Entries Tagged as 'wireless'

Thoughts on Clearwire

In this post, I share some of my thoughts on Clearwire’s presence in Portland, both as it relates to Portland and to Personal Telco. I stress that this is in no way a review of the service or a technical critique, as I’ve not used it extensively, nor do I have any in depth knowledge of their deployment. I’m talking here about the social, cultural and economic elements of their entry into the Portland market, and my opinions on how they might succeed or fail.

Sharing Not Allowed

[Update] We have been told by Bill Macfarlane from Clear that sharing IS allowed, and they have been selling service to many places specifically to operate hotspots.

My understanding is that sharing your Clear connection is expressly disallowed by their Acceptable Use Policy. In light of this, Clear is not a boon to those interested in hosting PTP nodes, or otherwise setting up hotspots in areas where DSL or Cable service is unavailable. While I find this disappointing, it is in line with the other major ISPs that operate in Portland, Qwest and Comcast, who both, by my reading, disallow sharing. It is also in line with the data plans of all the major wireless carriers.

Target Market

To me, Clearwire’s largest challenge is how they position themselves in the wireless and broadband markets. In terms of connection speed and availability, Clear looks best when compared to Edge and 3G wireless data services. Like these services, Clear promises an always-on data connection to laptops within the service area through the use of a fairly unobtrusive wireless card. If the current speeds hold with more users, Clear should be a compelling alternative to 3G and especially Edge, as it would provide a 1M+ service, where 3G seems to max out around 500k and Edge is little more than dial-up.

The potential problem for Clear is that smartphones have been steadily replacing laptops for many people’s on-the-go communication. Laptops are now typically only used in fixed locations such as airports and coffee shops, where WiFi connections tend to be readily available–and generally free in Portland, obviating the need for an additional subscription data plan. Additionally, smartphones usually come bundled with data services from the carrier, and are generally not capable of using peripherals such as WiMax cards, meaning that these users will not have compelling reasons to adopt Clear accounts, regardless of the speed benefits that the service would represent.

There is the possibility that smartphone manufacturers will offer phones that include WiMax chips, however, these are likely to remain niche products, similar to push-to-talk phones, since Sprint, by far the weakest national wireless carrier, will likely be the only carrier requesting such phones, at least until Verizon or AT&T develop their own 4G offerings.

Clearwire’s home service is probably well priced for users who do not subscribe to cable television, home phone or other services that provide bundled discounts on broadband. The lowest tier may also attract current dial-up users who do not otherwise use their home phone lines, as $20 represents at least a $5-10 discount over home phone service and a dial-up account, in Portland. Clear may also be invaluable to users who currently do not have wired options for broadband service, though it is unclear, to me, to what extent service availability rather than affordability remains a serious issue in Portland, so this benefit may be one and the same with my previous point.

Overall, as a fixed service, Clear is not game-changing or particularly novel, and I think it is unlikely that they will become a major player in Portland’s home and business broadband market. The appeal of their service will also wane as cable and (hopefully) fiber deliver exponentially higher speeds.

Performance Under Load

By far the biggest technical challenge for any new service is how it performs once it reaches thousands of subscribers. Currently, everything I’ve heard indicates Clear is performing as advertised, but that is, of course, with relatively few users. Clearwire’s infrastructure has a few potential points of failure:

1. Available bandwidth to the Internet. This is a fairly straightforward challenge, and one that should be very easy to improve as needed. Clearwire is positioned in one or more datacenters, where they likely subscribe to Tier One bandwidth providers. Within reason, accessing additional bandwidth is as easy as reconfiguring the circuit. Usage is monitored and adjustments can likely be made as needed. Like most (if not all) providers, Clearwire’s bandwidth will eventually be oversold, meaning that they sell more capacity than they actually have and bank on the expectation that only a fraction of their subscribers will use the service at the same time. Unless economic factors lead Clearwire to skimp on bandwidth, the network will likely not be oversold to the point that it will affect subscribers, as bandwidth expenses should be a relatively small component of Clearwire’s operating costs.

2. Throughput at the node. The greater challenge, especially with a wireless service, is the available throughput at the node (in this case, Clear tower sites). Clearwire’s technology should be more effective at managing client and backhaul connections (which may not be wireless) than MetroFi’s, but even so, there will be intractable technical limitations in the event that too many subscribers are attempting to access an individual tower (or radio). There’s no reason to believe that Clearwire hasn’t anticipated this by building more capacity than they expect to need in the initial years, and in fact, this is likely the case. Nonetheless, as I believe that this issue was responsible for much of the user dissatisfaction with MetroFi’s network, it bears mention as it has the potential to bring a service to its knees if it is poorly managed. As with their Internet backhaul, Clearwire will be oversubscribing their transmitters. The same calculus goes into it–that you won’t have every user active at the same time. Increasing tower capacity is more costly and not as easy as adding bandwidth, but with planning on the technical and financial sides, there is really no excuse for letting this degrade an individual user’s experience.

Net Neutrality

Clear’s Terms of Service are very, well, clear, they don’t give you Net Neutrality. Of course, neither does Comcast, and though some DSL providers might, they certainly aren’t going out of their way to support Net Neutrality. Anyone hoping that Clearwire’s competition would create a market that is more responsive to customer desires is probably going to be disappointed. By hewing to the status quo, Clearwire certainly hasn’t done anything wrong, but they don’t stand out either. This is a great example of why Personal Telco will continue to educate on, advocate for and build community networks. The best way to have a communications infrastructure that is responsive to your needs is to build it yourself.

Clearwire is not MetroFi

This should be obvious, but it seems like a lot of people (including a lot of “experts”) have wrapped together any project that seeks to provide wireless Internet access to an entire metro area together, regardless of the technology used or the parties involved. For all its failures, the MetroFi service was, to me at least, a much more interesting project from a technical standpoint. They attempted to build a city-wide network using technology that was nearly all unlicensed, which is similar to what Personal Telco volunteers have done over the years. Their motives may have been different and they certainly structured their business in a way that I would not personally have recommended, but on the technical side, watching them try and fail was fascinating.

Clear doesn’t hold the same interest for me, because I expect it to work, basically, as advertised. I’m sure there will be some business or customer relations hiccups, but, as far as I can tell, they are basically using licensed WiMax exactly as the engineers designed it. As a business test case, it will be interesting to see if Clearwire can compete with the wireless carriers for mobile Internet service, especially as their success will largely be tied to devices that don’t yet exist, but as a technical experiment, it just doesn’t feel very experimental. My guess is that Clearwire will succeed, however modestly, by still being here in five years, as long as Verizon or AT&T doesn’t introduce a competing high speed WiMax service that gobbles up most of the market share.

Personal Telco, of course, will continue to be here as well, educating Portlanders about the power and value of community networks, answering questions about wireless technology and maybe, if we realize as a community what’s good for us, leading the initiative to build a city-wide, public fiber infrastructure to provide equal access, and future-proof service to the entire city.

Portland MetroFi and Coverage

We keep seeing reports of “Metrofi is roughly 30% complete in covering Portland,” but is that really the right number? Even though numerous reports say you should deploy at least 35 nodes per square mile, MetroFi took the low-end number of 25. While I’m not sure who came up with that number considering the gear used, terrain, etc., they still stuck with 25 per square mile. Regardless of what the reasons are, and most will point to cost, the fact remains that you are not going to get your 95% coverage as MetroFi were contracted out to achieve.

We can agree this is not an exact science, but I will say this: telling the media you are ~30% complete in covering the city is incorrect. MetroFi only deployed ~550 nodes throughout Portland, and Portland is 134 square miles (less major parks like Forest Park). If you do the math, even by using their deployment number of 25 per square mile, that’s only 16.4% coverage. Meanwhile, at what “experts” have suggested (35/mile), that’s only 11.7%.

Even though nobody seems aware on what they’re basing the coverage on, and what areas are considered “Portland” and “worth covering”, it’s nowhere near 30% complete. The MANY complaints people have mentioned regarding overall usability, and mostly those outside buildings, should not be surprising at all.

Everyone has their own definition of the word “coverage”, but the residents of Portland have a different definition than MetroFi.

Hopefully people can take this opportunity to create community networks in their respective neighborhoods.

Oh and one more thing… Sorry for beating a dead horse!

Personal Telco on YouTube

This is slightly old news, but for those looking for video footage of the group, we have a channel on YouTube now with most of our videos from the past 8 years.

Check us out.

Personal Telco IRC

We just setup a browser-based IRC client that will make it easier for people to contact our volunteers for node support and general wireless questions. Feel free to check it out at irc.personaltelco.net

For more IRC information, check out the wiki entry.

Also note that this may not be supported by certain browser configurations. You can always install an IRC client, such as mIRC, XChat or Irssi if the applet doesn’t work for you.


A recent story about a wifi project in the small town of Philomath, Oregon was noticed by MuniWireless, but in the process they appear to have veered off the road of comprehension. The original story, published in the Corvallis Gazette-Times notes in its first sentence:

A local high tech company is working to provide limited wireless Internet service throughout the city of Philomath for free, possibly by next year.

Reading a bit further:

Internet users could access e-mail accounts, the city’s home page and other local business and information sites [...] Locals would need to pay to receive full Internet access.

Now, that sounds perfectly clear. As a reader, I have an immediate understanding of at least the broad outlines of how the network would work. It is pretty close to one of the early ideas of a “walled-garden” model for the Unwire Portland project. I see that as pretty distinct from the Free that Personal Telco nodes provide, and even pretty distinct from the ad-supported “Free” that MetroFi provides. MuniWireless doesn’t seem to draw the same distinction. In the first few sentences of their reference, we read this:

Free Wi-Fi comes to Philomath, Oregon
A local company in Philomath, Oregon, expects to offer free Wi-Fi services to residents next year.

Now, they go on to describe the actual deal, which gets us back to the “walled garden” description, but conflating that with “Free” just seems weird to me. So much so, that I thought it was worth submitting a comment, which thus far (about 24 hours later) has not been approved:

Carol, your use of the word “free” here is an odd choice. The english language is filled with perfectly good words that could have described that network, but to use “free” completely debases what “free” should mean. If the Philomath network is “free wifi”, then I don’t see why a network that provides a DHCP resolution without a credit card or a login shouldn’t be labelled “free” as well. It amounts to about the same thing. There *are* free wifi networks. I’d be a lot happier if you reserved your use of “free” for those.


We have fired this up to invite the community to respond to matters in a fashion that seems to be much easier than our other current options (the wiki, mailing lists, irc, etc). Hopefully this will enable more feedback and discussion.

Jason McArthur

Personal Telco Project launches a blog!

Thanks to Jason McArthur for the idea and the implementation. This will provide a venue for commenting on the state of the community networking world.