103 – Networks Assessment

This response will be awarded full points automatically, but it will be reviewed and possibly adjusted after submission.

There are two methods of sending across a network, circuit switching and packet switching.

With circuit switching a physical path using  or fibre optic cables is setup between the sender and receiver prior to  of the message.  The path remains open throughout the communication and is only released for use by others once the message is .  Although circuit switching is still used, it isn’t the best choice for large networks because it is difficult to  up and every connection has to be capable of supporting the maximum  that might be required.

For large networks, and the internet, packet switching is used instead.  Here the message is broken down into smaller parts (data ) which are sent individually through the network by the shortest\1}  route available.  The packets are given a ‘packet ‘ to identify the order they are in.  When the individual packets arrive at their destination they will be  back into the correct order.  Once all of the packets have safely arrived a message will be sent to the original computer as confirmation. If a packet is missing or  then the message will ask for that particular packet to be resent.

The data packets contain three important pieces of information:

1. a or return address to say where the packet originated from
2. a address to tell the packet where it has to go
3. a number so that the individual data packets can be reassembled in the correct order once they have all safely arrived at their destination

Benefits of packet switching include:

1. Robustness – there are many paths through which the packets can travel.
2. The connection is required for only a brief period of time, unlike circuit switching where it is held until the whole message is transmitted.
3. If a packet is lost or corrupted only that packet needs to be resent – unlike circuit switching where the whole message has to be resent.
4. Each packet carries its own correction codes, so if a few bits are corrupted, they can often be corrected.

However, one issue is that when one packet arrives more slowly than the others, it is missing or corrupted then delays can occur (known as ).

The speed, or bandwidth of a network, is measured in per second.  The more bits per second, the more that can be transmitted in a given period of time.   Most people have to share the bandwidth on a network and as there is a limited bandwidth available, the more people that are using the network, the slower it can become.  The term, ‘ ratio’ is the ratio or number of users compared to the bandwidth available.

Networks often use physical connections to join them together.  The most common choice is using cables or Ethernet cables which carry electrical signals.  These can be either Cat-5 or Cat-6.  Cat-5 cable is the least expensive and is fine for a network which is around 100Mbps.  However, if the network needs to be around 1 Gbps or above then Cat-6 is the better option.

Even faster than Ethernet cables are optic cables.  These work by transferring data using which is reflected along the inside of the cable.  They offer very high bandwidth but are expensive when compare to cables.