Telecom network architecture in 2004 is very much a child of many years of growth and technical evolution driven by telephone calls. Telephone calls require 64 Kbs and are point-to-point connections used for short periods. Telephone service economics where the longer the distance between two points of a call costs more money influenced most of the traffic into local and regional calling patterns. These regional calling patterns led to hierarchical network architecture with high capacity switch sites in fewer numbers piecing together circuits to enable regional and national telephone calls.
Figure 1 is an illustration of the hierarchical nature of the telephone network.
Figure 1: Classical Telecom Hierarchical Architecture
The access tandems and certain designated transmission facilities are generally referred to as the access network. These switches formed the technical and economic basis for a philosophy the FCC dubbed equal access. Next time you pay a telephone bill, look at the term FCC access charge. Nope, the money doesn’t go to the FCC. It goes into the owner of the access facilities. And according to FCC regulations, the owner of the access facilities in each LATA must afford equal access to any inter-exchange carrier (IXC) willing to arrange for a connection between their facilities and the local exchange carrier’s (LEC) facilities.
On the other side of that equation, equal access means the LEC must offer their subscribers the ability to select their long distance, or IXC.
The higher levels in the hierarchy vary, depending on the size and extent of the individual long distance carriers. AT&T operates much as it did before divestiture. Others, such as MCI and Sprint, have flatter or fewer levels in their switching hierarchy. All the long distance carriers have gateways to international networks and their own shares in international transmission facilities. Piecing together two or more SONET links end to end creates an extension of the physical boundaries of the network with the addition of more links (Figure 2).
Figure 2: SONET Transmission Segment
All the layers are extended across the interface. Each layer becomes an extension of itself each time a SONET transmission facility is connected to another. Linking two or more in series along a path or route, extends the boundary or domain as defined in the facility specification. Figure 2 shows a block diagram of SONET transmission infrastructure linked to provide continuous bandwidth that could be used to provision an E-1/T-1/J1 or E3/DS3/J3 private line. Such a facility could be used as a link in a data communications network or to connect two routers in an inter-network or intra-network link.
Another method of linking facilities is the mid-span meet, commonly used by carriers to link or interface between networks. This is depicted in Figure 3 and shows links between LECs and IXCs.
Figure 3: Linked SONET Transmission Facility
A variation on the light wave terminal equipment (LTE) is the add-drop multiplexer (ADM). SONET transmission capacity is built around the 51 Mbs STS. An LTE breaks out all the STS in the optical carrier (OC) and makes them available. If the LTE is used at a location where all STS are not terminated in local routing/switching equipment, then any transit streams must be connected directly to other LTE facing other routes. The ADM breaks out and terminates a limited quantity of STS from the optical carrier. For example, the office in which the terminal is located might only require 1 or 2 STS from an OC-3, OC-12, or higher rate transmission facility. Using an ADM would be less costly in terms of capital investment and operations expense.
