9.5 Protocols

The agreed upon rules of information transfer are called protocols. Network protocols are the key to reliable information flow within a network. Network protocols handle various services on the network, and each protocol has a specific function within the network. The services on a network are dependent on the type of task that has to be accomplished on the network. These tasks are made into modules such that one can work independently of the other, but each depending on the service of the other. One of the pioneering institutions in area was the University of Hawaii, where in the early seventies, researchers linked computers placed in various campus locations by means of radio signals. This very basic wireless network soon took on the name Aloha Net.

During this period of evolution in data networks, many of the well-established technology companies (vendors) such as IBM and DEC researched and developed their own version of networks. For example, IBM’s SNA network was very popular in the seventies and eighties, and is still present today. But with each vendor developing its own form of data communication networks, it soon became a problem when companies using different network solutions from different vendors could not communicate with each other. Therefore it was necessary for the computer industry as a whole to get together and to formulate a set of standards that all of the vendors followed, so that the networking equipment sold by one vendor will be compatible with the equipment sold by another.

In order for networked computers to successfully communicate with each other, there needs to be a defined set of ‘rules of engagement’ or protocol that each of those computers follow.  For example, when one drives on our roadways, it is important that one follows a set of road rules or protocol in order to get from point A to point B without getting into an accident. In this protocol for driving, each driver is expected to drive on the right side of the road, stop at a red light, stop at stop signs etc. If a driver from the United Kingdom, where the driving protocol is somewhat different (drivers drive on the left side of the road instead of the right side as in North America) was to drive in Canada using the British driving protocol then there will definitely be a collision! In a similar way, only computers using the same protocol can communicate with each other, unless one uses specialized devices used for protocol translation between two computers that are not using the same protocol. This is similar to a language translator. In the case of IBM’s vendor specific (proprietary) data network, the protocol is called SNA. In the case of DEC, the protocol is called DECnet and so on. 

While some protocols that are in use have been officially standardized and recognized by one of the standards bodies, other protocols are widely in use and are generally accepted by the industry, though the standards bodies do not officially recognize them. The former are referred to as ‘De Jure’ (in law) standards and the latter are referred to as ‘De Facto’ (in fact) standards. The following is a list of standards organizations which are responsible for the creation of the various standards as it pertains to data networks:

The OSI Model

In the late 1970s, the International Standards Organization (ISO) defined a communication architecture model called the Open Systems Interconnection Reference Model (the OSI Model) to facilitate network communication between networking equipment manufactured by different vendors. The OSI model is based on a protocol layering concept with 7-layers. This model was defined to facilitate problem free communication between networking devices manufactured by different vendors. The OSI model segments the communications process into seven independent functional divisions. This layered approach gives the flexibility for functionality in one layer to be changed without it affecting functionality in another layer. It is important to note that the OSI model in itself is not a protocol, but rather a set of guidelines that vendors can use to develop their networking products.

Layer 7 - Application layer. This layer provides the necessary interface between the end user and the underlying layers. Applications such as Netscape Navigator, and Microsoft Exchange are such interfaces.

Layer 6 - Presentation Layer. This layer provides such functionality as compression, formatting, and displaying the information.

Layer 5 - Session Layer. This layer provides such functionality as creating and terminating the various logical connections made between the software applications.

Layer 4 - Transport Layer. This layer ensures error free and orderly transmission of data between communicating devices by creating, maintaining, and terminating connections between the end devices.

Layer 3 - Network Layer. This layer provides addressing and routing the information from device A to device B. The network layer does not guarantee end-to-end information integrity during transmission, which is a function of layer 4.

Layer 2 - Data Link Layer. This layer is responsible for creating and monitoring boundaries during message transmission. These boundaries are referred to as frames. The layer also assumes responsibility for error detection, error correction, and message retransmission between communications devices.

Layer 1 - Physical Layer. This layer is responsible for defining the physical characteristics of the signal as well as the medium of transmission.  The physical layer also defines the conversion of data bits into electrical currents and light pulses by defining connector standards, voltage levels etc.

Currently there are a number of protocols in use globally. Due to the current trend of interoperability between equipment made by different vendors, much of the protocols in use today are standards based. Different protocols address different layers of the OSI model or stack; therefore, in order to have successful communications between a source node (computer-A) and a destination node (computer-B), it is essential that all communications devices in between computer-A and computer-B follow the rules set by the various protocols that operate at the various OSI layers.

Consider the case of a web search. Information typed in by an end-user, e.g. a student using Netscape Navigator to access his or her Yahoo email. This information filters down through the various layers of the OSI model and travels the network to the destination server, e.g. Yahoo mail server, and rises up the OSI model to the application layer of the mail server. It is beneficial to conceptualize this process in terms of each layer communicating with its peer as shown in figure 9.24. It is therefore necessary that respective OSI layers between communicating devices support identical protocols.        

Protocols such as Ethernet, Token Ring, X.25, Frame Relay, ATM, and SONET are sometimes referred to as lower level transport protocols as they define communications in the lower layers of the OSI stack (layers 1 & 2). These protocols typically do not concern themselves with end-to-end information integrity. Protocols such as TCP/IP and MPLS on the other hand define communications from an end-to-end perspective and function at layers 3 and 4. A further distinction is made between such protocols as Ethernet and Token Ring which are said to be Local Area Network (LAN) protocols and X.25, Frame Relay, ATM, and SONET, which are said to be Wide Area Networks (WAN) or Metropolitan Area Networks (MAN) protocols. LANs are typically networks that serve a company or a household. It is limited in its geographic coverage to a building or a campus. WANs are networks that serve a larger geographic area such as a city, and WANs are networks that serve an even larger geographic region such as a country, continent, or the globe. Today given the advances in technology, the boundaries between LANs, MANs, and WANs are blurring, with individual protocols being able to function in all three of the segments without the need to implement different protocols for the various segments.

A data network is made up of many types of devices, each with its own designated functionality. It is sometimes beneficial to think of a data network in terms of our postal service. If for instance Jack, who lives in Toronto, wants to send a CD to Jill, who lives in New York, then Jack would have to wrap the CD, write Jill’s address on the package and drop it off at the nearest post office. The post office would then sort through all of the mail and in the case of Jack’s package, would load the package into a plane that goes to New York. If there isn’t a direct flight to New York, then the package would be sent to a city near New York, say Boston, and then the post office in Boston would load the package on to a truck that is leaving for New York, based on the “To” address written on the package. Once the package finally reaches the post office in New York, it would then be delivered to Jill’s house. 

Computer networks work in a similar way. When computer-A needs to send some data to computer-B, the data is arranged into what is referred to as packets. Essentially a packet is comprised of the actual user data bits, also known as the payload, that are encapsulated with addressing information, as done in the post office example, along with several other bits to ensure data integrity, or in other words, detection and correction of bit errors along the way. In this scenario the “To” address on the packets would be that of computer-B. These packets are then sent by computer-A along the network being used to its nearest packet routing or packet switching device (device-1), like dropping off at the post office. Based on the “To” address on the packet, the router or the switch (device-1) would then direct the packet (route or switch the packet) to the router or switch to which computer-B is directly connected (device-3). If no such direct connection exist between device-1 and device-3 then device-1 would send the packets along the network to another device, device-2, that is nearest to device 3. Device-2 would then ultimately send the packets to device-3 based on the addressing information within the packet. Once device-3 receives the individual packets, it would then forward the packets on to computer-B, which is directly connected to it. The various pieces of information in the packets are organized as shown below.

9.5 Practice Questions                                                                                                                                           9.5 Practice Questions 1.     Why are protocols needed for network communication? 2.     What is the difference between a De Facto standard and a De Jure standard? 3.     Why are several standards organizations involved in setting standards? 4.     Name the 7 layers of the OSI model. 5.     At what layer of the OSI model does the user interact with the network? 6.     What are the advantages of using the OSI model? 7.     Explain what is meant by the term packet? 8.     What is an example of a protocol that operates at layer 2 of the OSI model? 9.     Explain the function of the session layer in the OSI model.