Development of computer networks, do you really know

Since the network into a pit workers, we first need to have a general understanding of the development of the network. This is like in the future someone asks you about your history of struggle, even if you own struggle said they do not understand that this may a little awkward. Well, without further ado, we come to the question.

Foreword

While the computer network has experienced only a few decades of development, but now it has become an indispensable part of people's lives. Some important features of the 21st century is the digital, network and information technology, it is a network-centric information age. To realize the information must rely on a sound network, because the network can transfer information quickly, the network has become an important basis for the development of the information society and the lifeblood of the knowledge economy.

A, Internet development process

The infrastructure of the Internet has gone through three stages of evolution. These three stages in time division is not completely separate, but there is some overlap, because the network evolution is gradual, not one day suddenly undergone a qualitative leap in particular.

  1. ARPANE process from a single network to Internet development.

1969 US Department of Defense created the first packet switching network ARPANET began as a single packet Internet (even when a network interconnection are not really). To connect all hosts on the ARPANET are directly connected to the nearest node switch. But by the mid-1970s, people have realized that is impossible to use only a single network to meet all the communication problems. ARPA then began to study a variety of network (e.g., packet radio networks) interconnected technology, which led to the emergence of the interconnection network. At that time the US Department of Defense just want to speed up the transmission of information through the network, in order to deal with the threat may face, therefore, the birth of the Internet was just an accident. They never thought that one day, the network will become an integral part of people's lives.
Here Insert Picture Description
At this point interconnection network has become the prototype of the Internet today. 1983 TCP / IP protocol as a standard protocol on the ARPANET, so that all computers using TCP / IP protocol can use the Internet to communicate. So people put time in 1983 as the birth of the Internet. ARPANET in 1990 announced the closure, because its experimental tasks have been completed.

  1. The completion of the tertiary structure of the Internet

Since 1985, its US National Science Foundation NSF (National Science Foundation) six large computer centers around the construction of computer networks, namely the National Science Foundation Network NSDNET. It is a three-level computer network, into the backbone, regional networks and campus network (or intranet). The three network covering the nation's major universities and research institutes, and a major part of the Internet. In 1991, NSF and other government agencies began to realize that the Internet will expand its use should not be limited to universities and research institutes. At this point, many companies in the world have access to the Internet, so that the sharp increase in traffic on the network, then the Internet has the capacity has been unable to meet the demand. So the United States government decided to turn to the Internet backbone private company, and started the unit to access the Internet charges. 1992 hosts on the Internet more than 1 million units. In 1993 the rate of Internet backbone up to 45Mbit / s (T3 rate).

  1. Multi-level structure of the Internet evolving ISP

Beginning in 1993, the US government-funded NSFNET gradually replaced by a number of commercial Internet backbone, government agencies no longer in charge of Internet operations, which appeared in a new term: the Internet service provider ISP (Internet Service Provider). In most cases, ISP is a commercial activity of the company. Such as China Telecom, China Mobile, China Unicom and other companies is the famous ISP in China.

ISP can apply on the Internet governing body to many IP addresses (hosts on the Internet must have an IP address to the Internet), also has communication lines (large ISP companies build their own communication lines, a small ISP companies lease large ISP's communication lines ) and routers and other networking equipment, so long as any institutions and individuals pay the prescribed fee to the ISP, you can use the right to obtain the required IP address and access to the Internet from the ISP. We often say that the so-called "Internet", meaning access to the Internet. IP address management agency will not assign a single IP address to a single user, but to a certain number of paid IP address leased to qualified through examination ISP (IP address management agency only engage in the wholesale, retail engage ha ha). Thus, now the Internet for countless large and small co-owned ISP in the world, this is the Internet also is called a "network of networks".

-----------------------------------------Dividing line------- -----------------------------------------

Second, the Internet standardization

The rapid development of the Internet began in the 1990s. By the European Organization for Nuclear Research CERN developed the World Wide Web WWW (world wide web) is widely used on the Internet, which greatly facilitates the use of the majority of non-network professionals network, the Internet has become the main driving force exponentially. The number of Web sites is also rapid growth.

The figure below shows 1993--2016 Nian Internet subscriber growth
Here Insert Picture Description

As the saying goes: "No rules, no standards." With the rapid development of the Internet, the world wide use of the Internet as a means of communication. But this time the Internet and there is no uniform standard, giving users a great inconvenience. Development of international standards is a very complex issue, there is both a lot of technical problems, there are many non-technical issues, such as competition for economic interests between different vendors (as well known as Cisco and Huawei equipment problems).

Because since 1992, the Internet is no longer owned by the US government jurisdiction, so the establishment of an international organization: the Internet Society (Internet Society, abbreviated ISOC), for the overall management of the Internet and promote their development and use in the world. ISOC technical organization under the jurisdiction of the Internet Architecture Board IAB (Internet Architecture Board), responsible for managing the development of Internet-related protocols.

IAB is also under the jurisdiction of two institutions: the Internet Engineering IETF (Internet Engineering Task Force), IETF is a forum made up of many engineering group WG (working group) consisting of these working groups is divided into several areas, each focused on a particular field specific short-term and medium-term engineering problems, mainly for the development and standardization of protocols.

Another institution under the jurisdiction of the Internet Research IRTF (Internet Research Task Force), a forum of several groups RG (research group) composition. Task is to study some of the problems require long-term consideration, including some of the Internet protocols, applications, architecture and so on.

All of these institutions to develop Web standards are in the form of RFC (request for comments) published on the Internet, all the RFC documents can be downloaded from the Internet for free. But not all network standards are RFC, will last only a few RFC documents become Internet standards.

Development of Internet standards to go through a formal look at three stages:
⑴ draft Internet - Internet draft is valid for six months, at this stage can not be the RFC documents.
⑵ proposed standard --RFC documents
⑶ Internet standards - after reaching the official standard, each standard will be assigned to a number STD xx. A standard can be associated with multiple RFC documents.

-----------------------------------------Dividing line------- -----------------------------------------

Third, the composition of the Internet

Although the Internet topology is very complex, but from the point of view of its working methods, it can be divided into the following two categories:

  1. Edge portions

Consists of all the hosts connected to the Internet. This part is used by the user directly, to communicate and share resources. The host is also known as end systems (end system), that is, we often say that terminal. Edge portion using a service provided by the core part, so that the host can communicate with each other and a number of exchange or share information.

We must first clear a concept: Host A and Host B communicate, actually refers to "run a process running on the host A and the other processes on the host B communicate." This more stringent argument usually referred to as "communication between computers."

Communication between end systems can be divided generally into two categories: client - server model (C / S) mode and a peer (P2P mode).
⑴ client - server model

Internet is the most common and traditional way. We mail. When find information, use is the C / S mode.
Customer (client) server (server) refers to the two application processes involved in communication, C / S model describes the relationship between processes and services with the service. Customer service requester, the server is a service provider. They have to use the services provided by the core of the network.
Here Insert Picture Description

⑵ peer connection mode

Peer connection means does not distinguish between two hosts in the communication which is a service requester, which is the service provider. As long as both hosts running peer to peer connectivity software (P2P software), they can be equal, peer to peer communication connection. At this time, both parties can download the shared documents stored on the hard disk of each other.
Here Insert Picture Description

  1. core part

The core network portion is the most complex part of the Internet, because a large portion of the edge of the core network hosts To provide network connectivity, so that the edge portion Renyiyitai host can communicate to other hosts.
In the central part of its special role a router, packet switching is a key member (packet switching), and its task is to forward the received packet. Exchange is divided into circuit-switched and packet-switched.

Circuit switched ⑴

Shortly after the advent of the telephone, people will realize that, to make all the phone twenty-two connected to communicate, it is very unrealistic. Its large workload, cost of manpower and material resources as much as people are unacceptable. So people want to use an intermediate device to connect all calls on this device intermediate so that all the telephone may communicate with each other. This is the telephone exchange, the exchange using a circuit switched.

When the increase in the number of the telephone, it is necessary to use a lot of switches connected to each other to complete the exchange of the task of the whole network. In this way, it constitutes a telecommunications network covering the whole world.

From the point of view of communication resources, the transmission line switching is to allocate resources dynamically in some way. Before using a circuit switched, dial-up request must first establish a connection. When the called user to hear the tone sent by the switch, and an off-hook, the calling terminal and the called terminal on the establishment of a connection is a dedicated physical channel, after the calling and called parties can talk to each other. This connection ensures that when the two sides talk communication resources needed, these resources are not being used when communicating parties other users. After the communication is completed hook, the switch is released before the physical channel established.
This must establish a connection (request and acquisition of a communication resource) → calls (persistent communication resource occupancy) → releasable connection (return communication resource) is called a three-step exchange circuit switching .

Important feature of the circuit-switched: the entire communication time, communication resource is always occupied by two users to communicate end to end. When using a circuit switched transmission of computer data, the line utilization is low.

⑵ packet switching

Compared with the instant switching circuit switched, packet switching is used store and forward technology. The following figure shows a packet re-transmission is divided into several packets before. Packet data units are transmitted in the Internet, the packet header contains important control information such as destination address and source address, each packet transmission paths can be selected independently in the Internet, and ultimately delivered to the correct transmission destination.

As the basic unit of packet transmission in the network
Here Insert Picture Description
during packet transmission, located at the edge portion of the network host computer to process the user information, and may exchange information through the network and other hosts . Router to forward the packet, packet switching .
After the router receives a packet, temporarily store it, check the header of the packet, the routing table according to the destination address, find a suitable forwarded out the interface. Thus by the N router forwarding packets will eventually be delivered to the destination host.

The figure below shows the comparison of each exchange
Here Insert Picture Description

-----------------------------------------Dividing line------- -----------------------------------------

Fourth, the development of China's Internet
development of the Internet in China began in the late 1980s. September 20, 1987, Professor Qian Tianbai by Italian public packet-switched network ITAPAC of PAD in Beijing issued our first e-mail to communicate with the University of Karlsruhe in Germany, opened a prelude to the Chinese people use the Internet .
September 1989, the State Planning Commission established the Zhongguancun area of education and research and demonstration network (NCFC). The main objective of the project is between Peking University, Tsinghua University, Chinese Academy of Sciences and 3 units in high-speed Internet, and to establish a supercomputing center. The project construction is completed in 1992.

October 1990, China formally registered in the DDN-NIC our country's top-level domain CN. In April 1993, Chinese Academy of Sciences Computer Network Information Center, part of the network of experts convened by the Domain Name System survey of States, pursuant to raise our country's domain name system.

1994 January 4, NCFC project connected to the Internet by the US company Sprint 64k international special line connection to the Internet full-featured, from our country has officially become the Internet.

Since 1994, chaired by the State Planning Commission, the Ministry of Posts and Telecommunications, the state religion of the Board of Education and Chinese Academy of Sciences, built China's four Internet, namely China Golden Bridge Information Network, China Public Computer Internet, China Education and Research Network and the China Science and Technology Network. In a few short years, these backbones will be put into use, form the basis of the national backbone.

Since 1996, development of the Internet into the application platform and value-added business development. China's Internet has entered a rapid development period of unprecedented growth. A large number of Chinese sites, including integrated "portal" sites and a variety of professional Web sites have appeared, providing news, technology consulting, software downloads and ICP entertainment and other services, as well as web hosting, domain name registration, free space and other technical support service. At the same time, a variety of value-added services has gradually expanded, mainly e-commerce, IP telephony, video on demand and wireless Internet access. Premise applications in the face of widening Internet penetration and rapidly growing number of Chinese Internet companies began to enter the overseas stock market Nasdaq, become an important symbol of China's economic development a new turn of the century.

1997 June 3, according to the decision Leading Group Office of the State Council Information Technology, Chinese Academy of Sciences Network Information Center set up the China Internet Network Information Center (CNNIC), at the same time, the State Council Informatization Leading Group Office announced the establishment of China Internet Network Information Center Commission.

In November 1997, CNNIC issued a "China Internet Development Statistics Report." As of October 31, 1997, China had 299,000 Internet computer, Internet users 62 million domain names registered under CN 4066, WWW sites 1500, international bandwidth is 18.64Mbps.

January 22, 2017, China Internet Network Information Center (CNNIC) released the 39th "China Internet Development Statistics Report." As of December 2016, Chinese netizens reached 731 million, Internet penetration rate reached 53.2%. The scale of mobile phone users reached 695 million, an increase of more than 10% for three consecutive years. Desktop computers, notebook utilization rates decline, the use of mobile phones continue to squeeze other personal Internet devices.

-----------------------------------------Dividing line------- -----------------------------------------
Fifth, Computer Network Architecture

Computer network development to today, has evolved into a large and complex system. In computer professionals, a complex system hierarchical organization system, that is a lot of relevant functional decomposition, to be individually interpreted and implementation.

  1. Form a computer network architecture

We can imagine a first simplest case: two computers in the network to transfer files to each other. Obviously, to accomplish this goal, there must be a path for transmitting data between computers, in addition, the following work needs to be done:
passage ⑴ activation data communication to ensure that the data can be correctly transmitted in this pathway and receiving.
⑵ detect whether the other computer connected to the network properly.
⑶ ensure that the client computer is ready to receive and store work files.
⑷ If the two computers are not compatible file format, wherein the at least one computer to complete format conversion function.
⑸ variety of errors and accidents that may occur, such as data transmission errors, duplicated or missing packets, a network node failure occurs, there should be measures to ensure a reliable peer eventually receives the correct file.

Thus, the two computer systems communicate with each other must be highly coordinated work of the job, and this coordination is quite complex. For the design of this complex network, as early as in the original design of ARPANET has proposed a layered approach. "Layered" can be large and complex problems into several smaller localized problems, while smaller local research issues and facilitates handling.

In 1974, the United States, IBM announced Systems Network Architecture SNA system (system network architecture). This famous network standard is developed by a layered approach. Soon after, other companies have also introduced architecture with a different name of their company.

After different network architectures appear, and by using different network architectures devices communicate with each other is difficult. However, the global economy so that users of different network architectures urgent need to exchange information with each other. To make the computer networks of different architectures can interconnect, network standardization organizations ISO organization established to study the issue in 1977. They proposed the OSI Basic Reference Model of the OSI / RM (Open Systems Interconnection Reference Model), referred to as OSI.
"Open" means non-exclusive monopoly, so just follow the OSI standard, a system can communicate with any system follow the same standards.
"System" refers to the various parts of the system in real interconnection related.
Formed in 1983, the official document of the OSI basic reference model, known as the ISO 7489 international standard, which is today the OSI reference model.

OSI trying to reach an ideal state, namely global computer networks follow this standard, making the global computer can easily interconnect and exchange data. In the 1990s, though the whole of the international OSI standards have been developed, but few manufacturers able to produce commercial products meet OSI standards, and was eager to TCP / IP Internet has been executed successfully in the global range. Therefore, OSI received only the results of theoretical research, but in terms of market failure. So now we often say, of the OSI international standards on law, TCP / IP is the de facto international standard . Now, the technical standards with a strong business atmosphere, the emergence of a new standard, sometimes not necessarily reflect their level of technology is the most advanced, but there is a certain market background.

  1. Agreement with the division level

In the network to do an orderly exchange data, they must abide by the rules agreed in advance. These rules specify the format and related synchronization problems of the exchanged data . The rules for the data exchange in the network and established standards or conventions called network protocol. Network protocols include three elements:
⑴ Syntax: the structure or format of data and control information.
⑵ semantics: what needs to send control information, which operation is completed and how to make the response.
⑶ synchronization: Event achieve a detailed description of the order.

Protocol usually two different forms, one is used to facilitate the reading and understanding of human textual description. Another is to use a computer program code can be understood. These two different forms of agreement must be able to make an accurate interpretation of the information exchange process on the network.

ARPANET research and experience has shown that, for very complex computer network protocol, it should be hierarchical in structure. Each level of open systems by a number of entities. Entities are software elements (such as process) or hardware elements (such as intelligent I / O chip) abstraction. At the same level, called peers, a level by multiple entities. For convenience of description, may be referred to as any layer (N) layer, its upper and lower face layer is referred to (N + 1) layer and the (N-1) layer.

The basic idea is that each hierarchical level are using the basic services provided by the lower layer, the upper layer to provide services, and higher-level service provided to run distributed applications. Another purpose is hierarchical independence between the guarantee level . I.e. (N) layer does not need to know how to lower it is achieved, only need to provide services to the upper layer and the lower interface by using the service provided by the lower layer. Since only achieve a relatively independent function in each layer, which can be a complex problem intractable problems into several smaller more easily handled. In this way, the complexity of the whole problem declined.

Here Insert Picture Description
In general, delamination has the following advantages:
independent layers between ⑴
⑵ good flexibility and
can be separated on the structure ⑶
⑷ capable of promoting standardization

Of course, stratification also has some disadvantages: as some features will be repeated at different levels, resulting in additional overhead.

It is a collection of computer network architecture and network protocol layers.

Has been mentioned previously, OSI seven layer protocol architecture, a complete theory, but complex and not practical. TCP / IP architecture is different, now has a very wide range of applications. TCP / IP is a practical application in the four-layer architecture includes an application layer, a transport layer, an Internet layer (interconnection of different networks to solve the problem) and the network interface layer. But in fact, the network interface layer and no specific content, so TCP / IP only the top three layers. In general, we integrated OSI and TCP / IP advantage, only to adopt a five-protocol architecture, so both simple concept can be explained clearly.
Here Insert Picture Description

So far, the architecture of the computer network will be formally established and perfected. As for the role of each layer, which has a protocol, not be discussed here, the latter will specifically write an article to explain.

-----------------------------------------Dividing line------- -----------------------------------------
END

Recently registered a number of public, intends to knowledge of computer networks and computer-based finishing scratch again to others. No public has just started, there is hope to get to know like-minded friends, we encouraged each other, and common progress. Wen is the end of my public number of two-dimensional code, welcome Tell me what concerns
Here Insert Picture Description

Guess you like

Origin blog.51cto.com/14671287/2481413