Red Pollard: Everybody thinks we found this broken down horse and fixed him, but we didn’t. He fixed us. Every one of us. And I guess in a way we kinda fixed each other, too.
– Seabiscuit

At last night’s monthly meeting of Portland’s Personal Telco Project, Russell Senior and Caleb Phillips described their independent performance evaluation of MetroFi Portland. MetroFi is the city’s municipal wireless contractor.

Portland’s phase one installation covers only about 5% of the 134 square mile installation. By May 2007, MetroFi expects more than 55,000 Portland residents and 10,000 businesses will be able to connect to it’s WiFi network and plans to finish building out the network sometime next year.

Portland Oregon’s municipal wireless network is being overseen by the city’s Logan Kleier, Project Manager, for Unwire Portland (right).

Currently about 72 SkyPilot, access points have been activated, covering roughly 2.5 square miles of the Phase I coverage area. The Portland deployment uses about 25 SkyPilot access points per square mile. Each of the pole-mounted access points are designed cover a 500 foot radius (with a 1000 ft diameter). A 5X5 matrix of canisters covers roughly one square mile.

But mileage varies greatly — it IS unlicensed WiFi and subject to interference, foliage and multi-path.

In the first-phase, “proof of concept” area (downtown and central city), they found that MetroFi’s network worked about half the time. According to the contract, MetroFi needs to show at least 90 percent coverage (outdoors) before it can go on to phase 2.

Senior and Phillips lost a bid to do the city’s evaluation, but they expressed no ill feelings towards Uptown Services which won the bid to test the system. Uptown’s report is due April 12.

The Personal Telco volunteers opted to do testing anyway — for free — out of curiosity. Russell Senior had already assembled a monitoring system built out of a Netgear access point and GPS hardware. They had the computer randomly select several dozen spots within the phase one area for testing.

Of 39 random spots placed within 500 feet of access nodes, they were able to connect 20 times, and failed 19 times. When they tested between the 500- and 1,000-foot zone, they were only able to connect 12 times, failing 27 times. Overall, that’s a failure rate of about 60 percent. Not so hot.

Still, the results are preliminary. Senior and Phillips will release a full analysis later at their website - Unwirepdx-watch.org.

Meanwhile Uptown’s Neil Shaw has produced a whitepaper (pdf) that “should be enough to convince any City Manager or CIO that they should definitely include third party testing as part of their future RFP and/or contract for citywide Wi-Fi services”. Uptown Services says they gather performance data every 100 feet across the advertised service area. Parameters include coverage, data throughput, delay, packet loss, and loss of entire files.

“Personal Telco’s own unofficial testing effort is clearly not independent or impartial,” said Metrofi spokeswoman Denise Graab in a written statement emailed to kgw.com. MetroFi advises consumers use an outdoor Engenius EOC-3220 ($170) with an integrated 9dBi directional patch antenna and 400mW radio when 500-1000 feet from an access point.

“If the study suggests accessibility figures are not above 90 percent, we would tell them to fix that,” said Logan Kleier, Project Manager for the city’s wireless network.

Also at last night’s meeting, Michael Burmeister-Brown, of the nonprofit NetEquality, explained how they use the $50 Meraki box to bring wireless Web access to low-income housing.

Michael Burmeister-Brown, who is a co-founder of Central Point Software and a key developer of Yahoo Messenger software, has worked closely with MIT’s RoofNet group and their commercial spin-off, Meraki.

Net Equality installs wifi repeaters inside of the apartments or in the halls. Signals need only to penetrate interior apartment walls, not external concrete walls or weatherproof glass, improving propagation.

NetEquality likes to bring in broadband to a low-income apartment complex via wired DSL. Then they use a $50 Meraki box to repeat the signal. One $40/month DSL line can be shared by a dozen or more aparments by relaying the broadband signal through 3-4 Meraki access points that mesh together using the RoofNet protocol.

NetEquality played a key role in developing Meraki’s backend management software, called Dashboard (above), which maps access points on Google Maps.

Net Equality’s management backend, developed in parallel with Meraki, is open-sourced and designed to be simple and maintaince-free. The goal, explained Michael Burmeister-Brown, was to provide a simple, reliable solution for low-cost internet access. NetEquality targets low-income communities. It’s designed to be a stable, inexpensive, easy-to-use solution for property managers and end users.

With hundreds of apartments in the Portland metro area now being served by NetEquality’s Meraki solution, they appear to be successfully accomplishing their goal.

What if MetroFi doesn’t cut the mustard?

Why couldn’t Portland make a Crazy Salad? Imagine solar-powered Meraki minis inside giant yellow pencils. Call it public art.

Mount them on streetlights. On coffeeshops. Everywhere. The stand-alone, self-contained “WriteSpot” repeaters might cost under $200 and provide a simple, reliable solution for areas with poor signal reception. Local content on flash memory.

The message is iconic and positive — your ideas have value.

Communicate.

More coverage of last night’s PTP meeting is available at Portland Mercury, Willamette Week, the Oregonian and Unwirepdx-watch.org.

Posted by Sam Churchill on Thursday, March 29th, 2007 at 11:05 am.