Previous Table of Contents Next


Different Types of Internet Connections

Three kinds of connections are available on the Internet.

Packet-Switched. Related to the idea of broadcast is that of packet-switched. A message is broken into packets, each packet is labeled as to its origin and destination and then is broadcast onto the network. Other nodes forward the packet in the general direction of the destination. It is possible that adjacent packets in a message will follow different paths to the destination. This is the opposite of circuit-switched networks, such as the voice network, in which a circuit or path is determined in advance and all parts of the message follow the same path. In a packet-switched network, an intervening node may see only a part of a message. On the other hand, it increases the number of nodes that may see a part of it.

Peer-Connected Nodes on the Internet are “peer connected.” No node dominates or controls another. Thus, by default, all nodes behave as if they trust all other nodes as themselves. The implication is that the level of trust is equal to that of the least trusted node.

Any-to-Any Connection Like the postal system, and except as otherwise restricted, any device connected to the Internet can send a message to any other device. There is no requirement for an answer but, at a minimum, the destination device must recognize the message and make a decision about it. For example, at MIT the softdrink vending machines are connected to the Internet. If the addresses are known, they may be queried from anywhere in the world.

Increasing Interoperability

If connectivity is the ability to send a message to any node, interoperability is the ability to get a meaningful answer back. Already, the Internet is better at answering questions than most individuals are at asking questions. The Internet can provide a report of freeway traffic in Los Angeles, hotel availability in London, or the schedule of every opera house in the world for the next two years. It can also locate all the bed and breakfast lodgings in most places in the world, and get an index to the treasures of the Vatican Library or of the British Museum. Individuals can locate and download graphics, moving images, and general and specialized software. A query on “Mona Lisa” returns references to both 1000 different prints of Da Vinci’s La Gioconda and a sound clip of the Nat King Cole song. If the necessary software is unavailable to interoperate with another system at a particular layer, software can be downloaded at another.

As protocols and interfaces become more standard, they become more useful. As the use of a standard increases, so does the propensity to comply with it. The less standard an interface, the more it must include information about its intended or productive use.

No Central Authority

Although there are authorities such as the IAB and the IETF, which make architectural and design decisions for the Internet, no one is obliged to follow them. The individual networks are independently owned and operated. There is no central authority that is responsible for the operation of the entire network. Because the network is global, it is not even subject to the authority of any single nation state.

INTERNET PROTOCOLS

The Internet can also be defined and described in terms of the communication protocols that it employs. One, somewhat pure, definition is that the Internet is that collection of interconnected networks that employ TCP/IP suite of protocols. A more practical definition is that the Internet is that set plus those networks connected to it by appropriate gateways. (For purposes of this definition, a gateway is a node that translates traffic from one protocol to another.)

The Internet Protocol

The fundamental protocol of the Internet is IP, the Internet protocol. IP is the network layer protocol for the TCP/IP Protocol Suite. It is fundamental in the sense that all other protocols are built on it. It is connectionless, best-effort, packet-switched, and unchecked. “Best effort” means that the network will do its best to deliver the packet, but there are no guarantees. “Unchecked” means that there is no redundancy in the protocol to enable either the sender or the receiver to know whether the packet was received correctly. There is no acknowledgment of the receipt of the message. The receiver cannot be sure that the message comes from where the origin address of the packet says that it comes from.

IP is to the Internet as the post card is to the postal system, limited in capacity, function, and intent. However, just as a message of any length can be sent by using multiple post cards, or by using one post card to acknowledge or to check on another, IP packets can be composed in such a way as to compensate for all of these limitations. These compositions make up the higher-level protocols.

The Transmission Control Protocol

The TCP, is the standard IP for the transfer layer. It defines how IP packets are sent back and forth between a sender and a receiver to provide many of the things that IP does not. However, even TCP does not provide security or the reliability of origin and destination. Both the sender and the receiver know that they are talking to someone that is orderly and well behaved, but they do not know for sure that it is their intended party, and they do not know if any one is listening in.

The Oldest and Most Widely Used Protocols

The following are among the oldest and most widely used protocols on the Internet:

  Telnet. This was originally intended for connecting host-dependent terminals to remote systems or applications. Today, it is used by terminal emulator programs on workstations.
  File Transfer Protocol. FTP is used to move files from one system to another.
  Simple Mail Transfer Protocol. SMTP is used for E-mail.

The applications of these protocols are discussed in subsequent sections.


Previous Table of Contents Next

Copyright © CRC Press LLC