials are driving this car.
In West Virginia, legislators wanted to change the bill that they introduced at the behest of incumbents. Here's the quote from a representative involved:
‘"We were moving toward a compromise with the industry," said Helmick, D-Pocahontas. "But they don't want the bill, and they're going to be against it."'
"encourage private investment and competition through:
· regulatory reform to diminish unnatural barriers to entry;
· reform of burdensome local taxes or fees;
· easier access to rights of way / easements / public land or buildings;
· if necessary, direct subsidies / vouchers for the neediest. (importantly, any
subsidies or vouchers should be means-tested."
All of these items transfer taxpayer money to private groups or reduce corporate tax receipts that go into government coffers to pay for services.
Posted by Glennf at 9:59 AM | Comments (1) | TrackBack
Talk of the Nation covers municipal networking: The group of guests is biased quite a bit toward the very-pro-municipal side (2:1 plus a reporter), but I haven’t had a chance ot listen to this program yet.
Posted by Glennf at 7:06 PM | Comments (1) | TrackBack
Macworld interviewed me for their first Geek Factor podcast, talking about Bluetooth 2.0: We had a very detailed conversation about the issues around Bluetooth 2.0+EDR (Enhanced Data Rate) and what it means for cell phones, headsets and headphones, and the future of the standard. I start about halfway through the podcast.
Posted by Glennf at 5:05 PM | Comments (0) | TrackBack
Spyguard endorsed by British government to reduce RF emission of wireless networks: The window film and a similar paint reportedly turn rooms into Faraday cages without all the tedious insertion of a metal mesh. But MI5 gets to monitor who purchases the materials.
The group selling it claims it was developed by the NSA and that they have exclusive rights. Sounds a little James Bond-ish. Why would the NSA allow such an item to be sold overseas? I checked, and I’m wrong: they have a list of technologies available for license. (Hey, what happens when you link to the NSA?)
Posted by Glennf at 10:36 AM | Comments (0) | TrackBack
Atheros will ship a/b/g and b/g chips for PCI Express in third quarter: This next-generation bus design supports a much improved architecture for maximum throughput across all cards, and Broadcom and Atheros are both interested in being on top of its deployment. Atheros says that they have a single chip solution that integrates into a single-sided PCI Express card; sampling is already underway to its best customers.
Broadcom announced in early April that they have a PCI Express chipset—ostensibly at least two chips—that’s was in sampling then. I expect a war of the words over throughput, cost of goods, and other factors in the months ahead.
Posted by Glennf at 8:41 PM | Comments (0) | TrackBack
The debate is much more reasonable in tone between differing parties in the opening days of Minneapolis’s RFP: Unlike the Philadelphia plan, which was widely and inaccurately criticized for months before its release—and which criticism hasn’t been updated to reflect the plan’s real content—the Minneapolis RFP is provoking more reasoned discussion among debaters of the merits of muni networks.
This interesting piece in the Minneapolis-St. Paul Business Journal reminds local residents once again of an ill-fated plan set in 2000 for Time-Warner Cable to build out a fiber network. That work hasn’t happened and Time-Warner won’t comment on it. But they might bid on the new municipal plan.
The city’s network planner, Bill Beck, said 20 companies have already responded with queries for the proposal in which no public money will be used and the network will be entirely privately owned. The city will give the network its telecom business and provide what appears to be preferential access to buildings, poles, and other facilities necessary to build out a wireless and fiber optic network.
I say this debate is more reasonable because Braden Cox of the Competitive Enterprise Institute provides what I would argue is not a strawman argument noting that there isn’t an economic imperative in MInneapolis. It’s not in the same state of broadband availability as Philadelphia. That’s indisputably correct. Cox is interested in the Minneapolis model, too, as it “doesn’t risk any taxpayer money…”
Steven Titch of the Heartland Institute, a group I have regularly criticized for not revealing any (if any) ties to incumbent telecom firms and for releasing a broadband report with a group that is owned by Issue Dynamics (whose clients include most major telecom firms), makes his usual argument here: that hotspots don’t attract business travelers. But Minneapolis’s plan calls for limited hotspot Wi-Fi; it’s all about residential and business use, and includes fiber optic as part of the plan.
He objects that the technology isn’t quite there and the timetable is ambitious. I’m finding more and more that I am agreeing with the “wait a little while” approach. With the emergence of faster and cheaper standards to the market, I’m not sure that a network planned today and ready in two years will represent the state of the art except on the fiber-optic side which requires so much more physical work to build out. It’s possible that the city could fully realize the fiber optic part first with its private partner and delay the wireless part until the backbone timeline was set. That won’t cost them more money—except in increased labor costs which should be offset by reduced equipment cost.
Esme Vos points out, however, that similar networks have been built in Europe and have proved that competition from municipalities has spurred cheaper prices and higher speeds.
I’m particularly interested in whether a city-franchised entity can be put together in such a way that it doesn’t have discriminatorily low rates for facilities—meaning that no other competitor could afford to build a similar network if they so chose—and that it becomes a third choice alongside cable and DSL. If you could have a city-backed but not owned network plus the duopoly in place and add onto that TowerStream and other business-oriented broadband wireless firms and a handful of smaller residential firms—that could be enough competition to spur innovation, higher speeds, and lower costs.
The entrenched interests may find themselves operating more trenching tools if the multiplicity of options continues to grow.
Posted by Glennf at 5:09 PM | Comments (0) | TrackBack
Long-time tech report Rafe Needleman editorializes in favor of sensibly free Wi-Fi: He’s not asking for a gift, but rather describes a set of strategies in which more free Wi-Fi could be made available to travelers and the public, and how to convince hospitality operators like hotels to include Wi-Fi in their amenities budget so that it’s just plain free.
I’m not so convinced that Wi-Fi needs to be free everywhere: rather, I would argue that roaming needs to be universal with a single login and that pricing needs to drop to the natural bottom, which is about $20 per month. When you can get consistent, reliable, supported service everywhere you go as a business traveler for $20 per month it’s worth the price. When you have to traverse several networks and a melange of free and fee networks all of which offer a differ set of costs and promises for availability, it’s not worth the price. Free can cost you money if you can’t use it and there’s no promise you can.
In fact, I still believe that some locations may offer both free and fee usage: for free, you get bandwidth limits, limits by service, time limits, and no support. For fee, you get full bandwidth, no service limits, and full tech support, plus secure login over 802.1X. Having two separate network names would aid this, of course: “Use Us Free” and “Use Us Fee.”
Rafe notes that municipalities may provide Wi-Fi soon, but I’d caution him to believe that it’ll be free. Most of the plans—and all of the sensible ones—that would roll out metropolitan Wi-Fi involve charging end users in homes and businesses but offering some or extensive free hotspot service in parks, city buildings, libraries, downtown areas, and other public places.
Posted by Glennf at 3:23 PM | Comments (7) | TrackBack
We called him crazy, but he just kept coming at us: Peter Judge writes about Extricom, the company that produced a barrage of what appeared to be overblown throughput claims last November, but now offers enough details to evaluate their technology. Their claims of huge throughput weren’t across the entire system—that is, 1 Gbps everywhere—but rather aggregated throughput from multiple cells on the same network using the same channels.
The system promotes channel reuse by leveraging the collision detection that’s at the heart of 802.11 and Ethernet systems to better use the empty spaces that are wasted in routine Wi-Fi communication. Each Extricom switch has multiple thin APs on the same channel. The switch decides which AP handles which client without switching channels, and thus the client doesn’t change its connection (which means handoff latency is reduced far below any conventional system) and the switch maximizes the use of the RF space.
APs are coordinated at the switch level to avoid interference, but the 802.11 specification can handle co-channel interference as well. Between those two parts, the amount of interference is dramatically reduced. The goal is to allow many simultaneous voice conversations by bringing each client’s available bandwidth as close to the maximum throughput for their particular standard.
The only complaint from a test site seems to be the current eight-AP limit on their first switch model. That model will ship in May for $8,000 to $14,000 based on quantity and options like Power over Ethernet, according to the Techworld report. A 32-port switch will follow in the fall.
It’s ingenious, and I’ve confirmed that this could work (if implemented properly) with a Wi-Fi expert. It’s too bad they didn’t explain this more clearly six months ago.
Posted by Glennf at 3:18 PM | Comments (0) | TrackBack
Strix Systems moves into the great outdoors with products announced now shipping: Their outdoor mesh routers are hitting distributors. Tempe, Ariz., will use Strix equipment through its integrator MobilePro. Strix, Tropos, and BelAir all offer multiple radio outdoor mesh networking gear that uses Wi-Fi. Tropos trash talks the competition in this article, noting that their use of 2.4 GHz for backhaul is actually a plus over BelAir and Strix’s decision to backhaul over 5 GHz.
The Tempe hotzone will cover 40 square miles, the article reports, passing 65,000 residential households and 1,100 businesses. First responders will use the network, which will also offer service across the 4.9 GHz emergency band, which sounds like a unique offering. Subscribers will pay a yet-to-be-determined cost, but the service will be free in the downtown district to attract visitors and businesspeople.
Posted by Glennf at 2:42 PM | Comments (0) | TrackBack
Sprint PCS has turned on Wi-Fi at the Memphis International AirPort: For $9.95 per day or as part of an unlimited usage $49.95 per month subscription, you can use the airport-wide network. The airport handles 11 million passengers entering and leaving the facility. Sprint PCS has struck a number of bilateral roaming arrangements lately, so Memphis should become available to subscribers to other national networks.
Posted by Glennf at 2:37 PM | Comments (0) | TrackBack
At a London WLAN conference a fake network pushed viruses, worms: According to this ZDNet article (which relies on comments from a single individual who works for a reliable company that sells corporate Wi-Fi monitoring tools), a fake network set up at the Wireless LAN Event in London last week pushed 45 “random” viruses and worms to users when they connected. The article claims that the viruses were “randomly” generated, but that isn’t a realistic description of how viruses work.
Posted by Glennf at 6:44 AM | Comments (2) | TrackBack
Superb article in The Boston Globe on the state and future of Wi-Fi and wireless in Our Fair City: I’m not a resident of Bean Town—which no one there every calls it, I know—but I’m a great fan of the city and have long been following Michael Oh’s efforts to spread free, commercially backed Wi-Fi. He’s profiled in this long and well-researched article. It lacks the flaws that many similar articles have had: it’s technically accurate, paints a broad picture, and doesn’t try to be too cute about the technology and its implications.
I’m not sure I buy that there will be 130,000 hotspots in the US by the end of 2005, but the number won’t be an order of magnitude off. JiWire.com says there are over 20,000 now. I’d ballpark it at 30,000 to 40,000 depending on how you count all the one-offs and special cases. Based on targets set by T-Mobile, Wayport, SBC, and Sprint, hitting 75,000 by year’s end seems possible.
One Boston city councilor wants to emulate other cities planning metropolitan-area Wi-Fi networks and have Boston jump onto that bandwagon. I hope he knows what he’s getting into. The Beacon Hill Institute, which issued a negative report last year on municipal cable and broadband (looking almost exclusively at fiber-optic systems) is just over the Charles River from him at Suffolk University.
Tobin cites the same vague desire to bridge the digital divide that I’ve heard elsewhere, but I haven’t seen any statistics that show that improving Internet access helps school test results, increase income, reduces crime, increase employment, or improves literacy. Does anyone have studies that show positive outcomes? Or any outcomes? It all seems a little nebulous. I disagree with those that say that $2,000 computers are needed to use Wi-Fi networks—about $150 would buy a refurbished computer and a Wi-Fi card—but I do wonder how the digital divide is spanned when you throw computers and broadband at folks.
Boston has unwired all of its municipal libraries and is experimenting with adding free Wi-Fi in city-owned buildings. The local transportation authority is planning to add Wi-Fi to the local subway and rail system, the T.
The article concludes with a lovely set of musings on the nature of human interaction, and a suggestion that Wi-Fi hotspots encourage a kind of interaction through solitude in the midst of community.
Posted by Glennf at 11:55 AM | Comments (0) | TrackBack
Jack Uldrich writes in the Minneapolis Star-Tribune that Wi-Fi is not enough: He takes the Minneapolis city-wide fiber/Wi-Fi hybrid network as a jumping off point for arguing for a national broadband policy that would provide the structure in which businesses and homes would have access to cheap gigabit Internet access. You can tell this is being written in Minnesota, because Uldrich wants the government to create a policy that encourage open competition and open access (read: non-discriminatory access in opposition to the way cable is run and the way that telcos tried to restrict DSL). But he doesn’t outright state that the government should build this network. It’s unclear who should pay for its construction if the market already doesn’t have the incentives to build it.
Posted by Glennf at 11:45 AM | Comments (1) | TrackBack
Update on the status of the first large-scale BPL effort in the US: This is the largest commercial rollout of BPL, which delivers data over powerlines by encoding information on high-voltage lines. High-speed broadband can span fairly large distances over existing wires through devices installed at points, which makes it awfully appealing.
So far, estimates—not by the power company, Cinergy—are about 8,000 homes have signed up out of 50,000 which could be served. Cinergy and its data partner Current Communications Group expects to pass 250,000 homes within three years. They aren’t giving out numbers themselves, they say, to forestall providing competitive information.
The article opens with line workers installing a bypass box—that’s because transformers don’t pass the encoded data. Some BPL models use Wi-Fi as the last-100-foot solution, instead of using BPL to the home. This could wind up being a great WiMax/BPL hybrid, too, with WiMax on a certain frequency of poles serving a group of homes.
The utility is starting to gear up using the BPL for its own monitoring purposes, too. In other cases, utilities have worked the other way around, building fiber-optic installations for monitoring that are then turned to the use of broadband for city or for-fee public use.
Some of the think-tanks and analysts who are either opposed to municipally run broadband on ideological, financial, or other grounds are promoting BPL as a method of adding competition without requiring more wire in the ground. But BPL has seen little commercial uptake yet as power companies apparently haven’t acted generally interested in it. Related to this is the issue that many electrical utilities are municipal entities, in which case that introduces that element back into their concern about rolling out a network: should they in the current climate?
Posted by Glennf at 11:35 AM | Comments (4) | TrackBack
T-Mobile cell and hotspot subscribers can use this to find out specific coverage areas: This is a marvelous tool and one all carriers should emulate. It shows specific levels of expected signal strength for T-Mobile’s cellular network—this would help plan for GPRS use as well as voice—and has a checkbox that lets you see T-Mobile’s hotspots as well. You don’t have to have a T-Mobile account to use it.
Verizon’s CEO recently stated that people shouldn’t have the expectation that cellular service works everywhere, like in their homes, or have access to detailed information about coverage
areas. The San Francisco Chronicle wrote, “Seidenberg said it’s not Verizon’s responsibility to correct the misconception by giving out statistics on how often Verizon’s service works inside homes or by distributing more detailed coverage maps, showing all the possible dead zones.” Apparently T-Mobile thinks otherwise.
(Seidenberg also said that citywide Wi-Fi was the dumbest idea he’d ever heard while his company is preparing bids to build such service for cities.)
I’ll also note that while T-Mobile has no good migration path for 3G cell data service in the US—they’ve got GPRS and some vague ideas about the future—they are the only carrier that owns and operates hotspot locations. SBC contracts their operation out to Wayport, which is fine, but T-Mobile has every reason to try to emphasize their street-level cell and hotspot operations. And to say nyah-nyah to Verizon. [link via Engadget]
Posted by Glennf at 11:12 AM | Comments (0) | TrackBack
Can you make users pick strong passwords?: It’s well known that users are the problem in security. (Yeah, right.) I mean, you tell people to choose passwords of 16 characters or more with no words found in any dictionary on the planet using a combination of letters, numbers, punctuation, and the symbol for the artist formerly formerly known as Prince—and they just choose “dog,” “cat,” or “1234.”
So we know that’s not really the problem. The problem is providing reasonable tools to allow users to select passwords or passphrases that don’t require them to memorize impenetrably long sequences that force them to write them down in order to have them available.
This Slashdot discussion linked to above links to a Security Focus discussion which runs through many of the issues. Even if you have a secure method to deliver password suggestions to a user, it’s unlikely that those suggestions would meet the user’s personal needs for recollection, identity with other passwords they use elsewhere, and security.
WPA’s preshared key version—WPA Personal—suffers directly from this problem. Although you can enter long phrases (up to 63 characters), none of the interfaces provided by manufacturers requires a key of sufficient entropy to ensure that it beats the WPA weak passphrase choice problem. If you choose a short WPA passphrase with words found in a dictionary, it’s possible for that passphrase to be cracked. This says nothing about the underlying WPA encryption, which is very good, but about the mechanism to generate the key used in that system.
Many systems solve the key problem by generating an extremely long and secure key that itself is secured through a passphrase that’s passed through a strong one-way encryption algorithm. For instance, PGP’s system protects the long private key that’s part of the public/private pair using a passphrase of arbitrary content and length that you create. The passphrase isn’t related to the private key—the private key isn’t derived from it—but it can be something simple to recollect but not simple in nature. I use a long English phrase to protect mine, for instance, which is an increasingly frequent recommendation.
Ultimately, there has to be a way to have strong keys that aren’t derived from passphrases and that can be used more easily by those who have proven their identity. This requires management, though, and it will take a long time for anything comprehensive to be rolled out.
I would argue that Apple’s Keychain is the closest element that’s currently available. Apple uses a passphrase to protect the Keychain, which is strongly encrypted. Keys can be retrieved as needed through the use of the passphrase. It stores keys, passwords, and certificates and can set individual access control limits per item, as well as managing multiple Keychains.
Posted by Glennf at 8:27 AM | Comments (0) | TrackBack
Google is sponsoring free Wi-Fi in Union Square in Frisco: I can’t find out more information about it than what’s in this photograph (click for larger image). One colleague I mentioned this to suggested that increasing Internet usage increased Google’s ad revenue so just offering free Wi-Fi could translate into more revenue for them indirectly. Possibly. It might just be more branding effort—as if they need that, especially with yesterday’s earnings news.
Posted by Glennf at 9:11 AM | Comments (2) | TrackBack
Excilan’s pay-by-cell-phone technology has been acquired by The Cloud’s Roampoint: Excilan went into bankruptcy in its headquarters of Luxembourg last Friday, but a last-minute deal with Roampoint, the aggregation and roaming service that’s part of UK’s large The Cloud network, saves its unique method of payment. Excilan’s system allows a cell-phone carrying user to enter their telephone number on a hotspot gateway page, receive an automated call, and authorize payment which is charged to their bill. The significant difficulty faced by Excilan was getting both hotspot operators and cell operators in the same countries to participate. Excilan has 40 hotspot operators and 17 cell operator contracts, but few overlap geographically.
Posted by Glennf at 7:22 AM | Comments (0) | TrackBack
Sprint now claims 19,000 hotspots in its aggregated network: The company announced that it will gain another 6,000 locations from Quiconnect, 3,800 from Fiberlink, and several hundred from Pronto, Opti-Fi, and Nomadix. (The Fiberlink locations are actually resold from Boingo’s aggregation platform, although that fact isn’t mentioned anywhere.)
Sprint previously had arranged deals with SBC, Wayport, Airpath, STSN, and Concourse, as well as limited bilateral roaming with AT&T Wireless (now Cingular) for airport access. Those locations must have totaled 10,000, although I’m having a little difficulty adding up all of the component networks.
One of the key elements Sprint is pushing is its Extended Workplace, a way of having a single user interface for connecting across all kinds of communications methods, including dial-up, Wi-Fi, cell data, and Ethernet. Extended Workplace provides companies with a way of enforcing end-user policies, like VPN usage or anti-virus protection—just as with software from remote-access providers like iPass.
Pricing for Extended Workplace is $120 per month per user for unlimited Wi-Fi and Sprint PCS Vision (its brand name for 1xRTT data service) with additional metered fees for dial-up and other connection services.
The article bizarrely quotes a Sprint business development manager stating that Sprint started building airport Wi-Fi service in 2000 and now has seven airports. Now I’ve been writing about and researching airport Wi-Fi since 2000, and I can state categorically that Sprint didn’t start getting into the business as a provider until 2003. If they were providing the back-end outsourced services, then they were handling it for Nokia and other companies without revealing their brand at the time. Nokia, Wayport, and MobileStar unwired the first airports in North America that I’m aware of all before 2001.
Posted by Glennf at 6:10 PM | Comments (0) | TrackBack
The two companies have similar hotspot footprints: Both firms offer service in airports, hotels, college campuses, apartments, and highway plazas. Their locations tend to be the overlooked second tier that has a substantially underserved market. For instance, while Los Angeles still lacks Wi-Fi across their airport, which handled 60 million passengers (combined embarkations and debarkations), ICOA fulfilled legislators’ dreams by putting in Wi-Fi in Sacramento’s terminals, which serve 10 million. Likewise, WiSE is in airports like Omaha, San Diego, and Baltimore.
Posted by Glennf at 5:56 PM | Comments (0) | TrackBack
The Star-Tribune unsigned editorial doesn’t dislike the idea of citywide coverage, but suggests careful planning before leaping in: This editorial isn’t opposed to the lovely notion of ubiquitous coverage through Minneapolis, as described in the city’s request for proposal. In fact, it describes the idea as “cool.” But it suggests that there’s not enough detail in the current RFP to answer many outstanding questions, and that until proposals are received from bidders, the details will remain unknown.
Minneapolis is proposing a network that’s of a larger scale than Philadelphia’s with a more extensive purpose and that uses Wi-Fi for access but is tied together with fiber optic which will be used for other purposes. The city won’t put a cent of funding into the plan, but will guarantee its telecom business to the project’s winner, thus ensuring early revenue. (Both Minneapolis and Philadelphia project large cost conservation by shifting their telecom spending to this kind of network.)
The editorial notes that a previous effort to have Time Warner, the city’s cable franchisee, install fiber-optic links among city buildings was agreed on in 2000—and still hasn’t been built. This proposal might be an effort to route around that blockade, but it does create essentially a new kind of franchise.
The editorial concludes, “As this ambitious proposal goes forward, its balancing of complicated interests will bear very careful scrutiny.”
I admire the stance of this editorial because it doesn’t throw any strawmen onto a pyre and fling matches at them. Instead, it looks at the reasons advanced by the city for having this network built, analyses the risks without overstating them, and generally endorses the concept while remaining patient about how the concept turns into reality.
Posted by Glennf at 9:30 AM | Comments (0) | TrackBack
Microsoft and VeriSign have own flavor of how to protect networks from infected computers: This new architecture will be based on Microsoft’s Network Access Protection (NAP) and VeriSign’s Unified Authentication platforms. It’s supposed to protect networks by checking that a laptop trying to connect over Wi-Fi has been issued a clean bill of health with the latest patches and virus definitions, among other factors.
But this announcement doesn’t mention a press release from yesterday from the Trusted Computing Group’s Trusted Network Connect specification will also work with NAP. The TNC spec allows computers that connect to a network through any medium to validated for security before being allowed access. It ties nicely into 802.1X port-based authentication. If a computer fails validation, it’s segregated on a protected VLAN that only offers access to patches and updates, but can’t reach the rest of the network.
Posted by Glennf at 8:36 AM | Comments (1) | TrackBack
Tor Amundsen whips up a Wi-Fi to 3G box: While you can purchase commercial systems like the Junxion Box to create a Wi-Fi network that’s uplinked to a 2.5G or 3G cellular network using a PC Card from a cellular provider, this is the first do-it-yourself project I’ve seen to achieve those results. He used a Soekris box, a well-known and well-liked generic embedded OS appliance box, and installed NYC Wireless’s embedded Linux. There’s some work with a drill press and lots of kernel fiddling, but it’s a neat idea, and a replicable project.
It also shows how the genie is out of the bottle for using cell data as backhaul whether the cell operators like it or not. The flip side? Many operators are monitoring usage and cut off those engaged in what they think are purposes contrary to their contract. [link via Engadget]
Posted by Glennf at 7:18 AM | Comments (1) | TrackBack
Robert Liu verifies early speculation that the Philly network needs indoor bridges: A CPE (customer premises equipment) could be a T1 modem, a cable modem, or a DSL modem—or, in the case of many Wi-Fi mesh networks, a Wi-Fi to Ethernet bridge with a high-gain antenna. This has been the missing piece in much of the coverage of Wi-Fi mesh networks.
Critics have said that cards built into laptops or added to desktop machines couldn’t receive signals from ubiquitous outdoor metropolitan networks. They’re generally correct. But until this fact-finding pre-proposal meeting in Philadelphia that Liu attended at which potential vendors had a chance to ask questions for the first time it was unclear that a CPE was required. The RFP states on page 10:
Should additional customer premise equipment (“CPEs”) be required or assumed in order to deliver this in-building coverage, Respondents are expected to state this in their Proposals and elaborate on this requirement and their assumptions.
Liu says this is now a given, not a “should.”
I disagree on this adding substantial cost, however, as Liu writes: a Senao 200 mW 802.11b bridge retails for under $100; in quantity, it would be substantially less. Because the network builder will be a wholesaler, they can’t recover the cost of this device from the consumer directly, but it appears they’ll be responsible for CPE selection. This could be built into the price of the network, representing a couple dollars per month of the wholesale cost. Or they could offset the cost to ISPs who recover it from end-user fees or leases. The Senao is being used in some Tropos deployments right now as a CPE.
Liu reports as well that the information provided to vendors who want to bid is probably inadequate for true RF planning: building footprints date to 1996; no building heights were provided. Other topographical details about poles and other city facilities was available, however.
Posted by Glennf at 11:15 AM | Comments (3) | TrackBack
Nintendo will use Broadcom to add Wi-Fi to “Revolution”: Only sketchy details are known about Nintendo’s next gaming system—more should be revealed at next month’s E3 conference—but they’ll definitely be using Broadcom’s Wi-Fi technology with several features turned on. It’s very possible that SecureEasySetup will be one of those features. Broadcom touts the technology as a painless method of ensuring that devices receive the maximum protection from WPA with the least effort by a user. Push a button on a router (or via a router’s software) and then on the device you want to add and you’re all done.
Posted by Glennf at 10:56 AM | Comments (0) | TrackBack
Glenn Reynolds asks from the right side of the aisle, what do incumbents fear? Reynolds, the No. 1 Google Glenn and author of the Instapundit blog, has a decidedly and honestly conservative viewpoint, but he’s as interested as folks to his left as to why the incumbent telecommunications providers are so worried about the interest by cities and towns to build their own broadband networks or have those networks built for them under franchise.
He writes, “There’s nothing illegal or improper, of course, about companies talking down competition, or hiring lobbyists to persuade cities to do things their way instead of somebody else’s way, but there’s nothing terribly impressive about it, either. In fact, the more those companies criticize the municipal wi-fi approach, the more it makes me wonder what, exactly, they’re afraid of.”
I have to agree. I’m not a fan of sub rosa lobbying, which is why I’ve written so much that complains about the attempts by incumbents to fund reports from groups that appear independent. What’s ironic, of course, is that if incumbents bid to build the networks that the reports say are impossible to run reliably and are unnecessary, how do the incumbents explain to their shareholders their participation in those projects?
Reynolds also has the very reasonable concern that city-run networks could be subject to city-run monitoring. I’ve heard this concern in poor contexts before; here it’s presented without any baggage. The most rabid pro-municipal-broadband supporters should acknowledge that under the current set of laws in the U.S. having municipalities directly responsible for the operation of new broadband networks could lead to personal information finding its way into government hands. Filtering laws, if they ever make it past the U.S. Supreme Court, might also affect these networks.
The proposals we’re seeing from cities that want to have networks built are morphing, though, probably due in part to the firestorm of Philadelphia’s initial reaction and the rash of laws spreading from state to state to restrict what some cast as a vital public utility that’s underbuilt and others view as municipal attempts to regulate on a local level what only states or the federal government should have the ability to micromanage or not micromanage.
Philadelphia’s plan hands off its network fundraising, build-out, and operation to a non-profit that will be ostensibly outside city control. This arm’s length plan would also ostensibly remove the city’s ability to monitor or be required to filter the network. Further, the non-profit would only sell wholesale access to ISPs. Minneapolis’s request for proposal says, “keep the city out of it!” A private company or consortium would receive essentially a franchise and a commitment for city telecom business. This would remove filtering, monitoring, and censoring from the pile of concerns as well.
Reynolds, like me, is interested in diversity, and that’s probably the broadest argument one can make for more competition of all kinds: “…municipal services are likely to be better when people have a standard for comparison, too. Being the only game in town is never good for service.”
Update: There are some great comments below. But a number of those commenting seem to have missed the point that many municipal networks are being planned as (or already set up as) hands-off affairs from the standpoint of funding, taxation, and operation. If a city can’t pay to have trash removed, but they’re not paying for the municipal network, you have to readjust your attitude. If the structure is set up so that taxpayers are completely insulated—that is, it’s encoded in a charter or local legislation that there’s no bail out—then the network might go belly up, but it didn’t cost you, as a taxpayer, any money.
Interesting side note here, too, on funding and taxes. Most municipal projects are criticized as having two tax and funding advantages: a city can issue tax-free bonds, and a city doesn’t pay a variety of taxes. The newest projects I’ve seen don’t provide that advantage to the municipal project. And if you look at one of the oldest broadband networks in the country run by Tacoma Power, they pay cable franchise and local and federal taxes. They even devote a page to their tax bill. Of course, Tacoma Power is separate organizationally from the local government.
Posted by Glennf at 10:06 AM | Comments (12) | TrackBack
Truck stops in Texas with free Wi-Fi may have to filter content: A Slashdot poster connects the dots in a Texas house bill that would require filtering on any state-provided wireless network on public property. This means the truck stops that have been equipped would need filtering. I don’t need to make snickering references here, as you can read plenty on Slashdot.
Posted by Glennf at 8:45 AM | Comments (0) | TrackBack
Trapeze has added support for several Cisco APs: WIth a command-line change, a Cisco AiroNet 350, 1100, or 1200 can be part of a Trapeze-managed WLAN switched network. This should make it an easier sell for Trapeze VARs walking into Cisco-oriented enterprises, especially with Cisco VARs and direct sales folk trying to push new Airespace equipment into existing installations. This announcement ranks up there with AirWave’s recent 3.1 version bump that allows AirWave’s software management tool for WLANs to handle Cisco Airespace devices, too.
Posted by Glennf at 10:55 AM | Comments (0) | TrackBack
It’s pretty extraordinary when there’s competition for giving away security: On the small-office/home-office (SOHO) end of the market for Wi-Fi security products, there’s an increasing trend to just give away products that can help very small number of users in the hopes that those users then adopt the for-fee products when they grow in scale or need. But it’s a very nice development for SOHO warriors, to be sure.
LucidLink joins this trend by dropping the price of their three-user package for secure WLAN connections from $99 down to zero. My guess is that it was a hard sell for that user category, and that becomes a great test tool for offices considering LucidLink. They can hook up this three-user version as a test before buying a multi-seat license. The three-user limit is simultaneous users. More can be registered in the system.
LucidLink is a locally hosted 802.1X server package that uses proprietary client software for Windows 2000 and XP to make the connection. The server software requires Windows 2000 (any edition), XP, or Windows Server 2003.
The next step up is a 10-user version of LucidLink for $449.
Posted by Glennf at 10:45 AM | Comments (0) | TrackBack
UCSD did a great job getting into the media this week with a fast Wi-Fi handoff technology: SyncScan drops a Wi-Fi adapter or appliance, like a VoWLAN phone, out of its associated mode for a few milliseconds at a regular, defined interval to check on signal strength. This avoids adapters swapping to a new AP only when signal strength becomes unusable or nearly so.
SyncScan relies on a feature in Atheros’s chipsets that’s available from the open-source madwifi drivers; it’s the same sort of feature (if not the identical one) that allows Atheros’s WLAN switch partners to offer RF monitoring on the same APs that are also handling client data interchange.
Two problems with SyncScan’s approach: first, it requires firmware to be installed on the access point, which is fine for experimentation and open-source projects, but otherwise needs signoff from major firmware developers and their manufacturing partners; second, it’s got that patent-pending label attached, which always has the caveat of causing resistance until fees are revealed.
SyncScan puts all APs within listening range of each other into a synchronized beaconing mode so that the “I’m alive” signals happen at fixed intervals. This allows adapters to only listen at discrete periods and to get a clear idea of precisely what’s happening in the local RF space. But this coordination adds overhead and there has to be a cost to synchronization and the inevitable resynchronization.
One of the IEEE 802.11 groups, 802.11f, was dedicated to fast reassociation through preauthentication—tokens exchanged among APs at Layer 3—but that doesn’t help with fast reassociation on an RF level, or Layer 1.
Posted by Glennf at 12:37 PM | Comments (0) | TrackBack
Hotspot payment system provider enters bankruptcy, halts services: I obtained information yesterday that Excilan, a Luxembourg-headquartered firm that offered a unique method of payment for hotspot services, has entered bankruptcy in that country Friday. The service will halt operations Monday at noon local time according to details I received.
Excilan allowed cell phone users at a Wi-Fi hotspot to enter their number on a gateway page and receive a confirmation call that they could use to agree to pay for the service. Customers of cell operators that hadn’t agreed to work with Excilan were granted short, free sessions which Excilan used as a market-research tool to find partners.
I don’t know the details of why the company has gone into bankruptcy; they were privately held. The company has been remarkably upfront about its usage, however, posting statistics on its home page. The most recent quarter saw 7,098 sessions, which at the highest possible commission and split might have meant just a few U.S.$10,000s.
Sean O’Mahoney joined the firm last year as CEO. O’Mahoney was formerly a founder and the CEO of Vancouver, B.C., Canada’s FatPort, which was an early Excilan partner. I was unable to reach O’Mahoney for comment since obtaining the news.
Posted by Glennf at 9:55 AM | Comments (0) | TrackBack
Vitriol spews from Verizon CEO’s lips on municipal networks, customer expectations, cancellation fees: Get this man a Valium, stat! Verizon CEO Ivan Seidenberg needs to calm down a bit. He thinks San Francisco’s municipal network idea is “one of the dumbest ideas I’ve ever heard” because the network has to be built, designed, and operated. Oddly, Verizon has no expertise in building campus-wide Wi-Fi networks, and thus isn’t really qualified to express that sort of opinion. I had the same reaction to reading a Comcast spokesperson’s critique of the potential of a city-wide wireless network: Comcast doesn’t build that; they’re not qualified to say it’s possible or not.
What’s even funnier about this quote is that Verizon will probably bid to build city-wide networks in Minneapolis and Philadelphia.
Posted by Glennf at 7:29 AM | Comments (1) | TrackBack
Voice over IP in the enterprise company winds down: Five short weeks ago, I was over at Telesym’s bustling headquarters east of Seattle having a great conversation recorded partly in this podcast about Telesym’s latest products and the tweaks they had made to their offering to better integrate it into enterprise phone switches.
Now the Seattle Post-Intelligencer reports that the company has laid off most employees with severance packages and hopes to return some capital to investors if it can sell its intellectual property. The CEO says fairly bluntly that their product doesn’t scale, so I’m not sure how encouraging that would be to potential buyers.
The company’s two co-founders were pushed out in previous months, and they’re a little bitter about how the products developed without their shepherding them to completion.
One of their competitors, Vocera, says the company floundered by not having a sharp focus and by—at least initially—targeting voice over PDAs. This disregards the success that phone/PDA combos have had in the marketplace, of course, but the company was too far ahead of that market. They were also too far ahead of convergence phones that would have benefitted from their integration on the enterprise and Wi-Fi side while roaming onto cellular as necessary.
My kiss of death interview record is unfortunately quite good: I interviewed the CEO of Cometa just weeks before they coasted to a halt.
Posted by Glennf at 7:22 AM | Comments (0) | TrackBack
Every few months, there’s another gas station chain adds Wi-Fi story: Estonia apparently was the first to pioneer this, and a colleague recently returned from there described the country as scary-wired. They’re pretty hip. But the rest of the world has taken a while to catch on, possibly because of quantity. With most chains, you want to make sure that people can consistently obtain the same services. This is what took McDonald’s so long to sign onto Wi-Fi: more than a year of testing with multiple partners led them to their current extensive plan.
Now part of Michigan has its Speedway and SuperAmerica stations unwired. FreedomNet Solutions is providing the service. They hope to strike gold in the way that apparently truck stops have. Truckers and mobile professionals alike need Wi-Fi on the road. It’s just plain easier to remain in the car or grab a bite and commute while still staying on top of deadlines and changes.
Posted by Glennf at 12:30 PM | Comments (0) | TrackBack
Sputnik is offering a dual-radio AP designed to work as a bridge repeater: Sputnik offers software that lets hotspot operators remotely manage hotspots. The new AP will let operators extend the range of their hotspots to cover larger areas. Sputnik also appears to be targeting the municipal market, arguing that its solution would be less expensive than today’s mesh offerings. However, I wouldn’t think that the Sputnik offering would be quite as flexible as the mesh deployments which allow for multiple hops enabling coverage to areas that are very hard to backhaul.
Posted by nancyg at 1:25 AM | Comments (0) | TrackBack
The possible puns regarding this story are practically endless, though the Register does a good job of exploiting most of them: Two Intel employees set up a hotspot at a research camp at the North Pole. It uses Iridium phones for backhaul, which means the access is probably excruciatingly slow, but people at the North Pole probably aren’t complaining. The deployment is purportedly meant to check out how the equipment fares in such cold, wet weather, though there are plenty of other such experiments that have been done for some time. Dave Hughes has done tons of work in Alaska with wireless networks and he was also instrumental in enabling the hotspot at Mt. Everest base camp in Nepal.
Here’s the press release from Intel in Russia: catch that “hot spot” in Cyrllic would be pronounced “zhot spote-um.”
Posted by nancyg at 1:04 AM | Comments (0) | TrackBack
VPN-for-rent services are multiplying: HotSpotVPN.com is the grandpa in the field, but WiTopia.net’s new personalVPN offering challenges their price and technology. In this article at Mobile Pipeline, I talk about both companies services. HotSpotVPN’s recently introduced SSL VPN service competes head-to-head with personalVPN but has higher grades of encryption available at higher costs. Read the whole article for the run-down.
Posted by Glennf at 8:50 AM | Comments (0) | TrackBack
The Brighton Express uses pre-WiMax gear to achieve 60 miles of coverage at 100 mph: Peter Judge reports from the London-to-Brighton line that although coverage is yet contiguous—that’s still to come—the service offers seamless performance across each base station zone by relaying Wi-Fi in the carriages to WiMax base stations along the route. T-Mobile is involved in this unwiring—which was carried out by Nomad Digital—and is offering the service at no charge while they tune the system.
The estimate is that the 37 802.16d-based Redline devices will need to be increased to 60, or a density of about one per mile, to provide complete coverage. The limit on speed right now is the ADSL backhaul of 2 Mbps per base station. The service switches to GPRS when out of range of WiMax bonding three GPRS devices. Only one train out of 15 is equipped so far.
T-Mobile will charge £5 per hour or £13 per day for service starting in June.
Posted by Glennf at 5:14 PM | Comments (0) | TrackBack
In-Stat says that WLAN switches will become more prevalent, but not as stand-alone devices: As Mobile Pipeline explains it, Ethernet switches will increasingly incorporate WLAN functions making the use of thin access points (most radio intelligence) a given but the centralized functions won’t require specialized hardware. The article specifically notes that Aruba and Trapeze may face difficulties on their own; Airespace was acquired by Cisco.
There’s another course for Aruba and Trapeze and similar companies to take, one that I think we’re seeing the early directions toward. Instead of selling centralized hardware and specialized APs, move to centralized software that runs on commodity PCs that integrators and VARs can configure. The value would move entirely to the switching software. Aruba and Trapeze’s SLAPP proposal is one step in that direction for removing specialized requirements from APs; the next step would be to agree on a standard featureset with extensions that could loaded by individual switches.
Posted by Glennf at 8:53 AM | Comments (1) | TrackBack
HotSpotVPN joins the trend to offer SSL-based full-service VPNs: The service, launched today, offers Blowfish (128 bit), AES-192 (192 bit), and AES-256 (256 bit) encryption using SSL via the open-source OpenVPN client. SSL VPNs are not all about application and browser-based tunneling! OpenVPN is a full VPN just like a PPTP or IPsec service tunneling all data.
The idea of HotSpotVPN and similar services is to bring corporate-grade network obscurity and data protection to those who don’t have an information technology department supporting them on the road.
HotSpotVPN is calling the new service HotSpotVPN2, and charges $10.88, $11.88, or $13.88 per month for 128, 192, or 256 bits of encryption. Multiply those numbers by 10 to get the yearly rate. For now, HotSpotVPN works with Windows, but the underlying OpenVPN software has been built for many platforms; expect more platform support in the future. A free PPTP account is included with HotSpotVPN2.
HotSpotVPN2 competes with WiTopia’s personalVPN offering, another recent addition to the traveler’s arsenal. There’s a little that varies in the setup, but WiTopia’s $79 per year service is available only with 128-bit Blowfish encryption.
Posted by Glennf at 9:15 PM | Comments (0) | TrackBack
EAP types define how 802.1X transaction are conducted with what credentials and encryption: The Wi-Fi Alliance has added EAP-TTLS/MSCHAPv2, PEAPv0/EAP-MSCHAPv2, PEAPv1/EAP-GTC, and EAP-SIM. EAP-TLS has been a part of the certification testing since WPA was introduced because, as one alliance member explained to me, you couldn’t test all of WPA without having at least one secured EAP type involved so that 802.1X could be tested in a reasonable manner. Devicescape and Meetinghouse supplicants join Funk and Microsoft clients in the testbed. These several types represent most of the secure EAP in use.
Interestingly, this article says that WPA2 was released in Sept. 2004. I know that certified devices have slowly been hitting the market and mandatory inclusion of WPA2 is scheduled for later this year. I was dealing with a new baby in September and so I somehow missed this and subsequent WPA2 announcements, but so did many of my colleagues.
The press release notes that the alliance has produced a white paper on deploying WPA and WPA2 in the enterprise, which is surely good reading.
Pos