TCP/IP protocol suite consists of Internet protocol which is used for data delivery over network. Remaining protocols TCP, UDP and ICMP in the TCP/IP protocol suite uses the Internet protocol to deliver the data from one host to another host. This associates with the network layer of OSI reference model.
The network layer is heart of any network based on the TCP/IP protocol suite. The network layer consists of the Internet protocol (IP), Internet Control Message Protocol (ICMP) and Internet Group Management Protocol (IGMP). IP performs most of the work within the network layer. ICMP and IGMP will assist IP in management of special network messages such as multicast and error messages. Network layer is normally referred as IP layer because it’s close relationship with IP.
IP Address (Internet Address)
The data moves across network with the help of IP in TCP/IP protocol suite. IP uses address which is known as IP address to transfer data between hosts. An IP address is 32 bits or 4 bytes wide.
For Example: 18.104.22.168 is an IP address which consists of 4 bytes in dotted-decimal notation. The IP address consists of network ID and host ID. Network ID is unique in a network to distinguish from other network. There are 5 types of networks categorized based on high-order byte in an IP address. They are Class A, B, C, D and E networks. This is developed to overcome address space limitation.
I am taking high-order byte in an IP address to explain about 5 different classes.
1111 1111 – Represent all network or a broadcast address.
0111 1111 – Represents Class A Network (High-order byte will be up to 127 in decimal). In this class the 1 st byte represents network ID and remaining 3 bytes represents host ID.
1011 1111 – Represents Class B Network (High-order byte will be up to 191 in decimal). In this class first 2 bytes represents Network ID and remaining 2 bytes represents host ID.
1101 1111 – Represents Class C Network (High-order byte will be up to 223 in decimal). In this class first 3 bytes represents Network ID and remaining 1 byte represents host ID.
1110 1111 – Represents Class D Network (Used for multicasting and High-order byte will be up to 239).
1111 0111 – Represents Class E Network (Reserved for future use and High-order byte will be up to 247).
Link layer address protocols:
There are two types protocols associated with the Link layer; they are Address Resolution Protocol (ARP) and Reverse Address Resolution Protocol (RARP). These protocols maps an IP address into a link layer address and vice versa.
ARP maps addresses in the network layer to the corresponding address in the link layer. Link layer address is network technology specific. There are two types of network technologies. They are Ethernet and ARCNET. Ethernet addresses are 6 bytes wide and ARCNET addresses are one byte wide. ARP uses link layer’s broadcasting capability to identify the hosts that leave or join network. RARP is used to maps the Ethernet address into IP address. This is designed to use with diskless computers.
Datagram is a self-contained unit of data which is delivered by a data delivery device. TCP/IP uses IP for data delivery. Application layer transfers the data to transport layer as application message. Transport layer converts this data into either TCP segment or UDP datagram. As the data moves to network layer it becomes IP packet. Then the IP packet converted into Ethernet frame.
IP packet consists of IP header and actual data. The network software creates an IP header in multiples of 32 bit words. It contains only 20 bytes of storage space. An IP header contains following fields.
|Field Name||No. of Bits||Purpose of the field|
|Version Number (VERS)||4||Version of IP used to create datagram.|
|Header Length (HLEN)||4||This specifies the IP header length. By default, HLEN is 20 Bytes wide.|
|Type of Service (TOS)||8||This specifies the Priorities for the IP Packet. First 3 bits denotes precedence, 4th bit denotes delay, 5th bit denotes throughput, 6th bit denotes reliability, 7th bit denotes cost and 8th bit is unused.|
|Packet Length||16||This field indicated the total size of the packet including IP header. Limitation of IP packet size is called MTU|
|Identification||16||Identification of the fragmented small pieces. Fragmentation occurs when the IP packet is larger than MTU.|
|Flags||3||All fragments this flags field’s last bit is set to 1 except last fragment.|
|Fragment offset||13||Used to hold the breaking point (the distance from the beginning of the datagram)|
|Time-to-Live||8||This specifies how long the packet can line out on the network.|
|Protocol||8||This field indicates the protocol created the data encapsulated within the packet.|
|Header Checksum||16||Checksum of the IP header field used to detect data transmission errors.|
|Source IP address||32||The 32 bit IP address of the source host computer|
|Destination IP address||32||The 32 bit IP address of the destination host computer|
Note: 160 bits means 20 Bytes, 32 bit words are 5. This is the maximum length of an IP header.