Chapter 2: Protocol Architectures and Standards
To achieve the complex functions for networks mentioned in section 1.4, we need an elaborate system comprised of hardware and software. The structure of this elaborate system is called a Protocol Architecture and its implementation is called a Protocol Suite. An example of protocol architecture is DPA (DOD Protocol Architecture) and its implementation is called as TCP/IP protocol suite (also known as Dept of Defense Protocol Suite). Other example architectures are IBM’s SNA (Systems Network Architecture), OSI Model, XNA (Xerox Network Architecture) and Novell Netware.
2.1 Protocol Architectures
**What is a protocol?
Since the overall functions for computer communications are very complex, usually the system is divided into several layers. This layering makes the whole system more manageable by having modularity (the ability to work on one layer at a time), modifiability (the ability to modify a layer without affecting other layers). The number of layers ranges between 4 and 7 depending upon the architecture. For example, TCP/IP architecture divides the functions into 4 layers and IBM SNA and OSI model divide the functions into 7 layers.
Each layer of a given protocol suite achieves a set of predefined functions. Every computer should employ the same protocol suite to communicate each other. Therefore, when two computers communicate each other, each layer on a machine communicates with its corresponding peer layer on the other machine. This mode of operation is called “peer-to-peer layer communication”.
This peer-to-peer layer communication must follow “a set of strict rules that govern the communication”. This set of rules for a given layer is called “Nth layer protocol”. For example, IP (Internet Protocol) is a 3rd layer protocol (if it is positioned in the OSI 7 layer architecture).
**protocol architecture diagram**
** Protocol Architecture Diagram
Peer-to-peer layer communication: Communication between a given layer on a machine with the same layer on the other machine. Example: An ftp protocol (file transfer protocol: an application layer protocol in TCP/IP Protocol suite) on a machine talking to an ftp protocol on another machine.
Nth layer protocol: A set of strict rules that govern the communications between peer layers of communicating machines.
Example: 3rd layer protocol on TCP/IP protocol suite is IP.
Protocol Architecture: A layered structure of a set of predefined protocols.
Example: DPA (DOD(Dept Of Defense) Protocol Architecture) that is essentially the same as Internet protocol suite, SNA(Systems Network Architecture of IBM), NETWARE(Novell’s protocol architecture)
Protocol suite (also called as a protocol stack): Implementation of layered protocols based on a protocol architecture. Example: Internet protocol suite (also called as TCP/IP protocol suite) is based on the DPA.
**TCP/IP architecture diagram
Equipped with the above concepts, we can summarize the computer communication as follows:
Each layer’s functions are defined by it’s protocol called nth layer protocol(1st layer, 2nd layer, etc.). The functions that a layer provides are called the layer’s services since a layer provides a set of functions(services) to it’s upper layer. In turn, the upper layer uses the lower layer’s services to achieve its own functions(services). A higher layer is called a service user since it uses the services provided by its lower layer. A lower layer is called a service provider.
**diagram showing layers and service users and service providers
**Virtual communication
Virtual communication: When the peer-to-peer layer communication is carried out, the peer layers do not have direct connections except the physical layers. For example, when an IP layer sends an IP packet to another machine, it has to use the services from the lower layers to deliver the IP packet. Since there is no direct connection between the peer IP layers, the peer layers achieve communication indirectly, in other words, virtually.
* Analogy for the layered architecture of protocols
A popular analogy is a communication between presidents of two countries. Imagine two presidents in their offices want to talk each other. They cannot pick up a telephone and start talking since they do not know each other’s languages. They need to follow a proper way(this is also called as protocol), so the communication may happen in the following manner.
1. President X writes a message(to president Y) and hands it to the foreign minister
2. The foreign minister interprets the message into another language and encrypt the message and hands it to a person responsible for the communication between countries
3. This form of handing the message down the layers may happen several times……..
4. At last, the message is physically delivered. The message may go through several vehicles(trains, automobiles, etc.).
5. At the receiving end, things happen from the bottom and go up all the way until the message reaches the president Y.
***protocol diagram which shows the Encapsulation/Decapsulation
*Encapsulation and Decapsulation
A layer composes a PDU (Packet Data Unit) which contains two parts. It is a packet destined to the peer layer on the other machine.
|
Nth layer header |
Nth layer data |
The header part contains information such as Address and Control information. The data part for the Nth layer follows the header. The whole thing is called as an N-PDU(Nth layer Protocol Date Unit).
When this data (N-PDU) is passed down to (N-1)th layer, the (N-1)th layer adds its own header to the N-PDU. Now, the packet looks like:
|
(N-1)th layer header |
Nth layer header |
Nth layer data |
This process of each layer adding its own header is called “Data Encapsulation”. Now, when the message arrives at the peer layer, let’s say (N-1)th layer, the layer consume the header((N-1)th header) for the layer’s control information and passes the remaining data to its upper layer(Nth layer). Nth layer will do the same process of stripping off the header and passing the remainder to its higher layer. This process is called as “Decapsulation”. When a layer handles a packet, the portion that is passed to its higher layer is untouched(not processed), therefore, “uninterpreted”.
When computers began communicating each other starting around mid 1960s and having evolved to a more elaborate networking, many protocol architectures were developed and implemented. For example, IBM had its own protocol architecture called SNA (Systems Network Architecture) which has 7 layers. DEC (Digital Equipment Corp: minicomputer manufacturer which was known to follow the IBM mainframe’s footsteps in 1960s, 1970s and 1980s) had its own protocol architecture called DNA (Digital Network Architecture). XEROX developed its own architecture called XNA (Xerox Network Architecture). As we can see here, there were many protocol architectures and their implementations in 1960s and 1970s. The main problem was that they were incompatible each other. In other words, these different protocol suites that are based on different architectures can not communicate each other without some kind of interpreters in between. They had many differences; number of layers different, layer boundaries, and layer’s functions.
These incompatibilities between architectures urged the international community to develop “standard protocols” so that all computers can talk each other. Starting around mid 1970s, many standards have been developed. It is essential to understand the existing standards, the organizations responsible for the standards.
2.2 Standards, Standard organizations
In a simplified view, a computer network is an implementation of the layered protocols of a given architecture. To achieve an orderly, error-free communication, each layer must be defined precisely and all implementations must follow only one definition. So, we need a standard for each layer that is adopted nationally and internationally.
The Computer networking world experienced its share of frustration in the 1970s when there existed many different, non-standard protocols (so called de facto standards) such as IBM’s SNA, DEC’s DNA, XEROX’s XNA, etc. There was a definite need for standardization.
Many standard organizations exist in national level, regional level, and international level and numerous standards have been adopted by them. Since it is not practical to know all of them, let’s concentrate on the major players and the most important standards from them.
According to their scope of coverage, we can classify them into three levels; national, regional, and international levels.
(a) National level standard organizations
· ANSI (American Standard Organization): It is a
non-profit organization responsible for adopting standards in the
· EIA/TIA (Electronic Industries Association/Telecommunication Industries Association): It is mainly comprised of members from electronic manufacturers, actual producers of hardware. Many companies involved in hardware manufacturing are members. EIA/TIA adopts its own standards that are followed in the manufacturing and the standards are forwarded to ANSI for adoption. The most well-known are EIA/TIA 568(Commercial Building Telecommunications Cabling Standard), EIA RS232 (Serial Interface standard), and many others.
· NIST (National Institute of Standards: NIST, formerly called as NBS(National Bureau of Standards), is under the Department of Commerce and controls all the governmental standards. NIST publishes FIPS (Federal Information Processing Standards) which mandates all the governmental bodies to follow the standards for purchasing. One exception is Department Of Defense. DOD has its own standards. NIST is one of the major contributors to ANSI standards.
· For
countries other than the
(b) Regional level standard organizations
There are many regional standard bodies around the world. Here are the major players.
- ECMA International (European Computer Manufacturers Association): ECMA International is an industry association founded in 1961, dedicated to the standardization of information and communication systems. ECMA International is a non-voting member of ISO and ITU-T but recommends standards for adoption.
- ETSI (European Telecommunications Standards Institute): ETSI is a non-profit organization whose mission is to determine and produce the telecommunications standards at regional level.
- Pacific Area Standards
Congress (PASC): Established in 1972, PASC is voluntary organization,
independent of national standards organizations. It is a forum to strengthen
international standardization programs of the ISO and IEC and to improve
the ability of the
- Other regional organizations: There are numerous regional standard organizations. A list is provided to illustrate the fact that there are many organizations:
Accreditation Cooperation (APLAC)
Pacific Accreditation Cooperation (PAC)
Asian Consultative Committee on Standards and Quality (ACCSQ)
European Committee for Standardization (CEN)
Committee European De Normalization Electronic (CENELEC)
Joint
Accreditation System -
Arab Industrial Development and Mining Organization (AIDMO)
Euro
(c) International standard organizations
· ISO (International Standards Organization): It is a
voluntary, non-treaty organization. Membership is by each nation’s national
standards organizations, such as
§ ISO 7498—OSI Reference Model
§ Quote from ISO web site (http://www.iso.ch/iso/en/ISOOnline.frontpage) : “The ISO 9000 and ISO 14000 families are among ISO's most widely known standards ever. ISO 9000 and ISO 14000 standards are implemented by some 610 000 organizations in 160 countries. ISO 9000 has become an international reference for quality management requirements in business-to-business dealings, and ISO 14000 is well on the way to achieving as much, if not more, in enabling organizations to meet their environmental challenges.”
· IEC (International Electrotechnical Commision): international co-operation on electrotechnical standardisation and related matters, such as assessment of conformity of standards. There are more than 60 participating countries. ISO and IEC work closely at certain areas.
· ITU-T (International Telecommunications Union-Telecommunications Standardization Sector): Another acronym that is often used is ITU-TSS. ITU-T is a newer name, formerly it was known as CCITT (Consultative Committee on International Telephone and Telegraph).
ITU-T
is an intergovernmental, treaty organization under United Nations. Within the
ITU-T, the public and private sectors cooperate for the development of
telecommunications standards. The ITU adopts international regulations and
treaties governing all terrestrial and space uses of the frequency spectrum as
well as the use of the geostationary-satellite orbit. It also develops
standards to accommodate the interconnection of telecommunication systems on a
worldwide scale. Membership is by governmental bodies of each country;
§ CCITT X-series standards—X.21, X.25, X.200
§ CCITT V-series standards—V.35, V.90
§ CCITT I-series standards—I.100, I-200
· Overlap between ISO and ITU-T: As far as standards for communications are concerned, there is lot of overlap between them. Even though each has its own set of standards, they usually cooperate closely when adopting standards. Each concentrates on certain areas; ITU-T concentrates on telecommunications and the lower layers of OSI model, ISO concentrates on computer communications (computer networks) and the higher layers of OSI model. When one of the organizations adopts a standard, it is common that the other adopts a similar standard. An example is the OSI model; the OSI model is an ISO standard which is numbered as ISO7498 and ITU-T standard numbered as ITU-T X.200.
· IEEE (
**Internet Standards
With the experimental nature and the openness of ARPANET, TCP/IP network protocols are perpetually renewed as the research advances. There is no designated governing body to issue directives and regulations for the Internet unlike a typical standards enforcement by a national standard body such as ANSI. Internet standards are mostly developed via cooperation by many participants/organizations.
Internet Society (ISOC) is the standardizing body for the Internet community. The Internet Society is a non-governmental, international organization for global cooperation and coordination for the Internet operations and the internetworking technologies, applications, and protocols. The Society is governed by its Board of Trustees elected by its memberships around the world. Members are from industries, governmental organizations, educational institutions, manufacturers, other organizations.
Quoting from www.isoc.org, “The Internet SOCiety (ISOC) is a professional membership society with more than 150 organization and 16,000 individual members in over 180 countries. It provides leadership in addressing issues that confront the future of the Internet, and is the organization home for the groups responsible for Internet infrastructure standards, including the Internet Engineering Task Force (IETF) and the Internet Architecture Board (IAB).
Since 1992, the Internet Society has served as the international organization for global coordination and cooperation on the Internet, promoting and maintaining a broad spectrum of activities focused on the Internet's development, availability, and associated technologies.”
Internet standards are governed by IAB (Internet Architecture Board) and under IAB there is the actual standard body called IETF(Internet Engineering Task Force) for adopting and issuing new standards. Under IETF, there are many Working Groups where new standards are discussed, modified, and adopted. The IETF itself is governed by the Internet Engineering Steering Group (IESG). For details, pay a visit to ISOC’s web site (www.isoc.org) and also IETF’s web site (http://www.ietf.org).
Internet standards are issued in the documents called RFCs (Request For Comments). Search the Internet with the key word, “RFC”. The process of standardization is summarized below:
· When a new specification needs to be approved as a standard, an application has to be submitted to the IESG where it will be discussed and reviewed for technical merit and feasibility. It is also published as an Internet draft document (as an RFC).
· After the final approval by the IESG, the draft is recommended to the Internet Engineering Taskforce (IETF) for inclusion into the standards and for publication as a Request For Comment.
· After it is published as an RFC, it may be revised over time or phased out when better solutions are found.
Following figure is from http://www.livinginternet.com/i/iw_mgmt.htm a good sight!
2.3 Network Infrastructure
Computers need to be connected to achieve communication. This connection structure is called “network infrastructure”. Usages of the phrase “network infrastructure” depend upon the size of the networks. A small company may have a small infrastructure and a large company may have a large infrastructure. An ISP (Internet Service Provider) may have a nationwide infrastructure. Depending upon its size, each network infrastructure may be classified into a LAN, a MAN or a WAN.
LAN (Local Area Network): The acronym LAN is used to designate a small scale network. It can range from a very small network with few computers to a large scale network with many computers. A large scale network may be made up of many smaller LANs and still called as a LAN, in this case it may be called as a campus LAN. The scope of the term LAN depends on its each usage.
MAN (Metropolitan Area Network): A larger scale network which covers a city. Examples of MANs include cable television network (where cable modems are used for computer communication) and DSL (Digital Subscriber Loop) network covering a city.
WAN (Wide Area Network): Any network larger than a MAN can be classified as a WAN. A WAN is a network infrastructure which interconnects LANs and MANs together.
Here is a conceptual diagram which shows the interconnections of LANs, MANs, and WANs.
From http://www.cybergeography.org/atlas/topology.html
From http://www.cybergeography.org/atlas/topology.html
A topology map of a core network of
a medium-sized ISP. It was created
using an automatic network discovery tool called Mercator developed as part of
the Scan project by Ramesh
Govindan, Anoop Reddy and colleagues, at the Information Sciences Institute,