Computer Networking a Top-Down Approach Featuring the Internet

Chapter1: Computer Networks and the Internet Introduction Chapter goal get context, overview, and feel of computer networking more depth, detail later in the course approach: descriptive use Internet as example This chapter describes the overall computer network and Internet, and its purpose is through the computer network / Internet composition, critical understanding of computer network technology, the overall situation (see the forest through the trees.) Same time, involving more involved in the knowledge point, this chapter also guide the introduction of the specific details will be given in later chapters. Roadmap What is the Internet? The Network edge The Network core Access networks and physical media ISPs and Internet backbones Delay & loss in packet-switched networks Protocol layers and their service models History of networking and the Internet

First, from two perspectives to understand what Internet: Internet and the specific composition of the services provided, as well as the concept of network protocols. Part of the learning at the network edge-to-end systems, network applications, architecture, connectionless and connection-oriented services and so on. Section involving the computer network switching technology, which circuit switching and packet switching, and computer network of the type / classification. Access network to learn the classification of the fourth quarter, the main form of access networks and physical media. Part of the ISP and Internet backbone connection between the compositions of the Internet to learn the structure of ISP. Section VI study the delay in packet switched networks and packet loss, delay involved in the type of packet loss caused by the main reason. VII, the main contents on computer networks / Internet architecture, namely, computer network protocols according to levels of components, layers and protocols is a collection of network architecture. The last part is about the development of computer networks and Internet history. Their own learning. 1.1 Whats the Internet? Nuts-and-bolts description The basic hardware and software components that make up the Internet. Computing devices, communication links, switches A service description Describing the Internet in terms of a networking infrastructure that provides services to distributed applications. Connection-oriented reliable service Connectionless unreliable service Internet difficult word to define. We introduce the following two perspectives from what is the Internet. 1, nuts and bolts: "Nuts and bolts" of the internet is the basic hardware and software components that make up the Internet. 2, service-oriented view: describing the Internet in terms of a networking infrastructure that provides services to distributed applications.

1 A Nuts-and-Bolts view End systems public Internet is a worldwide computer network that connects millions of computing devices Pcs workstations, servers, PDAs phones, even toasters. hosts The term computer network is being to sound a bit dated! Public Internet is a global computer network, connected to the distribution in many parts of the world's computing devices (Computing devices), such as traditional PC, Unix-based workstations and servers, PDA, mobile phones, environmental sensing equipment, and even home appliances (household appliances) and other equipment. Thus, in a sense, Computer Networks concept some old, out of date. All of these Computing devices are called host or end system. Currently, the number of Internet hosts has nearly 500 million. Host Count in the Internet

communication links connect the end systems together to form a network Fiber, copper, radio, satellite transmission rate: bits per second (bps)

End systems are usually connected to each other through packet switches. forward arrival packets from input to output A packet is a chunk of data. Router and Link-layer switch forward packets thru network according to the value of header field 2

Packet switches packets arriving from the input link from the output link forwards them. Packet switch there are two categories, one is called a router, a class called the link layer switches. The role of the two similar groups are forwarded, the difference is forwarding packets based on different information. Router's IP address according to packet forwarding packets, link layer packet switch according to the destination MAC address to forward packets. Example-A Packet

route / path links and switches sequence that a packet traversed from a source to the destination end system In the Internet, the route is dynamic, and not dedicated. packet switching Allows multiple communicating end systems to share a link, a switch, or the whole path. Routing and path: packet from sender to receiver through the transmission process of the sequence of communication links and switches. Packet Switched Network (Packet Switching Network), this path is not fixed, but dynamic, that is, between the source and destination host may send multiple packets.

End system access the Internet through the ISP. Internet Service Provider Each ISP is a network of packet switches and communication links. Protocols control the sending and receiving of information within the Internet. Protocols are rules must be obeyed. TCP/IP TCP: transmission control protocol IP: Internet protocol Internet standards RFC: Request for comments IETF: Internet Engineering Task Force IEEE: Institute of Electrical and Electronics Engineers Protocol: the two sides abide by the rules of communication, mainly used to specify the packet format, and each packet is received after the implementation of the action collectively referred to as Internet Protocol TCP / IP protocols, including TCP, IP, DNS, HTTP, PPP, etc. Protocol for the Internet is extremely important, the organizations responsible for developing protocol standards are: IETF Internet Engineering Task Force RFC: Request for Comments IEEE (Institute of Electrical and Electronics Engineers): IEEE802.3 3

 public Internet vs. private intranet Internet is a network of networks public Internet :the network that one typically refers to as the Internet, that is, the global network of networks extranet private Internet: heir hosts cannot exchange messages with hosts outside of the private network intranet Brief summary End systems/hosts Communication links Packet switches Route or Path ISP Protocols Public Internet and private Intranet

2 A Service Description The Internet enables distributed applications WWW, email, e-commerce, database, voting, Everything over IP! connection-oriented reliable service guarantees that data transmitted from a sender to a receiver will eventually be delivered to the receiver in order and in its entirety Connectionless unreliable service without any guarantees about the eventually delivery Internet allows distributed applications running on their end systems to communicate with each other, with the "Everything over IP" the trend, more and more types of network applications. For example, Web surfing, instant messaging (QQ, MSN, etc.), Video-on-demand, streaming media and so on. Internet services provided in two ways: Reliable connection-oriented service: to ensure that from the sender sends the data to the receiver will eventually order to reach the receiver intact Connectionless unreliable service: does not provide any guarantee for the transmission The current Internet provides only a service model, "best effort": the network to do their best to transfer the application packet, but not a packet transmission delay without any warranty. Internet is also not willing to pay more for the users with better service. The Internet without QoS (Quality of Service) capabilities. 3 Whats a protocol? human protocols: Whats the time? I have a question Introductions there are specific msgs we send, and specific actions we take in response to the received reply msgs or other events What's a protocol? Protocol is a buzzword in computer networking what does a protocol do? To facilitate understanding of the term protocol, we take the example of people talking to each other to explain. For example: Inquiry time, ask questions. Clearly, the message sent or received and when these messages are sent or received when the action taken, in the human protocol plays the most important role. If you talk to both sides saying the other side cannot understand the language, you cannot communicate, which is the same network protocol, but received not a human entity to send a message, but in the host's hardware and software entities only. Network protocols: machines, rather than humans, exchange msgs all communication activity in Internet governed by protocols Protocols define format, order of msgs exchanged between network entities, and actions taken on msg transmission, receipt or other events 4

Human protocol and computer network protocol:

1.1 Whats the Internet?-summary Nuts-and-bolts description The basic hardware and software components that make up the Internet. A service description Describing the Internet in terms of a networking infrastructure that provides services to distributed applications. Protocols Define format, order of msgs exchanged between network entities, and actions taken on msg transmission, receipt or other events. 1-2 The Network edge Introduction network edge: hosts and applications services network core: routers and switches Circuit/Packet switching access networks, physical media: physical media access types The edge of the network, we understand the main host, applications, etc., we also detail connectionless and connection-oriented services. In the network and the core part, the main understanding of circuit switching and packet switching technology, multiplexing. Part in the access network, we mainly need to understand the way access networks and access networks using a variety of physical media 1 End systems Computers connected to the Internet are referred to as end systems. sit at the edge of the Internet Run application programs, so also referred to as host. e.g., WWW, email Access to the Internet computer (computing device) in the network at the end is generally called the end system. End system is also called the host, because they hold the application. 1 Clients, and Servers Two categories of hosts : Clients and Servers Client (program) host requests, receives service from server (program) Client/Server Internet applications are distributed applications Client program runs on one PC, and the server program runs on another PC. 5

Host client and server are generally divided into two categories. Clients refer to the processing power in general, request and receive resources or services equipment; and the server refers to have a strong capacity to provide resources or services equipment. This client-server model is the most mainstream of the web application model, this time; Internet can be seen as a black box.

Peer to Peer model (P2P) host interaction symmetric e.g.: teleconferencing emule In fact, a host can be both a client and a server; even both client and server. For example, in P2P, a host of other peers as a client to request resources or services to another host and can provide resources or services. 2 Services End system programs use the services of the Internet to exchange messages. Connectionless service UDP: user datagram protocol User Datagram Protocol Connection-oriented service TCP: transmission control protocol Transmission Control Protocol End system using the Internet to communicate with each other provided services. The Internet provides two types of services: Connection-oriented: the communicating parties before a formal data transmission to send control packets to establish a connection establish and notify the parties about the connection status. The relationship between the communicating parties to establish a relatively loose, packet transmission path / route through which the packet switches are not maintenance-related condition, which is called "connection-oriented" and not directly called "Connect" service reasons. Connectionless Service: connectionless service. No handshake. 2 Connection-Oriented Service Connection-Oriented service handshaking setup (prepare for) data transfer ahead Sends control packets to each other to establish the connection. end system allocates resources for the connection buffer/bandwidth/states Why connection-oriented and not just connection? Connection established between hosts is very loose and only the end systems are aware of the connection. Only the end system allocates buffers, state, fort the connection TCP reliable, in-order byte-stream data transfer Connection is equivalent to a water pipe loss: acknowledgements and retransmissions flow control: a sender wont overwhelm the receiver Congestion Control senders slow down sending rate when network congested 6

TCP (Transmission Control Protocol, Transmission Control Protocol) is a connection-oriented protocol, before the data transmission through three-way handshake to establish a full duplex connection, before the real data. In addition to TCP is a connection-oriented protocol, but also to provide reliable, on-demand to reach the byte stream data transfer, flow control and congestion control. 1, reliable - in sequence - a stream of bytes transmitted: through the recognition and retransmission to ensure reliable data transmission; TCP entity in order to submit the application data; each byte has its own number, but whether each byte meaning. 2, flow control: The main thing to avoid low-end system will not be sending data to the end overwhelmed. The basic mechanism is the buffer. 3, congestion control: to avoid the network into a state called Gridlock, which detects when the network is congested conditions to reduce their sending rate. Gridlock is a situation that ultimately comes from a region such that all intersections are blocked by vehicles, no vehicle can move such a state; dining philosophers problem with this possibility. 2 Connectionless Service Connectionless Service = no handshaking UDP (User Datagram Protocol) Connectionless service is sending the data connection is not established before the process of application to send data on what time to send data, so applications using this service, the data from the source to the target side of the time may be short. However, this service generally does not guarantee reliable data transmission, does not provide flow control and congestion control. Generally not used to losing some of the lethal effects of data applications such as IP telephony, video conferencing. 2 Services TCP or UDP? Apps using TCP WWW, FTP, E-mail Apps using UDP Internet Phone, video-conferencing If the apps want control what and when the data sent, choose UDP. If apps want reliable data transfer, choose TCP. Summary End systems/hosts are sit at the edge of the Internet, and host applications. Client/Server P2P End systems use the services of the Internet to send msgs to each other Connection-oriented serviceTCP Connectionless serviceUDP Applications choose transport service according to their requirements.

1-3 The Network core Introduction Network core is a mesh of routers and links that interconnect the hosts. The main task of the core network from the edge of the network to reach its destination packet forwarding. For the network core switching technology is mainly two: 1. Switching techniques Two approaches to build a network core circuit switching: dedicated circuit per call telecommunication network packet-switching: data sent thru net in discrete chunks (packets) Internet

Circuit Switched (circuit switching): In the communication process, the network for data transmission in the transmission path reserve resources, these resources can only be the Communication used by both sides; Packet switching (packet switching): Data is divided into a one group, each group have to carry the destination address, not for the packet transmission network in Reserve resources along the packet switches, packet switches for each packet independently to determine the forwarding direction. Circuit Switching Reserve resources needed along a path for any call for the duration of the session. bandwidth and/or switch capacity dedicated resources allocation guaranteed performance call setup required
Circuit switching network, communicating parties must be established before sending data in a dedicated circuit, to reserve resources along the way, including bandwidth and switching capacity

Packet Switching Resources are used on demand Resources are shared with other sessions No performance guarantee Circuit Switching

For circuit-switched technology brief. The slide chart: 1) the use of four four-link circuit switches to connect to the network. Each link has n circuits (circuit), can support n-connected. 2) Host and a switch. When two hosts pre-communication, the network between the host to establish a dedicated end to end connection. For example, the host communication between A and B before, each link in the network first reserve a circuit (circuit). 3) As a link up to the n circuits, each circuit using the link bandwidth 1 / n.

Multiplexing in Circuit Switched Networks There are lots of circuits in a link. How to implement a circuit? network resources (e.g., bandwidth) divided into pieces a piece allocated to one call resource piece idle if not used by the owning call (no sharing)
In a physical link on a number of circuits (circuit), the basic approach is to achieve the circuit resources into multiple small pieces, a small piece points to a call (call), when this piece of free resources to other call cannot be used, that is resources are exclusive. Circuit-switched data in the resource exclusive cause low utilization, which is the computer network using packet switching the main reason. 8

Whats multiplexing Multiplexing is sending multiple signals on a carrier at the same time in the form of a complex signal and then recovering the separate signals at the receiving end. frequency-division multiplexing-FDM Wavelength division multiplexing-WDM time-division multiplexing-TDM code-division multiplexing-CDM
"Multiplex multiplexing" is the term commonly used in communications technology, is in the same transmission medium can transmit multiple signals simultaneously in the technology, aims to improve the utilization of communication lines. Commonly used methods are frequency-division multiplexing, TDM, CDM code division multiplexing, etc.

FDM The frequency spectrum of a link is divided into bands and shared among the connections established across the link. Bandwidth: the width of the band

FDM (Frequency Division Multiplexing, Frequency Division Multiplexing): frequency-division multiplexing is the use of different frequencies to transmit different signals simultaneously without any interference. Spectrum of the physical link is shared by multiple circuits, but use different frequencies for each circuit only. For example, the voice frequency range 4 KHz, a link will be divided into a spectrum of 4 KHz wide frequency band, a band assigned to a Call to use, as a Circuit. Bandwidth: The width of the band is called the bandwidth (bandwidth).

Interference

TDM For a TDM link, time is divided into frames of fixed duration, and each frame is divided into a fixed number of time slots. The network dedicates one time slot in every frame to the connection.

TDM (Time Division Multiplexing, Time Division Multiplexing): time division multiplexing is the use of different time slots send different signals. The time is divided into a fixed long Frame (structural, frame, etc.), each Frame is divided into a fixed number of time slots (time slot). When the connection is established, the network connection for each Frame in a fixed allocation of time slots (slot). Multiplexing in Circuit Switched Networks Circuit switching is wasteful because the dedicated circuits are idle during silent periods. Example: Calculate the time to transfer a file from A to B (in the slide) Suppose file size: 640,000 bits setup time:500 msec; link rate:1.536Mbps; a frame=24slot time=640000/(1.536Mbps/24)+0.5s=10.5s Packet Switching Long msgs are broken into packets. There has a destination address of the destination in each packet and used by the packet switches to determine output link Packets are transmitted over each link at full rate of the link resources used as needed, no setup Resources are shared among multiple user
The basic idea of packet switching: the data into a one packet, the destination address of each packet are carried along by the packet through the packet switches based on the destination address carried by the decision of its output link. And circuit switching, link, switch / routers and other resources shared by multiple users, the switch forwards a packet to its output link speed full speed.

Packet Switching: Store-and-Forward transmission Packet Switches must receive the entire packet before it can begin to transmit the first bit of the packet onto the outbound link. it introduces a store-and-forward delay at the input to each link along the path. The delay is proportional to the packets length Packet Length = L bits and the link rate =R bps Store-and-forward delay=L/R s Generally use packet switching Store and Forward (Store and Forward) technology, that switch must receive a complete packet before the packet transmitted on the outbound link first. Using Store and Forward will introduce a delay, this delay and packet length is proportional to the speed and inversely proportional to the output link. Packet length: L bits Q links from source to destination link rate: R bps no queuing delay and no propagations delay store and forward delay in each link is L/R s end to end delay is QL/R s

Packet Switching: Buffering in Packet Switches If a packet needs to be transmitted across a link that is busy now, it must be stored in the buffer temporarily and wait for its turn. Queuing delay Packet loss: if the buffer overflows Packet switching in storage group to use buffer (buffer / cache), cache is generally set at the switch output (this switch is called the output buffer switch). When arriving from multiple input links from the
11

same output link packet forwarding, the switch can only transmit a packet on the output link, the other group waiting in the buffer required output link is idle. Can be seen in the packet switches in the group will go through a line (queuing) delay. Queuing delay on busy with the switch, the size variable. If the cache is full when the packet arrives, the switch will discard a packet. Statistical Multiplexing of Packet Switched Network

Sequence of A & B packets does not have fixed pattern statistical multiplexing. In TDM each host gets same slot in revolving TDM frame. The picture shows a simple packet switch network. Suppose A, B also sends to the E group. Statistical multiplexing "is actually a time-division multiplexing technology. Full name is called" statistical time division multiplexing ", referred to as STDM, also known as" asynchronous time-division multiplexing. " The so-called "asynchronous" or "statistics" because it uses common channel "slot" method and the traditional time-division multiplexing methods, the traditional time-division multiplexing access terminals are fixed to each assigned a common channel of a time slot, is pigeon-holing. Because of the terminal and the slot is "condemnation", so they are "synchronous". The asynchronous time division multiplexing or statistical time division multiplexing is to implement public channel time slot, "according to need", that is only for those who need to transmit information or termination to work before being assigned to the slot, thus making all of the time slots are can get full use. Statistical analysis, statistical multiplexing than traditional time-division multiplexing to improve transmission efficiency 2-4 times. The main features of this multiplexing is to dynamically allocate channel time slot, so the statistical multiplexing can be called "dynamic multiplexing." Packet switching vs. Circuit switching Packet switching allows more users to use network than circuit switching! For Example: 1 Mbps link each user: 100 kbps when active active 10% of time circuit-switching: 10 users packet switching: with 35 users, probability > 10 active less than .0004

11

 Packet switches can send data at full rate of the outbound link. Example 10users, one suddenly generates one 1000-bit packet, while others remain silent. Link rate is 1Mbps, each user 100kbps Under circuit switching 1000/100000=10msec Under packet switching 1000/1000000=1msec 2. Packet Switched Networks datagram network: destination address determines next hop routes may change during session analogy: driving, asking directions virtual circuit network: Each packet carries tag (virtual circuit ID), tag determines next hop. And fixed path determined at call setup time, remains fixed thru call routers maintain per-call state There are two types of packet-switched network, classification is based on packet-based switch forwards the packet destination address or Virtual circuit identifier: 1, Datagram (packet) network: According to the packet destination address to forward packet network, eg Internet. Each packet contains the destination address and other information are Interest in the head, when the packet arrives the switch, the switch according to the packet destination address and a routing table to determine the packet forwarding direction. 2, Virtual Circuit virtual circuit network: number of forward packets according to virtual circuit network, eg, X.25, ATM, etc. Virtual circuit network, data transmission before Need to use signaling to establish a virtual circuit end to end, with the VC ID to identify and fill in each packet, called a tag. Switch according to the Group in the tag to determine its forward direction, which is no longer checks the destination address. Each virtual circuit switches the connection needs to maintain ongoing Access information. Network Taxonomy

Summary of the network core Network core is a mesh of routers move packet from source to destination Circuit switching and packet switching multiplexing in circuit switched networks store-and-forward (delay) buffering and queuing delay circuit switching vs. packet switching packet switched networks Network Taxonomy
12

1-4 Access networks and physical media Introduction Access Networks is the physical link(s) that connect a host to its edge router edge router :the first router on a path from a source host to the destination host. Categories: Residential access Company access wireless access Keep in mind: bandwidth of access network shared or dedicated Host access network is connected to the border router of the physical link (last mile), divided into the home access, access, and wireless access to three types of units. 1.1 Residential Access Connect home end systems into the Network (edge router). Dial-up MODEM ordinary analog telephone line + modem up to 56Kbps direct access to router (conceptually) access network = a host-side modem + a telephone line + a router-side model Cant surf and phone at same time The link is dedicated! Early home Internet access is usually dial-up networking, using ordinary telephone line modem (UTP cat.3) up to a maximum data transfer rate of 56kbps, this time at the border routers also need a MODEM. Therefore, at this time is the access network includes a point of the MODEM and a telephone line. As the lower rate, cannot make phone calls and Internet access. Broadband residential access DSL: (digital subscriber line) a kind of new modem technology running over traditional telephone line. restricting transmission distance to achieve higher rate DSLs link is dedicated The data rate are typically asymmetrical in the two directions. ADSL: (asymmetric digital subscriber line) Download:10Mbps Upload:1Mbps HFC: (hybrid fiber-coaxial cable)
Many households use broadband access technologies such as xDSL and HFC. xDSL is the analog telephone lines to transmit digital signals, it uses a new modem technology and limits the maximum transmission distance, it can transfer data at higher speeds. The use of ADSL, phone and Internet can be, both independently of each other. ADSL uplink and downlink speeds of different. Uplink speeds of up to 1Mbps, downlink speeds of up to 10Mbps.

xDSL uses FDM 50 kHz - 1 MHz for downstream 4 kHz - 50 kHz for upstream 0 kHz - 4 kHz for telephone DSL uses frequency division multiplexing, the communications link is divided into three frequency channels and do not cover, namely: 1,0 ~ 4KHz two-way voice channels 2,4 KHz ~ 50KHz uplink data channel 3,50 KHz ~ 1MHz downlink data channel
13

 HFChybrid fiber-coaxial cable ) extended from broadcasting cable television. require cable modem to convert signal also asymmetric up to 1 Mbps upstream >= 10 Mbps downstream shared link bandwidth with other users Television system is a broadcasting system In the cable TV system, cable head stations located all users to radio and television signals, television signals along the station -> the user the direction of transmission and amplification. HFC (hybrid fiber coaxial cable network), Host Modem need to use the device called a cable access network, Cable Modem uplink and downlink will link into two channels. As the channel is shared among multiple users, so there is network congestion and the size of the problem. Similar to ADSL, HFC upstream channel rate is lower than the downlink channel rate, and the entire channel is shared by all users. While ADSL uses Point to Point Channel, is the dedicated channel. 1.1 Residential Access - HFC

14

1.2 Company Access LAN company/units typically use LAN to connect end systems to edge router LANLocal Area Network Ethernet is by far the most prevalent technology for company access 10/100Mbps, Gigabit Ethernet, 10GE shared media twist pair ,coaxial cable, fiber Press coverage of computer network, divided into LAN, MAN and WAN. (Emphasis needs to know according to the classification of computer network coverage) LAN: local area network: a few hundred meters to several kilometers covered WAN: wide area network: MAN: Metropolitan Area Network: a city covered 1.3 Wireless Access Wireless LAN transmit/receive packets to/from a base station (access point, AP) base station connected to a wired router Wireless LAN standard: 802.11a(2Mbps) 802.11b(11Mbps) 802.11g(54Mbps)
In addition to WLAN, but also can use the wireless WAN technology, access to Internet, such as mobile phones. WAN technologies on the use of unlimited access, see p30 ~ 31

wider-area wireless access provided by Telco operator 3G ~ 384 kbps Will it happen? WAP/GPRS in Europe wireless access protocol General Packet Radio Service

15

1.3Wireless Access-Home networks Typical home network components: ADSL or cable modem router/firewall/NAT Ethernet wireless access point (AP) Head end Data Repeater 2.1 Physical Media One bit passes through a series of transmitter-receiver pair For each transceiver pair, the bit is sent by propagating electromagnetic waves or optical pulse across a physical medium. physical link lies between transmitter & receiver 2.2 Physical Media Categories of physical medium guided media The waves are guided along a solid medium such as copper, fiber or coax. unguided media The waves propagate in the atmosphere and in outer space, such as in a wireless LAN
Oriented media: media guide the direction of propagation of the signal Non-oriented media: general spread of the signal through the air, it does not guide the direction of propagation for the signal, such as short wave, microwave and infrared communications.

2.3 Twisted-Pair Copper Wire two insulated copper wires twisted together (TP) Category 3: traditional phone wires, 10 Mbps Ethernet Category 5: 100Mbps Ethernet Shield or Unshielded TP STP-Shielded TP UTP-Unshielded TP 2.4 Coaxial Cable two concentric (rather than parallel) copper conductors Base-band single channel on cable Broadband: multiple channels on a cable HFC Fiber Optics glass fiber carrying light pulses, each pulse represents a bit high-speed and lower error rate

16

Radio signal carried in electromagnetic spectrum no physical wire and bidirectional propagation environment effects: reflection obstruction by objects interference Radio Link Types terrestrial microwave e.g. up to 45 Mbps channels WLAN (e.g., Wi-Fi) 2Mbps, 11Mbps, 54Mbps wide-area (e.g., cellular) e.g. 3G: hundreds of kbps satellite up to 50Mbps channel (or multiple smaller channels) 250 msec end-end delay 1.4 Access Network and Physical Media Access Networks is the physical link (s) that connect a host to edge router Residential, company and wireless access Dial-up modem, xDSL, HFC LAN, Ethernet, Fast Ethernet, GE, 10GE WLAN, 802.11 the bit is sent by propagating electromagnetic waves or optical pulse across a medium. Guided media and unguided media Twisted-pair, coaxial cable, fiber optics and radio 1-5 ISPs and Internet backbones Internet Structure ISPs connect to each other and form a tiered hierarchy logical topology At very top of the hierarchy is a relatively small number of socalled Tier-1 ISPs. Tier-1 ISPs Tier-1 ISPs are special Directly connected to each of the other Tier-1 ISPs Connected to a large number of Tier-2 ISPs International in coverage Link speed is high known as Internet Backbone networks

17

Access network connected to the Internet via an ISP, ISP component hierarchy. The top level is called Tier-1 ISP. Tier-I ISPs are also called the Internet Backbone, which is characterized by: 1, with the other Tier-1 ISP is directly connected (to facilitate billing?) 2, with most Tier-2 ISP Internet 3, covering international, regional POP (Point of Presence): the points at which the ISP connect to other ISPs; POP belongs to a provider's network; A POP is a group of routers NAP (Network Access Point, not nap): are complex high speed switching networks, often concentrated in a single building, and can be owned and operated by the third party exchange huge quantities of traffic among many ISPs Increasingly, tier-1 ISPs interconnect with each other at POPs

Tier-2 ISPs Tier-2 ISPs: smaller (often regional) ISPs Connect to one or more tier-1 ISPs, possibly other tier-2 ISPs

Tier-2 ISP's coverage area, but a region or country. Tier-2 ISP usually with one or more Tier-1ISP connected, you can also connect with other Tier2 ISP.
18

Tier-3 ISPs and Local ISPs last hop (access) network (closest to end systems)

Internet Structure a packet passes through many networks!

1.6 Delay & loss in packet-switched networks How do loss and delay occur? packets queued in switch buffers and wait their turns to output packet arrival rate exceeds links capacity packets en-queue, wait for turn

Packet switching network, a router uses storage - forwarding technology. Input from the router to reach the group may need to wait in the queue the output link is idle. When the packet arrives at the location when the queue is not idle, the packet might be discarded. Packet source to the destination in the transmission process, in each node will experience delay. Packet delay in a node includes the node processing delay, queuing delay, send delay, propagation delay.
19

1. Types of Delay Nodal processing delay Time required to examine the header and determine where to direct the packet Queuing delay Time a packet waits to be transmitted in queue Transmission delay Time required to push all bits of a packet into link Propagation delay Time required to propagate from one end to another end of the link 1.1 Processing Delay nodal processing determine output link check bit errors on the order of microseconds or less after nodal processing, the router directs the packet to the queue After a node packet processing to the next node in the process, the delay experienced, there are four: 1, nodal processing delay (processing delay): refers to the router / switch, check the packet header to determine which output port to forward from the time required. Node processing delays may also include checking whether an error packet the time required. After processing delays, the packets are cached in the queue. 1.2 Queuing Delay Wait time for the packets turn to be transmitted onto the link depends on congestion level of router on the order of microseconds to milliseconds

queuing delay (queuing delay): grouping the output link is idle waiting for the time required. From the beginning to the grouping of packets into the queue to leave when the time difference. Queuing delay experienced by packets already in the queue depends on the total length of the packet and queue strategy. 1.3 Transmission Delay Time to push all of the packet bits onto the link Also call store-and-forward delay R=link bandwidth (bps) L=packet length (bits) time to send bits into link = L/R Transmission delay: the packet is sent to link all the bits on the time required, also known as Store and Forward Delay. 1.4 Propagation Delay propagation delay dprop = d / s s=2.0~3.0108m/s

21

Propagation delay: bit transmitted from one point to another time experienced, and the distance between two points and the signal propagation speed. Delay and propagation delay on sending easily confused. Need to remember that a packet transmission delay refers to the bit sent to the link on all the time, and the packet length and transmission rate, and the distance between two points is not in any way. The propagation delay refers to a communication from one end of the link to the other end of the time, and the link length and signal propagation speed.

1.4 Transmission vs. Propagation Delay The transmission delay is the amount of time required for the router to push out the packet. function of the packets length and links transmission rate The propagation delay is the time it takes a bit to propagate from one router to the next. function of the distance between the two routers 1.4 Transmission vs. Propagation Delay Caravan analogy car ~ bit ; caravan ~ packet ; tollbooth ~ router; highway ~ link

car travels at rate of 100km/h; tollbooth takes 12s to serve a car propagation speed and transmission time Suppose: when the first car arrives at a tollbooth, it waits at the entrance until the others have arrived. Q: How long until caravan is lined up before the 2nd tollbooth? time to push entire caravan through the tollbooth onto highway = 12*10 = 120 sec time for last car to propagate from 1st to 2nd tollbooth: 100km/(100km/hr)= 1 hr A: 62 minutes car travels at rate of 1000km/h; tollbooth takes 60s to serve a car Q: Will the 1st car arrive to 2nd booth before all other cars served at 1st booth? Yes! After 7 min, 1st car arrives at 2nd tollbooth and 3 cars still wait at 1st tollbooth. 1st bit of packet can arrive at 2nd router before packet is fully transmitted at 1st router! Nodal Delay

Node delay and the delay of the four parts. Processing delay (dproc) typically a few microsecs or less queuing delay (dqueue) depends on congestion transmission delay (dtrans) = L / R, significant for low-speed links propagation delay (dprop) a few microsecs to hundreds of msecs 2. Queuing delay Traffic intensity = La/R R: link bandwidth (bps) L: packet length (bits) a: average packet arrival rate

21

La/R ~ 0: average queuing delay small La/R 1: delays become large La/R > 1: more work arriving than can be serviced, average delay infinite! Design your system so that the traffic intensity is no greater than 1 !!!

Node queuing delay is the delay of the most complex and most interesting part. The reason why the most interesting, or refer to the current research work is carried out for the queuing delay, including scheduling, caching strategies. Queuing delay and the load status of network devices is closely related to different groups experienced queuing delay will change with the load.

2. End-to-End Delay End-to-End delay is a a packets delay form source end-system to destination end system. dend-end= (dproc+dtrans+dprop+dqueu) Real Internet delays and routes What do real Internet delay & loss look like? trace route ( tracert (XP)) : Measurement delay from source to router along end-end path towards destination. for all I: sends three packets that will reach router i on path towards destination router I will return packets to sender Sender times interval between transmission and reply.

Summary type of delays processing, transmission, queuing and propagating delay End-to-End delay from source to destination buffering and queuing delay buffering and packet loss delay and routes in the Internet

22

1.7 Protocol layers and their service models Introduction Networks are complex! many pieces: Hosts and packet switches links of various media applications protocols hardware, software Question: Is there any hope of organizing structure of network? Or at least for our discussion of networks?
The Internet is a very complex system; there are many different types of applications, protocol, end systems, connections. Is there a problem is structured to describe something / build Internet?

Layering of airline functionality

Layers: each layer implements a or some service(s) via its own internal-layer actions relying on services provided by layer(s) below
Hierarchical structure is conducive to modular processing of complex systems, is conducive to the realization of system services. For a complex system, the system may continue to implement changes in the hierarchy will help to adapt to these changes. Air passenger transport functions are divided into different levels; we discuss a framework to provide air travel. Each layer with the following features provided by the combination of level, have achieved a certain function, and called Service (Service). Layer in the ticket: to provide a passenger service counter to counter, as long as you buy a ticket, your airline can use the aircraft to the destination. Layer in the baggage: Passengers can check-in baggage, arrive at your destination can retrieve their luggage. Note, baggage only purchased tickets for passenger's baggage service.

Why layering? Dealing with complex systems: explicit structure allows identification relationship of complex systems pieces layered reference model for discussion modularization eases maintenance, updating of system change of implementation of layers service transparent to rest of system e.g., change in gate procedure doesnt affect rest of system
Agreement with the concept of layering and structural benefits. First, the agreement provides a discussion of hierarchical structured approach to complex problems; modular system to achieve beneficial updates and changes. Protocol layer has its disadvantages: 1) protocol layer may result in duplication of certain functions to achieve; 2) a layer of some services may need to use it with some of the other layers of information, in violation of the principle of layering. Such as TCP checksum calculation on the IP layer need some information.

23

Internet protocol stack application: supporting network applications transport: process-process data transfer network: host-to-host data transfer, routing of datagram's from source to destination link: data transfer between neighboring network elements physical: bits on the wire

The Internet is a very complex system, including hardware and software, including application, protocol, end systems, different types of communication media, routers / switches. The Internet architecture is also used in hierarchical structure, with the previously mentioned examples of similar passenger; Internet use for each layer is the layer of the upper or lower function to provide one or more services. Each protocol layer can use the software, hardware or a combination of both to achieve. Architecture: a collection of protocols and layers. Collection of the agreement also called protocol stack (Protocol Stack), Internet protocol stack, including: the application layer, transport layer, network layer, link layer and physical layer. Application Layer: contains a large number of applications generally require the agreement; application of the message called the message. Transport Layer: responsible for receiving messages from the application layer and transport the application layer message, the message will be handed over to the destination application. The group is called the transport layer segment. Network layer: the transport layer is responsible for the source Host segment to the network layer, network layer is responsible for the segment transmitted to the destination host's transport layer. Link layer: network layer is responsible for passing data between the source and destination, Link layer is responsible for packet transmission from one node to the next node. Link layer data transmission unit is called the Frame. Physical Layer: Link layer is responsible for a Frame from a Node to pass to the next Node, is responsible for the physical layer of each Frame (bit) transmitted from one end of the link to the other side.

Applications send and receive data is a complex process, the source data by the application layer vertical topdown pass, will message to the Transport layer, fill out the message that the application of information, fitted into segments, to the network layer; network layer protocol entities included in the segment to add the destination address outside of the head, loaded into a datagram (we used to call packet, packet) and turned over to the link layer; link layer in the packet's neighbor on the outside package node's link-layer address information, pretending to convert the physical layer frame by light signals transmitted on the media. After the middle of switches, routers forward the equipment to reach the destination after the data and then submit up layer by layer, each corresponding to submit the protocol headers are removed, and the final message to the application. It can be seen; the data transmission layer in the source Host to have a package (encapsulation) process, the destination Host opened a layer of the process. Data transmission may take one or more switches and routers. Both devices call Packet switch, but the highest level of different work. Routers work at the highest level of network layer; link layer switches work at the highest level of link layer. 24