Technology was developed that allowed computers in different organisations to be networked using what would now be described as a wide area network (WAN).

The computer could be a stand-alone system or some organisations chose to have more than one computer connected using a local area network (LAN).

The client-server model (or architecture) was first used in large organisations when they had installed internal networks.

The client-server mode of operation nowadays is different. The client is a web browser connected to the Internet. The server is a web server hosted on the Internet.

The server provides an application and the client uses the application.

There are five requirements for a data communications system:: a sender, a receiver, a transmission medium, a message and a protocol.

Data can be sent through the medium in different modes:

• simplex mode where data flow is one-way only
• half duplex where data can flow either way but not simultaneously
• full duplex where simultaneous both-ways data flow is possible.

A ‘message’ is any type of data, which can be sent as either:

• a broadcast, which is a one-to-all communication (as used traditionally for radio and television)
• a multicast, which is from one source to many destinations
• a unicast, which is a one-to-one communication.

• The use of certain wireless transmission frequencies is regulated by government agencies and so permission has to be obtained before wireless transmission is used. (无线传输频率受政府监管)
• Outside these frequencies, no permission is needed to use the air for transmission but cables can only be laid in the ground with the permission of landowners. (电缆只能在获得土地所有者许可的情况下才能铺设)
• For global communications, the two competing technologies are:: transmission through fibre-optic cables laid underground (or on the sea bed) and satellite transmission (discussed later in this section).
• Interference is much more significant for wireless transmission and its extent is dependent on which frequencies are being used for different applications. (无线传输存在干扰)
• Repeaters are needed less often for wireless transmission. 无线传输不需要中继器
• Mobile (cell) phones now dominate Internet use and for these, only wireless transmission is possible. (手机只能使用无线传输)
• For home or small office use, wired or wireless transmission is equally efficient; often, not having to install cables favours wireless connections for a small network. (家用效率相同，但是无线传输不需要安装电缆)

Nowadays, twisted pair cables are probably the most widely used networking connections, and fibre-optic cables are becoming more common.

A bus can be extended by linking two bus cables using a repeater. A repeater is needed because over long distances, signals become attenuated (reduced in strength), making communication unreliable.

Two segments are connected using a bridge. The bridge stores the network addresses for the end-systems in the two segments it connects.

The LAN port on an end-system is connected to a Network Interface Card (NIC). The NIC is manufactured with a unique network address that is used to identify the end-system in which it has been installed.

For a star network, the central device might be a hub, a switch or a router. The switch is by far the most likely. A switch is a connecting device that can direct a communication to a specific end-system.

WiFi (WLAN in some countries) is a term used to describe wireless Ethernet.

The central device in a WiFi LAN is a Wireless Access Point (WAP).

The WAP can communicate with an end-system in the WiFi LAN provided that the end-system has a Wireless Network Interface Card (WNIC) installed.

以太网(Ethernet)：指的是由Xerox公司创建并由Xerox,Intel和DEC公司联合开发的基带局域网规范。以太网络使用CSMA/CD(载波监听多路访问及冲突检测技术)技术，并以10M/S的速率运行在多种类型的电缆上。

它不是一种具体的网络，是一种技术规范。

以太网是当今现有局域网采用的最通用的通信协议标准。该标准定义了在局域网（LAN）中采用的电缆类型和信号处理方法。以太网在互联设备之间以10~100Mbps的速率传送信息包，双绞线电缆10 Base T以太网由于其低成本、高可靠性以及10Mbps的速率而成为应用最为广泛的以太网技术。直扩的无线以太网可达11Mbps，许多制造供应商提供的产品都能采用通用的软件协议进行通信，开放性最好。

而因特网是一个网络的网络(a network of network)。它以TCP/IP网络协议将各种不同类型、不同规模、位于不同地理位置的物理网络联接成一个整体。它也是一个国际性的通信网络集合体，融合了现代通信技术和现代计算机技术，集各个部门、领域的各种信息资源为一体,从而构成网上用户共享的信息资源网。

以太网以交换机为基础交换，因特网以路由器为基础交换。

CSMA/CD即载波侦听多路访问/冲突检测，是广播型信道中采用一种随机访问技术的竞争型访问方法，具有多目标地址的特点。它处于一种总线型局域网结构，其物理拓扑结构正逐步向星型发展。CSMA/CD采用分布式控制方法，所有结点之间不存在控制与被控制的关系。

CSMA/CD的全称是Carrier Sense Multiple Access with Collision Detection，即基于冲突检测的载波监听多路访问技术。CSMA/CD也是最初802.3中的核心，应用在10M/100M的半双工有线网络中，目前CSMA/CD的应用场景少了很多，大部分都直接基于全双工工作。

节点发送数据之前需要持续监听信道，一旦节点发现信道空闲，则立刻发送数据。在发送数据的同时，节点持续监听信道，"探测" 是否有别的节点也在该时刻发送数据。

若传输过程中没有检测到别的节点的传输，那么成功传输。在成功传输后，节点需要等待帧间间隔IFG（interframe gap）时间后，可以进行下一次传输。

若在传输过程中，探测到别的节点也在传输，那么则检测到冲突。发生冲突后，节点立刻停止当前的传输，并且发送特定的干扰序列（JAM序列），用以加强该次冲突（用以保证其余所有节点都检测到该次冲突），在JAM序列发送完之后，节点随机选择一个时间倒数进行backoff。当backoff完成之后，节点可以尝试再次重传”。

1. 开始 - 如果线路空闲，则启动传输，否则跳转到第4步。
2. 发送 - 如果检测到冲突，继续发送数据直到达到最小回报时间（min echo receive interval）以确保所有其他转发器和终端检测到冲突，而后跳转到第4步。
3. 成功传输 - 向更高层的网络协议报告发送成功，退出传输模式。
4. 线路繁忙 - 持续等待直到线路空闲。
5. 线路空闲 - 在尚未达到最大尝试次数之前，每隔一段随机时间转到第1步重新尝试。
6. 超过最大尝试传输次数 - 向更高层的网络协议报告发送失败，退出传输模式。

The standard has so far evolved through five generations:: standard or traditional, fast, gigabit, 10 gigabit and 100 gigabit.

In either topology, a transmission was broadcast type. Any message would be made available to all of the end-systems without any controlled communication exchange between any pair of end-systems. For each message received an end-system had to check the destination address defined in the message to see if it was the intended recipient.

collision

• If two end-systems were to transmit messages at the same time there would be what is described as a ‘collision’.
• The method adopted for dealing with this was CSMA/CD (carrier sense multiple access with collision detection).
• This relied on the fact that if a message was being transmitted there was a voltage level on the Ethernet cable which could be detected by an end-system.

step

• Check the voltage on the transmission medium.
• If this indicates activity, wait a random time before checking again.
• If no activity is detected, start transmission.
• Continuously check for a collision.
• If no collision is detected, continue transmission.
• If a collision is detected, stop transmission of the message and transmit a jamming signal to warn all end-stations; after a random time, try again.

The initial function of an Internet Service Provider (ISP) was to give Internet access to an individual or company.

We can also think of the Internet in terms of the connections that carry the most traffic, which consist of a set of fibre-optic cables laid under the sea and across land, which can be described as a ‘mesh’ structure.

Each router is connected to several other routers and its function is to choose the best route for a transmission.

The mobile phone, equipped with the appropriate software, communicates with a standard cell tower to access the wireless telephone network, which in turn provides a connection to the Internet.

The Internet is, as has been described above, an Internetwork.

By contrast, the World Wide Web (WWW) is a distributed application which is available on the Internet.

the web consists of an enormous collection of websites each having one or more web pages. The special feature of a web page is that it can contain hyperlinks which, when clicked, give direct and essentially immediate access to other web pages.

Cloud computing is the provision of computing services usually via the Internet. An organisation may choose to establish its own private cloud.

The alternative is a public cloud. This is created, managed and owned by a third-party cloud service provider.

Service::

• infrastructure provision - IaaS - Infrastructure as a service
• platform provision - PaaS - Platform as a service
• software provision - SaaS - Software as a Service

Streaming media make use of the Internet for leisure activities like listening to music or watching a video. Because of the file sizes involved, streamed media is always compressed to a sequence of bits - a ‘bit stream’.

on demand - In this case the delivery of the media and the playing of the media are two separate processes. The incoming media data are received into a buffer created on the user’s computer. The user’s machine has media player software that takes the media data from the buffer and plays it.

real-time（live transmission）- the content is being generated as it is being delivered such as when viewing a sporting event.

bit rate - The process of delivering the content is determined by the bit rate. example 300kbps or 128kbps

For the connection to the Internet the gateway would either first connect to a router or have the capability to act as a router itself.

On the Internet, all of the allocated IP addresses have a netID pointing to the router. The router then has to interpret the hostID to direct the transmission to the appropriate workstations on one of the LANS via a gateway.

The ISP will have available a large number of hostIDs. However, the number of users that the ISP is supporting could very likely be larger than the total number of addresses available. Fortunately for the ISP and for an individual user many of these potential users will not be engaged in Internet interaction. The normal practice is for the ISP to create a ‘dynamic address’ for a user.

The alternative is a ‘static address’ which never changes and can be provided if a user is prepared to pay an extra charge.

The domain name service (DNS) service allocates readable domain names for Internet hosts and provides a system for finding the IP address for an individual domain name.

There are more than 250 top-level domains which are either generic (e.g. .com, .edu, and .gov) or represent countries (e.g. .uk and .nl).

The domain name is included in a universal resource locator (URL), which identifies a web page, or an email address.

name resolution:

• If the domain is under the control of the server to which the query is sent then an authoritative and correct IP address is returned.
• If the domain is not under the control of the server, an IP address can still be returned if it is stored in a cache of recently requested addresses but it might be out of date.
• If the domain in the query is remote then the query is sent to a root server which can provide an address for the name server of the appropriate top-level domain. This in turn can provide the address for the name server in the next lower domain. This continues until the query reaches a name server that can provide an authoritative IP address.