ECV Wireless: Implementation
From LTWPwiki
Contents |
Status
Update: June 26, 2007
I conducted connectivity tests at ECV:
Baby step. I'm using Buffalo brand wireless routers (model WGR-HP-G54) that support a wireless distribution system. I positioned one in their office building as a primary (for Internet connection) and a secondary near the perimeter of the village. I attached a computer to the primary to pretend to be the Internet. Then I walked around with laptop to check signal strength and run benchmark file transfers to check speeds. Results were good enough that I believe we can cover the village (distance-wise) with 5 or 6 of these boxes.
The bad news is that I had to move the secondary box outdoors to make it happen, and it's not weatherproof. It's not clear to me whether we can position it adequately (like in an attic crawl space), or whether we'll have to supplement them with different omni or directional antennae. I'd like to stick with these boxes because they are cheap and off-the-shelf at Best Buy. I'm pondering this now.
See ECV Wireless Benchmark Test 1 for detailed report.
July 6
I returned to the site to check for antenna mount locations. There are numerous spots that look good. Some already have TV antenna mounts on them.
July 23
I met with Fred Vereen last week to discuss two proposed mounting locations and he confirmed that he'll be able to get it done. I ordered exterior omni antennae with associated mounting brackets, cables and connectors for two units. Our plan is to put two routers in place, assess service quality, then seek support and funding from other sources in order to build out coverage to the rest of ECV. Cost for all parts at this point is about $625 for positioning the two routers with exterior antennae. I saw five more routers on the shelf recently at Best Buy for $55 each, a $15 price drop.
Update: August 4, 2007
Yesterday I worked with Tonia Moore and Fred Vereen to install an antenna on the community building. The mast bracket attached cleanly to the fascia board near the roof peak with four 3 x 3/8" hex-head galvanized screws. The screws reach through the boards to the exterior wall behind. Drilled a 1" hole into the attic, inserted the antenna cable with N-connector leaving drip loop and caulked it. Need to return to replace ratty twist ties with proper cable ties. The mount seems reasonably secure. Maybe not category-3 hurricane-proof.
We ran out of time to mount the router in the attic, but we identified the final steps. I have ordered a 50 foot long cat6 cable to run from the router to the probable 'net connection point in an office below. There is existing Comcast service there which we will connect to for testing in the interim. AC power source undetermined; there are a few possibilities. The attic has AC for lights on a switch, but I don't know if there is an unswitched outlet up there.
I expect to return to complete this first installation within the next two weeks.
Update: August 13, 2007
We completed installation of the primary wireless router. It was mounted on a post in the community room attic. We attached a 50 foot Cat 6 cable to the WAN side and a power supply/AC cord that we ran down to an office below near the existing Comcast service connection point. Functionality was tested briefly with the Comcast service, demonstrating Internet connectivity via the Buffalo wireless service.
We walked around the village with a wireless laptop to check signal strength. I did not conduct data transfer speed tests, but simply observed signal strength readings displayed by the Atheros interface program on the laptop. Signal quality was noticeably improved as compared to earlier tests against the primary station without the outdoor antenna (as expected). It is likely that service will be adequate in many nearby apartments without any relay.
Next specific step will be to try a relay router in the community room where earlier tests were unsuccessful. It might be unnecessary to install an outdoor antenna there. Generally, we will try locating additional routers and conducting speed tests in order to formulate a plan for final placement of routers and/or outdoor antennae.
Update: September 3, 2007
Over the course of the past few weeks I worked with Tonia Moore on several occasions to select a location for a second router and to install it. The installation was completed August 31. We placed another antenna on a roof peak at the end of the westernmost corner of the the C cluster, the northern cluster of buildings. The router was installed in a utility room in the corner of the first floor. The 5 meter antenna cable was too short, so I purchased a 10 meter cable.
This morning I drove around all the parking areas to test signal strength with the wireless radio in my laptop. Signal quality ranged from 2-4 (5 scale) from one or both routers in all locations except parts of D cluster. I think that we might already have achieved adequate coverage for most of the apartments in the complex, but we won't know until we actually test inside apartments.
I also corrected a configuration error in the WDS array. I was unable to connect to the Internet, and I don't know why. The base router is connected via the WAN port to a router in the main office, and I could not ping that router. Also, the router reported no connection. I didn't have access to the office so I couldn't investigate. Thus, I did not run any speed tests with actual data transfers.
Current Coverage Map
The aerial photo is overlaid with marks denoting wireless deployment details as follows:
- known common areas where AC power supplies are available and are thus most suitable for router positioning (orange shading).
- Line-of-sight (LOS) notations for signal path between installed, or potentially installed routers(red).
- Distance circumferences at 100 foot intervals from currently installed routers (yellow for base station, purple for C cluster).
