Circuit and Packet Switched Networks

A circuit-switched network is one where there is a dedicated connection between source and destination, such as a telephone line designed to handle voice calls, which is dedicated to a particular single application for the duration of the activity. These were designed in the late 1900s, according to Copeland (2000), however in computer networks a circuit switched network can perform communication either by permanently static connections or ones that are established and used as needed by the network controller, which could be switching connections on a constant basis, however only a single point-to-point connection and used exists on each connection no matter how many times or how often the network may employ switching; the existing link is not shared by any other application in its duration.

A packet-switched network allows communications over a connection to be shared by different applications by transmitted, as the name suggests, packets of data which can be “routed, combined or fragmented, as required to get them to their eventual destination” Kozierok (2005) where the receiving device can use the information and data contained in the packets to reassemble the communication into meaningful and hopefully complete data. Thereby a packet-switched network can share communication lines.

As a result of their very nature, circuit-switched and packet-switched networks have quite specific uses and therefore each has their own advantages and disadvantages for intended uses however the main conclusion is that circuit-switched networks offer a performance guarantee in terms of bandwidth, barring no unforeseen technical issues, due to the single use of the communication line whereas packet-switched shared lines cannot offer such a guarantee. However conversely, packet-switched networks offer a greater flexibility due to their multi-use capability. In my experience this is evident when considering where to purchase fixed leased lines for external organisational communications versus the higher contention ratios shared services and often the difference in budget would often win out unless there were any special security considerations that over-ruled the budget.

Time Division Multiplexing (TDM) and Frequency Division Multiplexing (FDM) in circuit switched networks are techniques to provide multi-use applications of communication lines but to ensure that the required single connection status us preserved by preventing collisions of data. The main difference between TDM and FDM is the way in which they allocate bandwidth with TDM allocating time slots and FDM allocating frequency channels, as their names suggest. As a result the main disadvantage of FDM is that data has to wait for its frequency channel to become available, causing delays, while other frequency channels remain unused and resources are wasted as a result. Levine (2001) concluded that “results from queueing theory state that systems with multiple queues for multiple servers, as in fixed assignment protocols, engender longer delays and lower utilization than single queue systems” and also noted that FDM can be affected by interference which can cause data loss problems. Overall TDM provides a faster, less wasteful and more reliable circuit-switching multiplexing.

References

Copeland, L (2000) Packet-Switched vs. Circuit-Switched Networks [Online]. Available at http://www.computerworld.com/s/article/41904/Packet_Switched_vs._Circuit_Switched_Networks (Accessed 9 Apr 2011).

Kozierok, C (2005) Circuit Switching and Packet Switching Networks [Online]. Available at http://www.tcpipguide.com/free/t_CircuitSwitchingandPacketSwitchingNetworks.htm (Accessed 9 Apr 2011).

Levine, G (2001) CONTENTION CONTROL IN MULTI-ACCESS RESOURCE SYSTEMS [Online]. Available at http://alpha.fdu.edu/~levine/contention.txt (Accessed 9 Apr 2011).