Authors Top

If you have a few years of experience in Computer Science or research, and you’re interested in sharing that experience with the community, have a look at our Contribution Guidelines.

1. Introduction

Internet Protocol (IP) is the communications protocol in the network layer. Its main operation is routing. It forwards packets originating at a host to its destination, one router at a time.

In summary, the transport layer breaks up data streams into IP packets. Then, IP routers forward these packets through the Internet. At the destination, the network layer reassembles all the packets into the original stream. Finally, the transport layer takes and processes the data.

The Internet is made up of different networks. The hardware for these networks supports varying bandwidths. As a result, sometimes, the routers have to break the IP packets into smaller parts along the way. This is called fragmentation.

In this tutorial, we’ll discuss the IPv4 datagram in detail.

2. IPv4 Datagram

IPv4 was developed in 1981, and it’s still widely used. It uses 32-bit address space.

An IPv4 packet is called a datagram. It is made up of a header and a data part:


IPv4 header contains a 20-byte fixed mandatory part, followed by optional fields. Hence, the minimum size of an IPv4 header is 20 bytes.

The optional part contains options and padding. It can grow up to 40 bytes in size. So, the maximum header size is 60 bytes.

A datagram can carry up to 65,535 bytes of data (2^{16}), at least 20 bytes of it being the header. Hence, the maximum amount of data it carries is 65,515 bytes.

The header part contains 14 fields. The first 13 are mandatory. The last one is the optional field called options. Let’s look at the individual header fields.

2.1. Version

IP version field, set to 4 for IPv4.

2.2. Internet Header Length

IHL is the number of 32-bit words making up the header field. The minimum available number is 5, as the first 20 bytes are mandatory.

2.3. Differentiated Services Code Point (DSCP)

It specifies the type of service for differentiated services, such as voice over IP.

2.4. Explicit Congestion Notification (ECN)

When supported, ECN carries a network congestion notification without dropping packets or wasting bandwidth.

2.5. Total Length

It’s the total size of the datagram, including both header and data parts.

2.6. Identification

It’s mainly used to indicate a group of fragments if the datagram is fragmented.

2.7. Flags

Different combinations of the flags control the fragmentation and indicate fragmented datagrams.

We assume the don’t fragment flag is set when the destination can’t assembly the fragmented packet. Therefore, the packet is dropped if the flags mark the datagram as not to be fragmented, but it must be.

2.8. Fragment Offset

It contains the offset of a fragmented packet.

2.9. Time to Live (TTL)

The number of hops a packet lives. Each router on the way decrements the TTL field. If it reaches 0, the packet is discarded. This way, looping packets are eliminated.

2.10. Protocol

It’s the protocol for the data part.

2.11. Header Checksum

It’s a checksum for error detection, covering only the header field. Each router checks this value and discards the packet if an error is detected.
The data part’s checksum is not calculated.

2.12. Source Address

Source host IPv4 address.

2.13. Destination Address

Destination host IPv4 address.

2.14. Options

Additional options are stored in this field when present.

3. Conclusion

In this article, we’ve learned about the most widely used network communications protocol. Then we investigated the structure of a datagram.

Authors Bottom

If you have a few years of experience in Computer Science or research, and you’re interested in sharing that experience with the community, have a look at our Contribution Guidelines.

Comments are closed on this article!