Packet Switched Networks

Packet Switching
Image A


What are packet switched networks exactly? Well, they are networks that send and receive data in the form of packets. When a document is sent, it is broken up into a series of packets that make usually contain around 1,000 or 1,500 bytes of information – packets can also be referred to as frames, blocks, cells, or segments. These packets then are sent to their destination via the best available route. Since the introduction of packet switching, it has gained widespread acceptance over the conventional circuit-switched network.


Paul Baran
Image B - Paul Baran
The idea of using small “packets” of data was first introduced by Paul Baran in the early 60’s. However at about the same time, Donald Davies in the UK developed the same basic ideas independently. Additionally research in queuing theory – would be important in packet switching – was being done by Leonard Kleinrock. Baran came up with this concept while working at the RAND Corporation for the United States Air Force. In 1961 Baran briefed the Air Force on his concept, and late in 1962 and 1964 published papers on message block switching. Baran’s concept helped to influence Lawrence Roberts to adapt the concept into the development of the ARPANET, the first packet-switched network. 

The Advanced Research Projects Agency Network (ARPANET)

The first ARPA Network Nodes
Image C - First ARPANET Nodes
The ARPANET as previously stated was the first operational packet switching network. It would eventually become what is now known as the global Internet. The ARPANET was originally funded by ARPA – later DARPA – within the US DoD for use by universities and research laboratories. It was based off the designs of Donald Davies and Lawrence Roberts. ARPA awarded the contract to build the network to BBn Technologies in 1969. The BBn team proposed a network that used Interface Message Processors to connect local resources. These would later become routers. These Interface Message Processors would be interconnected with modems that were connected with leased lines. The ARPANET was designed so that it could communicate even with subordinate-network loss. Common myth says that it was specifically designed to survive a nuclear attack. The original ARPANET consisted of four Interface Message Processors. These were located at the University of California, Los Angeles, the Stanford Research Institute, University of California, Santa Barbara, and the University of Utah. The first message sent was the word login. However, the l and o were transmitted, and then the system crashed. So the first thing sent over ARPANET was lo. 

How it Works

In general all packet-switching networks have the following:
-all the data is divided into short chunks of data that have a max length
-the packets usually contain a header for addressing and contains error control information.
-the packets move very fast between nodes
-node computers do not store the data, after received an error check is performed to make sure it arrived at the right place, and then the node deletes the data.
Packet switching requires a data switching exchange or DSE. A DSE a network node that links three or more paths. The packets move from DSE to DSE in a way that lets multiple packets to pass through the same path in a consecutive time sequence. Breaking the information into multiple packets, allows for multiple users to send data at basically the same time as opposed to the early circuit-switched networks.

Circuit-Switched Networks

A Phone Switchboard
Image D - A Switchboard

Ok, imagine a switchboard like the one im Image D. Back when these were used you would pick up the phone and reach an operator. You would ask the operator to patch you into whoever you wanted to talk to. If they were local, you could be directly patched through, however, if they were not the operator would have to go through another operator where the person who you wanted to call lived. In order to do this there would have to be a direct connection between the operators. This meant there would have to be a large cable running for miles. Early on, networks used a very similar system. But this only allowed for one person to be connected at any one time. Also people would often have to have multiple terminals that would be connected to different networks. With the introduction of the packet-switched network multiple people were able to be connected at the same time. The connections would only be used when actually sending a packet.


X.25 is an example of a packet switched wide area connection or WAN. This network has packet-switching exchange nodes with physical links consisting of leased lines, telephone cables, or ISDN connections. The public data network was a common named used to describe the collection of international X.25 providers. It was commonly used during the 1980’s and 90’s, covering a large amount of the globe. X.25 provided the first international and commercial packet switching network.


TCP – Transmission Control Protocol – is an interworking protocol that uses packet-switching nodes. It is one of the main protocols of the Internet Protocol Suite which is often referred to simply as TCP/IP.
This is the protocol that is used by major Internet applications. This includes the World Wide Web and email.

Connectionless and Connection-Oriented

Wireless Router
Image F - A Wireless Router
Networks that use packet switching can either be connectionless or connection-oriented. Examples of connectionless networks include Ethernet and IP, while connection-oriented examples are X.25 and TCP. In connectionless networks, each packet is labeled with an address for its destination and source aw well as the port number, or a sequence number. This creates the need for a dedicated path to allow the packet to reach its destination. This also causes the header to contain much more information. Then after the packets reach their destination, they are reassembled in order based on the sequence number in the header.
In connection-oriented networks, packets are labeled with connection IDs instead of addresses. Address info is only transferred during a connection set-up phase after the route taken by the packet has been determined. Connection-oriented networks contain much smaller headers because they only have to contain the connection ID and some other information.


Packet Switched Networks break up information into smaller packets that are then transmitted. Each packet usually contains a header, the body, and an error check. They allow for information to be sent quickly between nodes, as well as allowing multiple connections at the same time. Packet Switching plays a major part in network communication, and helped connect the planet.

Works Cited

Image A -
Image B -
Image C -
Image D -
Image E -
Image F –