By early summer of 2009, Medicine Bow had been told that the local incumbent phone company was going to apply for a Community Connect grant, and would then install dsl for the town, establish a community center with public access computers and provide a couple of jobs within the center. The condition for the town was that they could not allow anyone else to provide broadband in the town or the grant would not be approved, thus the city council had voted to deny my company access to their water tank for the placement of a wireless access point.
After studying the terms of the Community Connect grant and talking to the person in charge of the program for Wyoming, I determined that the deadline for approval of the grant was August 15th. Armed with this knowledge, I went back to the City Council and came up with a compromise deal. I was willing to wait until September, well after the grant was approved or disapproved, before deploying broadband. This was a win/win situation for Medicine Bow. If the grant was approved, they would have their community center, jobs and two choices for broadband. If the grant was not approved, they would at least have broadband. We finally had the green light, now we just had to figure out how to connect to the water tank, 55 miles away from the nearest Vistabeam tower with a couple of sizeable ridges in between.
Going the direct route from Laramie was going to be a challenge. As at least two intermediate towers would be required and there is very little population between Laramie and Medicine Bow. We did have equipment on a tower in the mountains east of Laramie at about 9000 foot elevation that could see as far as Rock River, 20 miles from Medicine Bow, which would leave us with at least one 40+ mile backhaul link. Typically, most of our backhauls are in the ten to twenty mile range, but we did have one 65 mile shot running from the mountain east toward Scottsbluff, and many other WISPs had done long links as far as 82 miles so it had been done before.
Looking at a wide area outdoor broadband network from high overhead on Google Earth, it is a series of dots on a map with lines connecting the dots. The dots are the access points and backhaul towers, and the lines are the microwave links between the locations. Cell phone companies typically use expensive link planning software to plan and budget their backhaul links. Fortunately, an excellent open-source package called Radio Mobile was available for doing less sophisticated link planning. When looking for tower locations to connect two points, it is possible to put in the GPS coordinates and tower heights then run a model to determine locations where both of the end points are visible. I had used this feature several times in the past to find suitable tower locations. I put the information in with the water tank visibility in yellow and the mountain’s visibility in blue. The program generated a map with a few scattered green areas where we could connect the two locations. Most of the locations were ruled out because they were either inaccessible or on the tops of mountains, but there was a green area on a ridge just west of Rock River that looked promising.
I fired up Google Earth and zoomed in on the GPS coordinates of the green area to see if I there might be anything there that would work. The ridgetop was mostly populated by wind generators, but there appeared to be a couple of towers on the hill and one large, indeterminate structure. A glance at the FCC database showed that a utility company owned one of the towers and the other tower appeared to be too small to be of use. On the way home from a day working in Laramie, I decided to drive up to the ridge and see what the big structure was. Much to my surprise, it turned out to be an intact AT&T Long Lines facility, with the coax, antennas and racks still inside. An inquiry with the landowner determined that it was abandoned by the previous owners and was now ranch property, so I negotiated a deal to put our equipment on the tower.
While inspecting the tower, I noticed writing on the cables coming from the antennas that indicated the destination points. One end pointed to Crow Creek, which was the old Long Lines name for the upstream tower that would feed this one. The other end pointed to Hanna, just down the road from Medicine Bow, where a large structure identical to the Rock River tower stood on a hill above the town. I contacted the owner of the Hanna tower, negotiated a rental agreement and now our project included the towns of Medicine Bow, Hanna and Rock River – giving us a total population between the three towns of about 1200 people. I found it ironic that 20+ years after the Long Lines system was dismantled and many parts of it abandoned, it was coming back to life bringing connectivity to rural areas. Our network includes six of the former Long Lines towers, and there are many more across the US that are home to WISP backhauls and access points.
With the towers lined up and the RF software indicating that it would work, I put together the equipment list. Many WISPs use equipment from well-known vendors such as Motorola, Cisco or Alvarion, but there is also a large contingency of operators that use equipment that is interoperable and based on the 802.11 standards. I standardized on 802.11-based equipment for all of my deployments in 2001, and had good luck with a linux based wireless operating system called StarOS. StarOS was originally written by a group of Canadian software programmers that had also started their own WISP. True to WISP form, they didn’t like the commercial products and decided to write their own software. After nine years of development, it had matured into a very reliable and flexible product that was capable of serving the multiple roles of backhaul, access point, router and bandwidth controller. The current standard radio that we use is called the X-4000. It is a 533mhz CPU with four wireless cards and a price of about $350US. Each card can run in either backhaul or access point mode, so we would need only a single X4000 on each tower site.
For customer premise radios, we decided to utilize a mix of Tranzeo and Ubiquiti units. I had been using the Tranzeos since 2001, and watched the product line evolve and get consistently better over the years. The Ubiquiti radios were newer and did not have the track record of the Tranzeos, but were inexpensive and worked better in situations where an external antenna was needed. Each type of radio was in the $100 price range, which would help to keep the installation costs down for new customers. With no other outdoor deployments in the area that could interfere with the system, I decided to use the 2.4ghz spectrum for the last mile portion of the service. By running Atheros chipset radios on 10mhz channels, a 2.4ghz deployment can deliver broadband speeds to about 50 customers per access point and maintain a very strong resistance to interference.
The final hurdle that we had to overcome had nothing to do with technology. Medicine Bow is a three hour drive from our office in Gering, and a lot of our potential profit would be eaten up in gas and six hour round trip “truck rolls” to handle installations and service calls. Although there was no shortage of potential installers in Laramie, the State of Wyoming requires that satellite and wireless Internet installers have a low-voltage electrical license. Four of our regular subcontractors already had these licenses, but the cost of labor for them to do the work in Medicine Bow would be double the regular rate. Obtaining the low voltage license requires several hundreds of hours of experience working under a licensed electrician, so few people have them. Ironically, cable and telephone company employees are exempted from these licensing requirements.
By a stroke of luck, a new customer in Laramie asked if we were looking for installers and was interested in doing the work for us. He had experience with tower climbing, worked under an electrician in Laramie and was good with computers and networking. He also had a great personality and a strong work ethic. Of all the different variables and hurdles that we had to deal with to get broadband to Medicine Bow, finding a qualified, capable and hard working person to handle the installs and service work was the last piece that needed to fall into place to make things happen.
Everything was ready to go. It was time to put the gear in the air.