ROUTING | Network Technology

Routing and switching mean different things to different people. For example, both terms apply to switching and transmission facilities. Routing in the circuit switched or voice services world means that a call is routed according to service configuration parameters in a PBX or end office switch. For example, least-cost routing is established when the originating switch is programmed to use the least expensive route between the origination point and termination point for a telephone call. Alternate routes may be a point-to-point private line, a virtual private network (VPN), or the public network, where the private line is the lowest and fixed cost, the VPN is the next least expensive, and the public network is the most expensive.

Wholesale routing of traffic occurs when the traffic is moved from one transmission facility to another. For example, traffic normally routed from New York to Atlanta may go direct, but an alternate route may pass through Cincinnati. A national fiber ring might have a southern route and a northern route, as a regional ring might have an eastern route and a western route. Traffic normally on an Intelsat transponder facility might be moved to another route using a transponder on a PanAmSat satellite.

In the data world, routing becomes more of a technical issue because of the underlying network technology. From a classical perspective, data networks were built using leased or private line facilities provided by carriers on a 24/7 basis at fixed prices. For an enterprise with a headquarters located data center and several field operations, private or leased lines are used to connect computer terminals to the data center. Depending on the number and geographic locations, all the field offices may be connected directly to the data center in a home run arrangement. However, if two or three of the remote operations were physically close to one another, it may make sense to hub them into one common location, aggregate traffic, and connect to the data center over a common facility. In this case the hub becomes a routing point. All traffic from the other nearby locations is routed through the hub to the data center.

Data communications routing can become very complex and confusing because of the proliferation of various flavors of ATM, Ethernet, frame relay, and IP techniques, to say nothing of classic SNA, X.25, HDLC/SDLC, BISYNC, and others.

In circuit switching, routing intelligence is in the user’s head and the network, and is used to tell the network how to route a call or make a connection. The same basic principle applies to ATM, frame relay, and IP networks as well. In the circuit switched network, the routing intelligence resides in the common channel interoffice signaling system and its configuration software. In ATM, the intelligence is included in each cell header and distributed across switching machine configurations. In Ethernet, frame, and IP networks, it’s in the packet headers and distributed across configuration parameters in switching machines swooned over and lovingly called routers—except frame relay, which is typically a meshed, point-to-point arrangement and therefore has limited connection capability.