IP Addressing
Back

Characteristics-
IP has many important characteristics, which make it the predominant network layer protocol in use today. Internet Protocol is a connectionless protocol, responsible for transmitting data. IP is classified as connectionless because it does not have to first establish a connection with the receiving system to transmit, and its working partner TCP provides the connection at the transport layer. IP also provides data encapsulation, which is basically packaging the data into “letters” for transport. It also provides the very important function of addressing, where each NIC is identified by a unique IP address (unique on the network for private IP’s, and unique worldwide for public IP’s). These addresses are extremely useful, as they do not use the hardware address printed on the card (MAC address) so they can be used in any type of network or computer platform (IP still uses ARP, address resolution protocol, to supply the data link layer protocol in use with the hardware address). IP also provide address routing, which means it will choose the most efficient way to the destination address, or the path with the least amount of “hops”. IP also fragments packets into smaller portions and repackages them, to fit into the different media types and data link protocols which the information may need to traverse or be delivered to. IP also functions as a protocol identifier, which will tell the destination system which protocol generated the Data.

Assigning IP’s-
IP addresses are assigned by two different sources, depending on their type. If the IP address is a local IP on a private network, the network administrator will assign the addresses as he sees fit, either manually (small networks) or through Dynamic Host Configuration Protocol, which is recommended for larger networks. These IP can be given freely, as they will be used only for identification on the network to other machines on the network. However, if the IP is going to be a network identifier and be able to talk to other machines on the web or outside of its own network, it will need to be a public IP. These IP’s are generally used for routers acting as gateways, and need to be unique worldwide, so there are no conflicts. Thus, a computer on a network will be identified by computers not on the same network through its gateway, which is a unique address, in addition to its host address, which is unique on its own network but not globally. To prevent duplications and IP conflicts globally, the IANA is the root source of all global public IP addresses. However, these IP addresses are generally obtained from an ISP (internet service provider), who will in fact obtain them from the IANA first.

IP Classes-
There are three main different classes of IP address, which are used to provide support for different sizes of networks. Class A addresses are used in smaller networks with many clients, and can be identified by having a first byte value of 0-127. Class B addresses support a greater number of networks, and an equal number of host identifiers. They can be identified by the first byte value of 128-191. Class C addresses support the most amount of network identifiers, and the least host identifiers. They are identified by a first byte value of 192-223. Despite the fact that some of the Classes are more limited in network identifiers, these can be expanded even more by applying a subnet mask. This is a division of the network address, used to represent different part of that address and expand it. A subnet mask is used to specify which bits are the network identifier and which bits is the host identifier. The mask of a Class A address is 255.0.0.0, Class B 255.255.0.0, and Class C are 255.255.255.0.

Routers
Routers are the “gateways” through which packets must travel to reach their destination addresses. Each “hop” that a packet makes on its way to its destination is a stop on a router that the packet has been configured to take. The routers are called intermediate systems, while the source of the packet and the destination of the packet are both call end systems. When an IP packet passes through a router, the packet will only travel up as high as the network layer, so that the packet can be send properly to the next system. Thus, the router will never access the information in the datagrams data field. The router will strip off the data link layer frame and build a new one, making changes to the header and time to live field, however the actual information is never touched (each time the packet hits a router, the time to live field is decreased by one. When this value reaches 0, the packet is discarded and must be resent using Internet Control Message Protocol or ICMP). Note- a command called traceroute (tracert in dos) will map the number of hops at the routers a package takes, by generating packets with successively larger TTL values. This will generate an error through ICMP, which is assembled by traceroute and displays the entire route the packet takes to its destination.