The Fiber Initiative meeting on Friday was designed to look at how to get better Internet access in the Fort Scott area. Fort Scott is well behind many of the nearby cities. Here is a list of cities along with the highest bandwidth package available for residential users:
- Frontenac – 35Mbps
- Pittsburg – 35 Mbps
- Girard – 15 Mbps
- Uniontown – 15 Mbps
- Iola – 35 Mbps
- Nevada – 25 Mbps
(thanks to Nick Graham for researching this)
In Fort Scott AT&T DSL offers up to a 7Mbps (download) / 768kbps (upload) connection while Suddenline tops out at 3Mbps/512kbps. In some places AT&T is only available at lower speeds and in many others it isn’t available at all.
Right now the only bandwidth on fiber optic cable that comes into Fort Scott comes from AT&T. The only alternative for people who need significant bandwidth is some form of wireless. This is what FSCC is using–a microwave link provided by Cox. Unfortunately these types of connections are not particularly reliable and they tend to cap out at much lower speeds than what can be achieved over fiber optic.
Suddenlink says they can do fiber runs in town that make use of the bandwidth they purchase from AT&T. It appears a 5 Mbps connection would cost around $1,300 per month. (For those of you in the meeting who heard me say that it would cost $3,000 per month, I mispoke and realized my error after consulting my notes from last year.) For comparison purposes the college is paying around $3,000 per month for a 45 Mbps connection and a significant portion of those costs are for KanREN network management services–the actual bandwidth is closer to $1,000 of that.
There are two basic problems. First, the limited bandwidth coming into Fort Scott is very expensive and only available from AT&T. The second problem has to do with actually getting bandwidth to the houses and smaller businesses.
To solve the first problem, we need to get more fiber running into Fort Scott. It appears that at least Cox and Quest have fiber runs that follow the railroad through Fort Scott, but there are no “junctions” so no one can tap into it here. The previous meeting was attended by Cox and there was discussion about trying to get a “junction” to allow businesses in Fort Scott to tap into the bandwidth from Cox. It sounds like the cost to dig up the fiber, put in a building with a junction, etc. would be roughly somewhere between $30k to $100k.
Larger cities have what is known as “carrier hotels” where anyone can buy bandwidth from multiple providers. This helps drive the price down because carriers can quote prices based on just the cost of bandwidth–not the cost of building out to various buildings. This type of arrangement also means there is more competition to drive down prices and allow companies that need redundant connections the ability to easily provision connections to multiple carriers. The client is then responsible for getting the bandwidth to their datacenter or office. The clients in this type of setup would be ISPs and larger companies that can afford the build out to move the data from the “carrier hotel” to wherever it is needed. Fort Scott needs to be able to get reasonably price bandwidth into the city before it is going to be cost effective to provide reasonably priced bandwidth to individuals and businesses in the area.
Better bandwidth to Fort Scott from multiple carriers could make the city a more attractive place to put datacenters, call centers and other bandwidth intense businesses.
The second problem has to do with how to get the bandwidth to the individuals and businesses. This is often refered to as the “last mile” problem. The two big players in this market with wired connections are AT&T and Suddenlink. RTS and Valnet offer some wireless connections in and around Fort Scott.
AT&T DSL service only works with wire lengths less than about 14,000 feet from the central office. While this seems like a long distance in Fort Scott, wire doesn’t go in a straight line and wasn’t originally installed to minimize the distance back. Also older infrastructure often can’t support the needs of DSL. There are many situations where one person has a good DSL connection, but their neighbor across the street can’t get service. In repeated calls to AT&T it does not appear that they are interested in building additional infrastructure in Fort Scott to support better connectivity to homes. (Although they may change their mind with pressure from someone other than me calling in as an end user.)
Suddenlink says they plan to eventually support 10 Mbps connections in Fort Scott, but doesn’t have a timeline for any upgrades. When I spoke with them in August 2010, they said they might start offering a premium tier in the first half of 2011. It sounds like they are hesitant to increase the bandwidth because their current network design within Fort Scott, is not adequate to handle the existing traffic much less traffic that would come if their speeds were greater. Suddenlink says that 50% of the Fort Scott traffic in the evenings is from Netflix. Interestingly they have their max bandwidth right below what it would need to be for HD Netflix streaming which would consume significantly more bandwidth.
From a technical perspective of what is possible, DSL typically maxes out at about 15Mbps while normal cable modems can typically support speeds up to 50 Mbps. (There are all kinds of different technologies to make this go faster, so I’m talking about typical networks with typical equipment.) Suddenlink does have some networks where they have 50Mbps and even 107Mbps connections. However, it appears that they only offer this in markets where they are trying to match speed with competition. AT&T DSL coverage is spotty enough that it doesn’t seem to be a big competitive driver.
The places that are offering extremely high speed connections are doing it with fiber. Google is planning to offer 1000Mbps connections in Kansas City in the near future. The price hasn’t been set, but it will probably be $30 to $70 per month. Chattanooga offers a 1000Mbps connection for $350 per month. They also have a 50Mbps plan that is priced in the same range as Fort Scott’s options.
I think the current plan is to come up with a plan for where the city wants to be at certain points in time and then see if the current providers are interested in helping to implement it. For example, here are some very rough ideas of possible targets from looking at what other cities are doing:
In 1 Year:
- 10Mbps/2Mbps connections for less than $50 per month.
- 3Mbps for less than $35 per month.
- At least one fiber bandwidth provider other than AT&T
- Free Wifi at Library and downtown
In 5 Years:
- 100 Mbps connections for less than $100 per month.
- Network reliability sufficient to run 911 services over the connections.
- A “carrier hotel” with at least three carriers providing bandwidth.
In 10 Years:
- 1000 Mbps connections for less than $100 per month.
Keep in mind, this isn’t a plan that is being proposed, it is just my quick attempt to demonstrate what a plan could look like.