Menu
Home
Log in / Register
 
Home arrow Computer Science arrow Understanding Network Hacks
< Prev   CONTENTS   Next >

2.10 UDP

UDP (Unified Datagram Protocol) is, like TCP, a protocol of the transport layer, but in contrast to TCP it lacks session support and is therefore classified as stateless. Further on it doesn't care about packet loss or order and only implements addressing of programs through ports. A typical UDP header can be seen in Fig. 2.14.

UDP works by the principle of “fire and forget” and is mostly used for streaming services like internet radio or television, but its also the most common used transport protocol for DNS. The advantage of UDP is the size its header adds to the packet and therefore the much higher speed.

2.11 An Example Network

An Ethernet/TCP/IP network is what you nowadays think of if you hear the term network, because it is by far the most common one. Its constructed of five layers instead of the theoretical seven layers of the ISO/OSI model. For short refreshing: Ethernet is on Layer 2, IP (Internet Protocol) on Layer 3, TCP (Transport Control Protocol) or UDP (see Sect. 2.10) on Layer 4–6 and services like HTTP, SMTP, FTP on Layer 7.

Lets see how a HTTP packet passes all those layers one after another. In our example we want to get the index page of springer.com. First our computer parses the URL springer.com into the following components: HTTP as application protocol to be used, the hostname www, the domain springer, the Top-Level-Domain – TLD for short – (com) and at last the resource we try to receive in this case /.

Armed with these information our computer constructs the following HTTPHeader (Layer 7):

GET / HTTP 1.1

Host: springer.com

Next we head on to TCP (layers 4–6). It establishes a connection by the use of the Three-Way-Handshake addressing the destination port 80 (HTTP) and a random source port to connect the browser with the network.

IP (Layer 3) recognizes that it cannot use springer.com for addressing since it can only use IP addresses such as 62.50.45.35 so it makes a DNS query to resolve the IP for the hostname. We will learn more about DNS in Chap. 6. Now IP checks if the destination host is in the same network as our computer. This is not the case therefore a lookup into the routing table is necessary to retrieve the address of the next hop. There is no entry for the destination network thus the default gateway is used to send the packet to the outside world. Last but not least IP writes the address of the network card used to send the packet into the source address and our packet travels to the next layer.

On layer 2 the packet gets received by the ethernet protocol. ARP takes care about resolving the MAC address of the destination IP address and remembers them in the ARP cache this ensures it doesn't have to ask the network for every packet. Ethernet writes the MAC of the outgoing network card as source into the header and forwards the packet to the last layer (physical) in this case the driver of the network card, which will translate the packet to zeros and ones and transmit it on the medium.

 
Found a mistake? Please highlight the word and press Shift + Enter  
< Prev   CONTENTS   Next >
 
Subjects
Accounting
Business & Finance
Communication
Computer Science
Economics
Education
Engineering
Environment
Geography
Health
History
Language & Literature
Law
Management
Marketing
Philosophy
Political science
Psychology
Religion
Sociology
Travel