PA-ComNet Meeting May 2, 2001

The meeting began at 7:37 AM with Keith Cooley describing how the local loop has changed in recent years. It now means the whole communications facility for an area. It typically is 18 to 25 year old copper wire. Copper is being used for broadband data service, but DSL is reaching the limit of bandwidth capacity. And has limited expansion capability.

Things have changed significantly in communications. For over 100 years there was basically 1 owner of wired voice service. The phone system was 95% owned by AT&T until 1984, after which the system was broken up into local companies plus AT&T. In 1996 the law was changed to allow CLECs to use the phone companies local loop. The intent was to create competition for local phone service like the competition that developed for long distance. Despite this opening of local service, 88% of local loops are owned by the Bells. For residential 97% is owned by the Bells, for business they own about 83%.

A new local loop installation completely bypasses the Bell system for local users. Teleport started this in New York using fiber to connect businesses successfully. By 1995 the alternate phone companies were real competitors, inspiring the Bells to invest in their own fiber. All of the activity and competition was for businesses, none of it was for residential.

After the 1996 Act was passed companies promised to provide DSL to residential customers using leased lines from the Bells. Deregulation actually has been counterproductive because it inhibited investing in alternate loops, and encouraged investing in DSL, which is transitional and doesn’t scale well. Fiber is the ultimate need in order to create a proper local loop.

An alternative local loop should be installed with a fiber infrastructure. Other technology such as wireless using local lasers is possible too. Comparing costs for the various options, investment in fiber is the best bet now. The fiber that is installed should offer open access by allowing many users to lease fibers or to lease wavelengths on a single fiber. This encourages competition and lowers the barriers for entry. Co-location of equipment in the same space is important because it reduces costs and makes maintenance less expensive. The exchange or peering points for Internet access must be well-defined along the network. Encouraging networks to peer eliminates most interchange costs.

In the late 90s there were a number of companies building dark fiber networks. Some like Quest and Level 3 make it harder for others to access their dark fiber by requiring long term up-front payments. All of the fiber networks designed the routing to serve major companies, not residential except by accident.

Utility companies also have dark fiber networks, but they are not well used. The best bet for residential fiber networks is to provide it with municipal systems. In Sweden every municipality has installed dense dark fiber within a km of every home. Other cities in California as well as Palo Alto are installing dark fiber rings. The Palo Alto ring would be better if it had more entry points, was closer to more homes, had a co-location facility already set up, etc.

Chicago is doing a very good job of setting up a dark fiber ring which offers inducements for private companies to provide service. There are competing agencies which use local loops that may cause complications - police, schools, transit, and the city all have fiber systems. The network is designed to be all IP. Ameritech’s facility and operating system is in poor shape, and they have been fined heavily each month for poor performance. Local fiber distribution is possible via the large storm drain system or overhead lines that run in back yards all over Chicago. It should be relatively cheap and easy to install a public fiber network in and around Chicago.

The presence of dark fiber in a city is important because cities believe it will lead to economic expansion. Palo Alto is held up as an example by other cities. Several successful companies that started in Palo Alto by leasing dark fiber have done very well. It is important to serve both businesses and residences. Some of those bidding to provide these services in Palo Alto are not well known here.

In Davis, they plan to require all new subdivisions to be wired with fiber to the home. Fiber deployment should be dense so that the actual deployment to the end user is a short distance. Telecordia encourages cities to put in dark fiber and Telecordia then will manage it for the city.

Municipal fiber is the only way to serve homes. Commercial fiber providers have shown they are interested only in business uses. A municipal fiber deployment must show how the money will be spent, and who will benefit from the cities’ investments.

In Italy the city of Bologna started a joint venture with a Swedish company to provide FTTH. Milan is doing the same thing using a dark fiber loop. Cities need someone to operate the system and assure open, equal access to make it work. It’s unrealistic to expect cities to have lots of foreign investors and not have the investors get a good return on the investment.

Keith is putting a team together to provide the operation and service of the local loop, but he needs a larger experienced partner. There are several possibilities that he is working with.

Palo Alto didn’t put in enough density into the present fiber loop. Downtown almost all 288 fibers are leased already. Other areas, particularly residential, have far fewer users and the fibers are not as heavily used. Lease rates for fiber are at the high end, but costs for multiple fiber lease and maintenance costs are reduced form the initial rates.

Meeting adjourned at 8:30 AM

Submitted by Bob Moss