Most of the time, when our Inveneo engineers are setting up WiFi antennas it’s to link schools or villages in rural Africa. Sometimes, though, we get to do a little networking right in our Bay Area backyard.
Testing the 24 dBi grid antenna
The Challenge: Test the Performance of Two Antennas Across the San Francisco Bay
Yesterday, we loaded up two teams with wireless networking gear and sent them out to locations on opposite sides of the San Francisco Bay. The day’s goal was to compare performance of two antennas: the Teletronics International 24 dBi grid dish and the ARC Wireless 19 dBi patch for use in an upcoming deployment.
Prior to the test, we used radio link analysis software called Radio Mobile to identify locations with similar conditions to what we expect to see in the field for this project. We needed a 30 km WiFi shot from one point high on a hillside to another with lowland in between.
Our site on Google Earth
Once we had prospective test sites identified, we used Google Earth to make sure there were actually roads leading to the spots we wanted to reach. For permanent installations, we’re glad to hike in equipment and coordinate installation of power infrastructure (we and our ICIP partners have done so many times).
For this quick temporary setup, though, we wanted to roll up, throw the antennas on tripods, and power the systems from car cigarette lighters using small AC inverters.
With this analysis complete, we had our sites: a secluded spot near the top of a ridge in San Bruno and a location in the Berkeley Hills.
The Equipment: Ubiquiti Antennas
For one end of the test link, we used a Ubiquiti PowerStation2-External (400 mW) and at the other end we tried out a Ubiquiti Bullet (100 mW). Ubiquiti radios are our wireless workhorses due to their attractive price points, low power consumption, and easy but versatile user interface.
Our mobile testing setup
The type-N connectors on the PS2-Ext and the Bullet let you pair the radios with whatever antenna is right for your needs for a completely customizable solution. We hope to deploy more Bullets in the future, the small form factor is great. So far, we have had a hard time getting our hands on more than one or two. We’d love to try out the Bullet-HP (high power) model, which offers a full 1 W of power.
Once the radios were powered and the antennas were mounted on the tripods, the last step was to aim and fine tune the links. Radio Mobile gave us true bearing from one station to the other, but as we’re on the west coast of the US, we have significant magnetic declination to compensate for. Depending on where you are, this difference between true north and magnetic north can really throw off your antenna aim.
In most of Africa, declination is less than 2 degrees and isn’t significant for aiming with a handheld compass. In San Francisco, though, a declination of 14 degrees east makes an rather large difference. This handy Google maps mashup will let you check your own region to know what you’re dealing with.
The Ubiquiti antenna alignment tool
After coarse aiming by compass, fine tuning the aim one antenna at a time is easy with the Ubiquiti align antenna tool. It gives a clear, continually updating display of received power in dBm as you make slight adjustments to the aim to maximize the reception.
Our Result: Success!
With field testing, it’s all about results. The main goal of this test was to determine whether a patch to patch solution was appropriate for a 30 km link under these conditions, or if it would be necessary to upgrade one or both ends of the link to a grid dish. We try to use patches whenever possible because they’re easier to mount and position, but we never want to skimp on link margin where a dish is necessary.
The numbers told a pretty clear story: dish to patch gave us consistent received power of -67 dBm and a steady 11 Mbps link while patch to patch gave us -76 dBm at best with a 1 Mbps link.
Given these results, we’re definitely going to be deploying a dish on one end of the link for our upcoming project. We might recommend a patch to patch solution for anything under 20km, but with greater distances we’ll be swapping out one end with a dish for stronger signal.
Sorry, the comment form is closed at this time.