Circuit- vs. Packet-Switched Networks: Which Does Your Enterprise Need?
Circuit networks have connected telephone systems since 1878, when they first came into general use. These networks are direct, point-to-point systems that make one connection between an originator and a destination in a linear fashion.
In circuit-switched networks, signals will usually pass through several switches before the final connection is established. When a circuit-switched network is creating a connection, no other network traffic will be able to connect via the switches that are in use. This is a hindrance to certain types of communication, which is why packet-switched networks have piqued the interest of enterprises.
While telephone calls are served well by circuit networks, packet-based networks are designed for computer-to-computer communications. In this type of network, messages are broken into small data packets that seek out the most efficient route as data paths become available.
Because each packet can effectively choose its own route, the packets that comprise a message may travel on different routes to their destination. The header address tells the packet group making up the message where to go and describes the sequence for reassembly when it reaches the destination computer.
These packet-switching networks are used to connect computers rather than the chatter of people — with the exception of voice over IP (VoIP). But even that, at its heart, is just connecting one computer to another.
Switching Circuits in a Network
Circuit-switched networks reached their heyday when brick and mortars with the Bell System logo on the outside were popping up across the U.S. The idea of data has changed, however, since circuit networks were first used. The data wasn’t the variable voltage that came off a carbon mic and represented human speech; that data was analog, not digital. Although the data voltage gets switched with relays — which may have some digital functions applied — it remains analog throughout the network.
Enterprises Get Comfy With Hybrid Networks
As such, the trend these days is away from circuit-switching networks. Joel Maloff, president of consultancy Maloff Group International, summed it up well when he told Computerworld, “A circuit-switched network is good for certain kinds of applications with limited points to go to. If you’re doing voice applications solely, it’s great. But if you have multiple locations to get to and large amounts of data to transmit, it’s better to break it down into packets.”
That’s not to say enterprises will be able to switch to solely packet-switched networks tomorrow. Private branch exchanges are still there and working, which means enterprises will likely use hybrid networks for years to come. After all, while circuit-switched networks are limited in what they were designed to do, they do voice well and can alleviate IT departments of the hassle of setting up the IT infrastructure required to send voice over data networks.
Both of these network types have their benefits and their scope. Integrating both seems to be the best way for an enterprise reap the advantages of each and transmit all types of data efficiently — at least for the foreseeable future.