The IPv4 Header

The IP header has the following sizes and fields:

• Version—The first 4 bits of the IP header identify the operating version of IP; for example, version 4 or version 6.

• Internet Header Length—The next 4 bits of the header contain the length of the header, expressed in multiples of 32.

• Type of Service—The next octet contains a series of flags that can be used to specify precedence (that is, absolute priority relative to other IP packets), delay, throughput, and reliability parameters for that packet of data. The precedence flag is 3 bits long whereas the delay, throughput, and reliability flags are each 1 bit in length. The remaining 2 bits are reserved for future use.

• Total Length—This 16-bit field contains the total length of the IP packet measured in octets. Valid values can range up to 65,535 octets.

• Identifier—Each IP packet is given a unique 16-bit identifier, which is used to identify the fragments of a datagram.

• Fragmentation Flags—The next field contains three 1-bit flags that indicate whether fragmentation of the packet is permitted, and if it is used. The first bit is reserved and always set equal to 0. The second bit indicates whether that packet’s data can be fragmented. If this bit is equal to 0, the contents can be fragmented. If it is equal to 1, it cannot be fragmented. The third bit has significance only if the second bit was set to 0. If that bit was equal to 0 (and the data can be segmented across multiple packets), this bit indicates whether this particular packet is the last in the series of the fragment, or whether the receiving application can expect additional fragments. A 0 indicates that this packet is the last one.

• Fragment Offset—This 8-bit field measures the offset of the fragmented contents relative to the beginning of the entire packet. This value is measured in 64-bit increments.

• Time-to-Live (TTL)—The IP packet cannot be permitted to roam the WAN in perpetuity. It must be limited to a finite TTL. The 8-bit TTL field is incremented by one for each hop the packet makes. After reaching its maximum limit, the packet is assumed to be undeliverable. An ICMP message is generated and sent back to the source machine and the undeliverable packet is destroyed.

• Protocol—This 8-bit field identifies the protocol that follows the IP header, such as VINES, TCP, UDP, and so forth.

• Checksum—The Checksum field is a 16-bit error-checking field. The destination computer, and every gateway node in the network, will recompute the mathematical calculation on the packet’s header as the source computer did. If the data survived the trip intact, the results of these two calculations are identical. This field also informs the destination host of the amount of incoming data.

• Source IP Address—The source address is the IP address of the source computer.

• Destination IP Address—The destination address is the IP address of the destination computer.

• Padding—Extra zeros are added to this field to ensure that the IP header is always a multiple of 32 bits.

These header fields reveal that IPv4’s Internet Layer is inherently connectionless: The packet forwarding devices in the network are free to determine the ideal path for each packet to take through the network. It also doesn’t provide any of the acknowledgments, flow control, or sequencing functions of higher-level protocols such as TCP. Nor can IP be used to direct the data contained in an IP datagram to the appropriate destination
application. These functions are left to higher-level protocols, such as TCP and UDP.