A packet in computer terms is not that dissimilar from a packet of paper documents, since both involve grouped information. Packets are made up of bytes (units of computer storage) of information. For example, if you email someone, the computer doesn't recognize the email as one entity, but rather in separate packets that consist of bytes. The transfer of packets over a network is called packet switching.
History
Donald Davies and Paul Baran are credited with the invention of packets and packet switching. Davies and Baran realized in the 1960s that sending information over a network would be much easier for computers and networks to handle if that information was in the form of smaller packets instead of one large block of information.
The invention of the packet by Davies and Baran laid the foundation for the way our modern PCs and networks handle and relay information.
Significance
Without packets and packet switching, the speed at which we are used to sending information over the Internet would decrease significantly.
The reason why packet switching is usually faster than if information was sent as one huge piece of data is because packets are easier for networks to read, interpret and then send.
Compare it to the way that we read novels. Few can read an entire novel in one sitting and still be able to retain, process and analyze all of the information. To make novels more readable, they are divided into chapters. In this way, we can read through a section of the book and, if we are interrupted or need to take a break, we can stop at the end of a chapter. When we return to the novel, we can begin at the next chapter instead of starting over from the beginning of the book.
Because data is sometimes lost while being transmitted over the Internet, the advent of packets allow our transfers to pick up at a certain "chapter" of the larger file, rather than starting over from the beginning of the entire transfer.
Function
All information relayed through a network is made up of binary code (combinations of 0's and 1's that take the form of data we see on a computer). Packet switching will take chunks (packets) of binary code and send them to their intended destination. When these packets reach their intended destination, the network will then put the packets together in the correct order.
Types
"Packet" is the most common word used to refer to chunks of information sent over a network, but you might also see or hear them referred to as "frames," "blocks," "cells" or "segments." Even though the word may differ, the function is the same.
Packets are usually divided into three main parts:
Header--The information contained in the header carries instructions about the data within the packet, i.e. length, packet number, origination and destination.
Payload, or body--The payload, or body, of a packet contains the actual data that comprises the packet, i.e. part of a message or image. The bulk of the packet is contained within the payload or body.
Trailer, or footer--The trailer, or footer, tells the final destination device (i.e. your mom's computer) that the packet has ended.
Considerations
Although packet switching is a generally safe and reliable way of transmitting information, people should be aware of how files can become corrupted.
Packet switching divides a large piece of data such as an email into several smaller pieces of data. Because a network now has several different packets to transmit, there is a greater chance that one of these packets can become lost or misinterpreted during transmission, resulting in an incomplete piece of data.
There are different ways a packet can become lost or misinterpreted. Every once in a while a glitch in your network can completely miss a packet that is being sent, resulting in the loss of that packet. Also, a network may misread a packet and interpret it in a way other than the way it was originally intended. These problems are rare, but if you receive or transmit data over a network and an error occurs, then chances are the problem is with one or more packet(s). Because most files are complex, an improperly transmitted or missing packet will corrupt the entire file. If this happens, it is best to try to delete the file and re-download it.
Finally, one of the biggest problems with packets is that it is much easier for a virus to attach itself to an individual packet than it is to one large piece of data. This is because viruses are more obvious to your security software or firewall when they are part of a larger piece of data than a smaller one. Because of this problem, the makers of anti-virus software have added extra virus protection to standard security software. However, this extra protection can sometimes be a double-edged sword when it comes to packet switching, because sometimes security software will identify a virus in a packet when there is none at all, resulting in loss of some data.
Tags: packet switching, piece data, over network, security software, become lost, become lost misinterpreted, binary code