What is Internetworking?

In this tutorial, you will learn the basic concepts of internetworking. After reading this tutorial, you will learn about the connection of different networks, network issues, and fragmentation of packets.

Contents:

  1. Concept of Internetworking
  2. How Networks are Different?
  3. Issues of Networks
  4. Connection of Networks
  5. Tunneling Process
  6. Internetwork Routing
  7. Packet Fragmentation
  8. Transparent and Nontransparent Fragmentation

Concept of Internetworking

When two or more networks join together, they form an internetwork, which we call the Internet. But the problem also arises when more than two networks are connected. For example, the Ethernet networks and satellite networks are different. If they both connect, there is a transmission problem.

  • Larger networks are more valuable than smaller networks because many connections can be made to a larger network.
  • Internetworking is the concept where different types of networks are interconnected so that users on the network can communicate with other users on different networks.
  • If each network uses the same networking topology, it becomes easier for the Internet to function or connect to the network. But in the real network, the network is not only used for sending packets. Also, communication is complicated if the internetworking of different networks is not present.
  • We will look at internetworking-related issues in this tutorial.

How Networks are Different?

Different types of networks exist in the world. Networks differ in many ways such as frame formats, modulation techniques of the physical layer, etc.

  • When a sender sent the packet to the receiver, it travels through various networks known as foreign networks to reach the receiver. Also, there are many problems during transmission.
  • For example, the packet enters an Ethernet network, then a WiMAX network, then a connectionless network, and a connection-oriented network. This causes delays as it may be possible that the specific network cannot handle the load of many packets.

The diagram below illustrates how networks can differ.

advertisement
advertisement
illustrates how networks can differ

As shown in the figure, networks differ in many ways such as services, address, broadcast, packet size, quality of service, reliability, security, and parameters.

Issues of Networks

As we saw in the above section, there are many ways in which networks can differ. Some of the issues of networks are listed below.

  • Now, imagine a scenario where the network varies by considering the factor packet size. How does a router pass a 9000-byte packet through a network that has a maximum size of 1500 bytes?
  • It may be possible that packets first pass through a connection-oriented network, then through a connectionless network, resulting in packets arriving in a different order than when they were sent.
  • The above issues are prevented by joining two networks by using gateways that generate separate packets for each destination resulting in better network support for multicasting.
  • If the packet size is large, it can be broken into smaller pieces and sent in pieces. When the receiver receives the packet fragments, it reassembles the fragments and generates an original message. For that, they store the received packets in a buffer and deliver them in sequence.
  • The problem also arises when one network has a strong quality of services, and the other network has a best-effort system.
  • Therefore, whenever the network is prepared, these issues should be focused on.

Connection of Networks

After discussing the issues, let’s see how we can connect different types of networks. First, to connect different types of networks, we can build devices that convert packets from one type of network to another. The devices that convert the packets are placed at the boundaries of the network.

The diagram below shows how the different networks connect.

Packet Passes through Different Types of Networks
  • As shown in the figure, the network type from PC-1 to Wireless Router-1 (WR-1) is WLAN (802.11 Wireless LAN), from WR-1 to R-2 the network type is MPLS (Multiprotocol Label Switching), and the network type from R-2 to PC-2 is Ethernet.
  • Here 802.11 and MPLS networks provide connectionless service and connection-oriented service, respectively.
  • Now, PC-1 sends the packet to PC-2. When the packet reaches R-2, the network type is Ethernet. Ethernet can handle packets up to 1500 bytes in size, but packet sizes can be larger because 802.11 can work with larger frames than Ethernet. To prevent this problem, the packet is divided into small pieces.
  • The encapsulation and de-encapsulation process continue until the packet reaches its destination, as shown in the figure. The previous header is removed, and a new header is added to the packet. The packet contains its source and destination address, which tell the device where to send the packet.
  • The internetworking concept is not easy as we saw in this example until we add switches to the network.

Tunneling Process

The tunneling process is used when the source and destination hosts have the same type of network, but a different type of network exists between them. Tunneling is the process widely used to connect different networks or isolated hosts.

The figure below explains the tunneling method.

advertisement
Tunneling Method
  • As shown in the figure, PC-1 and PC-2 are on IPv6 network-1 and IPv6 network-2, respectively. There is an IPv4 network in between them.
  • Now, PC-1 wants to communicate with PC-2. So, it will generate an IPv6 packet and send it to the multiprotocol router-1, which connects the IPv6 network-1 to the IPv4 internet.
  • When Router-1 receives an IPv6 packet, it encapsulates the IPv6 packet into an IPv4 packet and sends it over the IPv4 network.
  • The packet sent by Router-1 passes through an IPv4 network path has a large tunnel extending from multiprotocol Router-1 to Router-2.
  • When Router-2 receives the packet, it finds the original IPv6 packet in the IPv4 packet and sends it to the destination PC-2.

Internetwork Routing

Different networks use different routing algorithms to process packets. For example, one network may use link-state routing, and another network may use a distance vector routing algorithm.

  • Different networks use different algorithms to find the path to send the packets. Some algorithms focus on finding the shortest path with the least delay, while others focus on finding the cheapest route.
  • The Internet is much larger than any individual network, so routing algorithms are used for better performance. Routing algorithms are used in intradomain (Interior Gateway Protocol) and interdomain (Exterior Gateway Protocol) networks.
  • In intradomain networks, router algorithms are used only within specific networks, whereas in interdomain networks, routing algorithms are used within specific networks and between two networks.
  • An intradomain protocol can be a link-state routing, and an interdomain routing protocol is BGP (Border Gateway Protocol).
  • All networks on the Internet are independent of each other. Each independent network is known as an autonomous system (AS).
  • Routing policy is also defined for the routing protocol. So that, if the routing protocol crosses international borders, different laws may apply to it.

Packet Fragmentation

Each network on the Internet can carry packets of some specific size. All networks impose some maximum size on their packets. Some of the reasons for these limitations are as follows:

  • Ethernet Frame Size – Hardware
  • Buffer size is 512 bytes – Operating system
  • Bits of the length field in the packet
  • Complies with national/international standards
  • Attempts to reduce error-induced retransmissions
  • Prevents packets from occupying the channel for a long time
  • So, if the network designer designs the packet, then he should check all the above factors about the packet, as he cannot choose any old maximum packet size as he wants.
  • The payload or data size in an Ethernet, 802.11, and IP is 1500 bytes, 2272 bytes, and 65,1515 bytes, respectively.
  • The source does not know how the packet will travel to the destination, and what type of network it will travel through. Therefore, it does not know how small packets should travel to the destination, the size of this packet is known as the path maximum transmission unit (path MTU).
  • Different networks handle packets of different sizes, so fragment is the method that allows routers to break packets into pieces.
  • When packets are fragmented into smaller packets, each piece of the packet is known as a separate network layer packet.
  • Packet switching networks have the problem of reassembling the packet fragments into the original packet.
  • Transparent fragments and non-transparent fragments are types of fragmentations.

advertisement

Transparent and Nontransparent Fragmentation

When the packet is broken into smaller fragments, two strategies are used to recombining the fragments to the original packet. They are transparent and non-transparent fragmentation.

The below diagram explains the transparent and nontransparent fragmentation.

Transparent and Nontransparent Fragmentation
  • As shown in the figure, in transparent fragmentation, the packet enters network-1 and arrives at router-1.
  • Router-1 breaks the packets into smaller pieces and sends them to Router-2 in the same network. As soon as router-2 receives the fragments, it combines the fragments back into the original packet and sends them to router-3 of network-2.
  • The same process will be repeated in Network-2 as well. The process continues until the packet reaches its destination.
  • In the non-transparent fragment, the packet enters network-1 and arrives at router-1. Router-1 breaks the packets into smaller pieces and sends them to Router-2.
  • Router-2 only sends fragments of network-2 to router-3. This non-transparent fragmentation process continues until the packet fragments reaches the receiver.
  • In nontransparent fragmentation, fragments do not recombine until they reach the receiver.
  • Routers don’t have to do much work in non-transparent fragmentation compared to transparent networks.
  • But in these methods, if even a fragment of the packet is lost, the whole packet will be lost. Path MTU discovery process is used to solve this problem. In MTU discovery, the sending host knows what length of the packet to send.

Key Points to Remember

Here is the list of key points we need to remember about “Internetworking”.

  • Internetworking is the concept where different types of networks are interconnected so that users on the network can communicate with other users on different networks.
  • When a sender sent the packet to the receiver, it travels through various networks known as foreign networks to reach the receiver.
  • To connect different types of networks, we can build devices that convert packets from one type of network to another. The devices that convert the packets are placed at the boundaries of the network.
  • The tunneling process is used when the source and destination hosts have the same type of network, but a different type of network exists between them. Tunneling is the process widely used to connect different networks or isolated hosts.
  • Different networks use different routing algorithms to process packets. For example, one network use link-state routing, and another network use a distance vector routing algorithm.
  • The payload or data size in an Ethernet, 802.11, and IP is 1500 bytes, 2272 bytes, and 65,1515 bytes, respectively.
  • When the packet is broken into smaller fragments, two strategies are used to recombining the fragments to the original packet. They are transparent and non-transparent fragmentation.

If you find any mistake above, kindly email to [email protected]

advertisement
advertisement
Subscribe to our Newsletters (Subject-wise). Participate in the Sanfoundry Certification contest to get free Certificate of Merit. Join our social networks below and stay updated with latest contests, videos, internships and jobs!

Youtube | Telegram | LinkedIn | Instagram | Facebook | Twitter | Pinterest
Manish Bhojasia - Founder & CTO at Sanfoundry
Manish Bhojasia, a technology veteran with 20+ years @ Cisco & Wipro, is Founder and CTO at Sanfoundry. He lives in Bangalore, and focuses on development of Linux Kernel, SAN Technologies, Advanced C, Data Structures & Alogrithms. Stay connected with him at LinkedIn.

Subscribe to his free Masterclasses at Youtube & discussions at Telegram SanfoundryClasses.