Introduction Computer Networks and the Internet

Learning Outcomes 159.334 At the end of this session, the students should be able to: Explain what the Internet is all about Explain what is a protocol Describe what comprises the network edge Describe what comprises the network core Explain connection-oriented service Explain connectionless service Compare circuit-switched network against packet-switched network Answer the short exercises given in the session

Introduction What’s the Internet? UNIX-based workstations laptop Digital cameras Email server When we think of the internet, so many things come to our mind. Nowadays, the internet connects almost anything imaginable. To take a peek of the few items connected to the internet, we have… portable device assistant, and so on and so forth. You must be wondering why a toaster appears in here.. Well, out of curiosity, I looked for that myself.. :http://news.bbc.co.uk/1/low/sci/tech/1264205.stm Automobile Web-page server WebTV PDAs with wireless Internet connections toaster Household appliances HOSTS or END SYSTEMS

Nuts and bolts of the Internet Networking Infrastructure What’s the Internet? Nuts and bolts of the Internet Hardware components Software Networking Infrastructure provides services to distributed applications infrastructure where new applications are being constantly invented and deployed We shall define the Internet in two ways in our discussions. First, we shall look at it based on its working parts or elements – and so we’re using the term “nuts and bolts”, then we shall also look at it in an abstract perspective, by looking at it as the underlying foundation or basic framework where new applications are constantly invented and deployed.

What’s the Internet? “Nuts and Bolts” View global network of networks Interconnects hundreds of millions of computing devices provides: Global communication Storage Computation infrastructure The internet is viewed as the global network of networks. It now interconnects millions of computing devices, providing global communication, storage and computation infrastructure. Before delving into the nitty-gritty part, let’s have a look at the bigger picture first in a much simpler perspective… that is, in terms of an END-to-END SYSTEM. End-to-End System: Core End System End System “edge” “edge”

SETI@HOME—MASSIVELY DISTRIBUTED COMPUTING FOR SETI What is SETI@home? SETI@home is a scientific experiment that uses Internet-connected computers in the Search for Extraterrestrial Intelligence (SETI). You can participate by running a free program that downloads and analyzes radio telescope data. Search For Extra-Terrestial Intelligence http://setiathome.ssl.berkeley.edu/

The Network Structure network edge: applications and hosts network core: routers network of networks access networks, physical media: communication links

What’s the Internet? “Nuts and Bolts” View End-to-End System End System=HOST Core End System End System Access networks Physical media Where much internet architecture complexity is placed Connect end systems to the network core Communication links Switches Transport data Much of the Internet architecture can be found in the END SYSTEM and likewise, the bulk of work. Rapid application development has been seen on this component, triggering an expansion of Network technology because of the high demands of evolving applications. To mention a few, I think most of us are familiar with Peer 2 Peer networking, where you can get everything for free (that application however is barred from our network – because it disobeys copyright laws – it is illegal). Moreover, there is another interesting application called Skype that provides internet telephony – with that, you can call anyone in the world for free (if the other person is using Skype) in stereo quality. On the other hand, what can be found in the core are the communication links and switches that transport data, as well as Access networks and Physical media that are responsible for connecting the END SYSTEMS to the network core. The communication links are made up of different types of physical media like copper wire…. And are characterized in terms of bandwidth or link transmission rate, measured in terms of bits/second. Communication links Characterized in terms of bandwidth Made up of different types of physical media: Link transmission rate Coaxial cable Measured in bits/second Copper wire Fiber optics Radio Spectrum

Where is the Network Core? Let us just identify where in the figure is the network core…. All components colored RED in the figure corresponds to the network core. The structure that connects the end systems to the internet is the network core.

What’s in the Links? Router Router X NETWORK CORE Sender Receiver Router End System End System X End System End System path or route What’s in the Links? End System Router Internet uses packet switching to allow for multiple communicating end systems to share a path, or parts of a path, at the same time Takes a chunk of information arriving on one of its incoming communication links and forwards that chunk of information on one of its outgoing communication links The most common component that can be found in the network core is the router. As you can see in the figure, the function of a router is to provide a path from a node on one network to a node on another network. The figure is a simplified illustration of how routers allow for the connection, but in real networks, the two networks may be actually separated by several intervening networks and, possibly, by many miles. The router provides the path by first determining a route and then providing the initial connection for the path. packet

Now let’s see how are links formed in terms of clusters Now let’s see how are links formed in terms of clusters.. Let’s see the bigger picture

Get http://www.massey.ac.nz/ What’s a protocol? a human protocol and a computer network protocol: TCP connection request Hi Hi TCP connection reply Got the time? Get http://www.massey.ac.nz/ 2:00 <file> time

What’s a protocol? Human Protocols: Network Protocols: Something we execute all the time Offer a greeting Wait for a response Analyze the response Act accordingly Network Protocols: Similar to human protocol, except that entities are machines rather than humans all communication activities in the Internet are governed by protocols In order for protocols to work, both entities must observe the same protocol. There is a set of conventional actions taken when messages are sent and received. Networking – understanding the what, why and how of networking protocols

What’s a protocol? A protocol defines: format and order of messages sent and received among network entities and actions taken on the transmission and/or receipt of a message, or other event * All activities in the Internet that involves 2 or more communicating entities are governed by a protocol. There are protocols in: Routers Protocols determine a packet’s path from source to destination NIC hardware-implemented protocols control the flow of the bits on the “wire” End Systems congestion-control protocols control the rate at which packets are transmitted between sender and receiver Communicating Entities: Hardware, Software components Different protocols are used to accomplish different communication tasks:

What’s the Internet? “nuts & bolts” view Protocols - control the sending and receiving of information within the Internet - run by End Systems, routers, etc.; TCP Two of the most important protocols in the Internet (principal protocols) IP TCP – Transmission Control Protocol IP – Internet Protocol – specifies the format of the packets that are sent and received among routers and end systems The internet is governed by a set of predefined rules in order to allow processes to communicate and exchange data.. In internet jargon, that is referred to as protocols. INTERNET STANDARDS Made possible through standards developed by (IETF) Internet Engineering Task Force RFCs (Request for Comments) define protocols such as TCP, IP, HTTP, SMTP

What’s the Internet? A Service View Provides a communication infrastructure that allows distributed applications running on its end systems to exchange data with each other. Remote login email Web surfing Instant messaging Internet telephony “the Web” – distributed application that use the communication services provided by the Internet Communication services provided to distributed applications: Connection-Oriented Reliable Service Guarantees that data is delivered orderly and completely (sender to receiver) Connectionless Unreliable Service Delivery is not guaranteed

Question ? Why would we opt for a connectionless unreliable service when there is a connection-oriented reliable service that is available? Hold on to that thought for a while…

A closer look at the Network Edge What happens in the network edge? The sending End System doesn’t know how messages are actually sent. It only needs to know what services are provided, and so the “nuts and bolts” of the Internet serves as a “black box” that transfers messages between distributed communicating components. There is some level of abstraction that hides the nitty-gritty part of the communication process between two end systems Client/Server Model - Most prevalent structure for Internet applications; although not all applications are purely client, or purely of server type (e.g. P2P file sharing)

A closer look at the Network Edge What happens in the network edge? End Systems (Hosts): run application program e.g., WWW, email at “edge of network” Client/Server Model client host requests, receives service from server e.g. WWW client (browser)/server; email client/server Client/Server Model - Most prevalent structure for Internet applications; although not all applications are purely client, or purely of server type (e.g. P2P file sharing)

The Network Edge “Connection” between two End Systems: (e.g. Web application or Internet phone application) Nothing more than allocated buffers and state variables in the End-Systems “Connection” Internet provides two type of services to End-System Applications: Connection-oriented service – (TCP) App’s using TCP: HTTP (WWW), FTP (file transfer), Telnet (remote login), SMTP (email) The term connection corresponds to nothing more but allocation of buffers and state variables. Only the communicating End-Systems are the ones who knows that they’re connected.. not even the routers of the intervening packet switches knows about any connection-state information between the communicating End-Systems. Connectionless service – (UDP) App’s using UDP: streaming media, teleconferencing, Internet telephony

Network edge: connection-oriented service performs handshaking Goal: data transfer between end system. handshaking: setup (prepare for) data transfer ahead of time Hello, hello back human protocol set up “state” in two communicating hosts Q* TCP service [RFC 793] Provides: Reliable data transfer: loss: handled using acknowledgements and retransmissions Flow Control: Ensures that the sender won’t overwhelm receiver Congestion Control: Instructs senders to “slow down sending rate” when network is congested Prevents gridlock Any protocol that performs handshaking is a connection-oriented service. We use the term “connection-oriented” because the end systems are connected in a loose manner, only the end systems are aware of the connection. The routers or intervening packet switches do not maintain any connection-state information about the connection. The services provided by TCP, such as reliable data transfer, flow control and congestion control are by no means requirements of a connection-oriented service. One or two of them may be missing, and yet still the network could offer connection-oriented service. Let’s just have a look at the definition here, then we shall see some animations later demonstrating them. DISCUSS THIS JUST QUICKLY – use the next animation slides to explain Transmission Control Protocol (TCP) Internet’s connection-oriented service

Network Edge: TCP Service 3-way Handshake Control packet CONNECTION ESTABLISHED DATA acknowledgement request CLIENT SERVER CLICK TO ANIMATE Reliable data transfer is achieved through acknowledgements and retransmissions * Data is delivered without error and in proper order

Network Edge: TCP Service Handshaking Procedure: Case: Retransmission of Request Control packet Client assumes packet was lost, decides to retransmit Client is waiting for Acknowledgement DATA acknowledgement CLIENT SERVER CLICK TO ANIMATE Reliable data transfer is achieved through acknowledgements and retransmissions * Data is delivered without error and in proper order

Network Edge: TCP Service Problem occurs when one communicating End-System transmits faster than the other End-System CLIENT This End-System does not receive an acknowledgement yet, and so it issues another packet CLIENT Control packet CLIENT CLIENT SERVER CLICK TO ANIMATE Flow control forces the sending End System not to send too many packets too fast for the receiver TCP/IP provides the Flow control service

Network Edge: TCP Service Problem: Gridlock sets-in when there is packet loss due to router congestion The sending system’s message is lost due to congestion, and is alerted when it stops receiving acknowledgements of packets sent CLIENT SERVER Due to router congestion, the packets sent by the sending End system is lost. When that happens, the sending end system is alerted of congestion if it does not receive an acknowledgement for the packets it sent. TCP provides a congestion control service that forces the End systems to decrease the rate at which packets are sent during periods of congestion. You can view the applet from Kurose’s site demonstrating network congestion. Congestion control forces the End Systems to decrease the rate at which packets are sent during periods of congestion

Network edge: connectionless service No handshaking procedure; End-Systems just simply send the packet Goal: data transfer between end systems same as before! UDP - User Datagram Protocol [RFC 768]: Internet’s connectionless service unreliable data transfer no flow control no congestion control There is no handshaking; therefore, data can be delivered sooner, but there is no reliable data transfer guaranteed The sending program simply sends the packets The source never knows for sure which packets have arrived at the destination Internet telephony uses the connectionless service because of the application’ demand for speed and the application doesn’t necessarily require acknowledgement of data all throughout the communication session

Something to ponder on ? Transmission rate of the link (Bandwidth) – how many bits per second a network can transport Propagation delay (Latency) – how many seconds it takes for the first bit to get from the client to the server Besides bandwidth and latency, what other parameter is needed to give a good characterization of the quality of service offered by a network used for digitized voice traffic?

Answer A uniform delivery time is needed for voice, so the amount of jitter in the network is important. This could be expressed as the standard deviation of the delivery time. Having short delay but large variability is actually worse than a somewhat longer delay and low variability. Jitter – irregular random transmission time in the network

The Network Core mesh of interconnected routers

The Network Core the fundamental question: how is data transferred through net? circuit switching: dedicated circuit per call: telephone net packet-switching: data sent through net in discrete “chunks” Approaches to building a Network Core: So how is data transferred in the NETWORK CORE? There are primarily two approaches, either by circuit switching or by packet switching. Circuit-switching is used by the ubiquitous telephone network, while packet switching is used by the Internet, and is said to be the future of telephone networks.

Circuit Switching vs. Packet Switching The Network CORE Circuit Switching vs. Packet Switching A Restaurant Analogy What resources must be reserved? Circuit-switched Networks Resources are reserved for the duration of the communication session Restaurant which requires reservation With a reservation, you can order right away when you get there guaranteed seats Packet-switched Networks Let’s have a look at a simple restaurant analogy to be able to easily understand the difference between the two switching schemes. You can think of Circuit switched networks as being analogous to restaurants requiring reservations, giving you guaranteed seats. On the other hand, packet switched networks are analogous to restaurants that do not require any reservations whatsoever, therefore seats are not guaranteed; The network equivalent simply dispatch messages, which take up on resources on demand; therefore, messages may have to wait on a queue to be transported. Messages use the resources on demand; thus, may have to wait (queue) for access to a communication link Restaurant which does not require any reservation you may have to wait on a queue to be served no sure seats

Network Core: Circuit Switching End-end resources reserved for “call” Reserved link bandwidth, switch capacity Switches on the path between sender and receiver maintain connection state for the duration of the session Resources are dedicated; thus, no sharing Advantage: circuit-like (guaranteed) performance call set-up required (unless infinite resources are available) “Circuit” The resources reserved for the duration of the call are the link bandwidth and the switch capacity. Such resources will be made available solely to a specific user in the network. Therefore, there is absolutely no sharing of resources between users or circuits. The advantage of this scheme is that circuit-like performance is guaranteed.

Network Core: Circuit Switching How is it implemented? By dividing the link bandwidth into “pieces” frequency division time division Inefficiency: Resource piece is idle if not used by owning call (no sharing)

Circuit Switching: FDM and TDM 4 users (or 4 circuits) Example: FDM frequency time 4KHz Network dedicates a frequency band to each connection for the session TDM frequency time Frame Slot In FDM, the frequency spectrum is shared among the users of the link. The link dedicates a frequency band to each user for the duration of the connection. (FM radio station – share microwave frequency spectrum, Telephone networks – freq. band = 4kHz or 4,000 cycles/sec). On the other hand, in TDM, the network dedicates one time slot in every frame of the connection. Two simple multiple access control techniques. Each mobile’s share of the bandwidth is divided into portions for the uplink and the downlink. Also, possibly, out of band signaling. As we will see, used in AMPS, GSM, IS-54/136 Used solely by one End System Network dedicates one time slot in every frame of the connection

Further assume that propagation delay is negligible. Question ? How long does it take to send a file of 640,000 bits from Host A to Host B over a circuit-switched network? Assume that all links in the network use TDM with 24 slots and have a bit rate of 1.536 Mbps. Also suppose that it takes 500 msec. to establish an end-to-end circuit before Host A can begin to transmit the file. Further assume that propagation delay is negligible.

Question Answer GIVEN: Size of file to send: 640,000 bits SOLUTION: How long does it take to send a file of 640,000 bits from Host A to Host B over a circuit-switched network? Assume that all links in the network use TDM with 24 slots and have a bit rate of 1.536 Mbps. Also suppose that it takes 500 msec. to establish an end-to-end circuit before Host A can begin to transmit the file. GIVEN: Size of file to send: 640,000 bits SOLUTION: Each circuit has a transmission rate of (1.536 Mbps)/24 slots= 64kbps (or 64,000 bps). So, it takes (640,000 bits)/(64,000 bps)= 10 sec. to transmit the file. Considering the circuit establishment time, we add 0.5 sec; therefore, It takes 10.5 sec. to transmit the file. The transmission time would be 10 sec. if the end-to-end circuit passed through 1 link or 100 links. (but the actual end-to-end delay also includes a propagation delay) Establishment time + transmission time

Network Core: Packet Switching each end-end data stream divided into packets user A, B packets share network resources each packet uses full link bandwidth resources used as needed Resource Contention: aggregate resource demand can exceed amount available congestion: packets queue, wait for link use store and forward: packets move one hop at a time transmit over link wait turn at next link Q* In contrast to circuit switching, packet switching allows for resource sharing. Data is transmitted on demand, without reservation, in terms of packets, using the full link bandwidth. This switching however could be fazed with a problem of catering to a multitude of packets exceeding the switch capacity, congestion, and store-and-forward delays Bandwidth division into “pieces” Dedicated allocation Resource reservation

We stopped here last time 

Network Core: Packet Switching Statistical multiplexing - on-demand sharing of resources 10 Mbs Ethernet C A statistical multiplexing Q* 1.5 Mbs B queue of packets waiting for transmission at the output link 45 Mbs Sender: Nodes A and B D E Receiver: Node E The figure demonstrates a simple packet-switched network. Assuming that Hosts A and B transmit a sequence of packets towards Host E. In the first packet switch, such packets will be received in random order. The link is not reserved for any sender, but is used on demand. In the jargon of networking, this on-demand sharing of resources is referred to as statistical multiplexing. The router employs statistical multiplexing to schedule the packets for transmission (this sharply contrasts circuit switching). sequence of A & B packets has no fixed timing pattern bandwidth shared on demand: statistical multiplexing. Compare this to TDM: each host gets same slot in revolving TDM frame.

Network Core: Packet Switching Consider a message that is 7.5 x 106 bits long. Suppose that between source and destination, there are 2 packet switches and 3 links, and that each link has a transmission rate of 1.5 Mbps. Assuming that there is no congestion in the network and negligible propagation delay, how much time is required to move the message from source to destination with packet switching? (7.5 Mbps/1.5 Mbps) * 3 = 15 sec. Transmission delay

Packet Switching: Store and Forward Behaviour Example: store and forward behaviour: break message into smaller chunks: “packets” Store-and-forward: switch waits until chunk has completely arrived, then forwards/routes Let us analyze how packet switching transports a message of size 5,000 packets. Take note that the unit of time used here is msec. From source to destination, taking the first packet, we see that packet one reaches the first packet switch at time = 1 msec. By the time packet 2 is transmitted, packet 1 is also being transported to the second switch. At this point, simultaneous transmission occurs, therefore, at time = 2 msec., packet 2 reaches the first switch, and packet one reaches the 2nd switch. Using this pattern, the last packet, 5,000 th packet reaches the first switch after 5,000 msec or 5 sec. adding 2 more switching, we get 5.002 sec = the time to reach its destination. Pattern that can be deduced from the packet flow depicted in the Figure: Time of arrival = packet_num + 2

Packet Switching vs. Circuit Switching Suppose that users share a 1 Mbps link, where each user alternates between generating data at a constant rate of 100 kbps, and periods of inactivity. Also assume that each user is active only 10% of the time. Compare the performance of Circuit Switching against Packet Switching.

Packet Switching vs. Circuit Switching Packet switching allows more users to use network! Example: 1 Mbit link shared by all users each user: Generates 100Kbps when “active” (at constant rate) active 10% of time circuit-switching: 10 users can only be supported 1,000,000 bits/sec divided by 100,000 bits/sec. packet switching: with 35 users, probability > 10 are active is less than .0004 probability <= 10 users are active is 0.9996 1 Mbps link N users Implies that 10 users can be using the circuit without competing, just like circuit-switching (bandwidth is equally distributed) Packet switching can cater to up to 35 users using the given 1 Mbps link. Using statistics, the probability that less than or equal to 10 users are active at a time is 0.9996 (almost equal to 1). When that happens, 10 users can be serviced in the same speed as circuit-switching. Packet switching allows for more than 3 times the number of users as compared to circuit-switching

Question (Transmission delay) ? A factor in the delay of a store-and-forward packet-switching system is how long it takes to store and forward a packet through a switch. If switching time is 10 µsec, is this likely to be a major factor in the response of a client-server system where the client is in Adelaide, Australia and the server is in Auckland, New Zealand? Assume the propagation speed in copper and fiber to be 2/3 the speed of light in vacuum. Speed of light = 3 x 108 meters/sec.

Question (Transmission delay) Answer A factor in the delay of a store-and-forward packet-switching system is how long it takes to store and forward a packet through a switch. If switching time is 10 µsec, is this likely to be a major factor in the response of a client-server system where the client is in Adelaide and the server is in Auckland? Assume the propagation speed in copper and fiber to be 2/3 the speed of light in vacuum. No. The speed of propagation is 200,000 km/sec or 200 meters/µsec. In 10 µsec the signal travels 2 km. Thus, each switch adds the equivalent of 2 km of extra cable. If the client and server are separated by 5000 km, traversing even 50 switches adds only 100 km to the total path, which is only 2%. Thus, switching delay is not a major factor under these circumstances.

Demo Total delay across a link = Transmission delay + Propagation delay

Network Core: Packet Switching Advantages: Great for bursty data resource sharing no call set-up Drawbacks: Excessive congestion, packet delay and loss protocols needed for reliable data transfer, congestion control Issue: How to provide circuit-like behaviour? bandwidth guarantees needed for audio/video apps this is still an unsolved problem!

Access networks and physical media Q: How to connect End- Systems to edge router? residential access nets institutional access networks (school, company) mobile access networks Keep in mind: bandwidth (bits per second) of access network? shared or dedicated?

Dial-up Modem uses existing telephony infrastructure telephone network Internet home dial-up modem ISP modem (e.g., AOL) home PC central office uses existing telephony infrastructure home directly-connected to central office up to 56Kbps direct access to router (often less) can’t surf, phone at same time: not “always on” Introduction 1-49

Central Office Example: A central office in Dakota, U.S.A. http://www.flickr.com/photos/afiler/3825218687/sizes/m/

Digital Subscriber Line (DSL) telephone network DSL modem home PC phone Internet DSLAM Existing phone line: 0-4KHz phone; 4-50KHz upstream data; 50KHz-1MHz downstream data splitter central office uses existing telephone infrastructure up to 1 Mbps upstream (today typically < 256 kbps) up to 8 Mbps downstream (today typically < 1 Mbps) dedicated physical line to telephone central office Works only within 5 to 10 miles from the CO. Introduction 1-51 For more info: http://www.systemtek.co.uk/modules.php?name=Content&pa=showpage&pid=18

Residential access: cable modems uses cable TV infrastructure, rather than telephone infrastructure HFC: hybrid fiber coax asymmetric: up to 30Mbps downstream, 2 Mbps upstream network of cable, fiber attaches homes to ISP router homes share access to router unlike DSL, which has dedicated access Introduction 1-52

Residential access: cable internet access Shared broadcast medium Diagram: http://www.cabledatacomnews.com/cmic/diagram.html Introduction 1-53

Cable Network Architecture: Overview Typically 500 to 5,000 homes cable headend home cable distribution network (simplified) Homes can be up to 100 miles from the cable headend Introduction 1-54

Cable Network Architecture: Overview server(s) cable headend home cable distribution network Introduction 1-55

Cable Network Architecture: Overview cable headend home cable distribution network (simplified) Introduction 1-56

Cable Network Architecture: Overview FDM: Channels V I D E O A T C N R L 1 2 3 4 5 6 7 8 9 cable headend home cable distribution network Introduction 1-57

Fiber to the Home optical links from central office to the home ONT OLT central office optical splitter Shared optical fiber optical fibers Internet Optical network terminator Optical line terminator optical links from central office to the home two competing optical technologies: Passive Optical network (PON) Active Optical Network (AON) – switched ethernet much higher Internet rates (download [10,20Mbps], upload [2,10Mbps]); fiber also carries television and phone services Introduction 1-58

Ethernet Internet access 100 Mbps institutional router to institution’s ISP Ethernet switch 100 Mbps 1 Gbps 100 Mbps server typically used in companies, universities, etc (Users:10 Mbps, 100Mbps), (Servers:1Gbps, 10Gbps Ethernet) today, end systems typically connect into Ethernet switch Introduction 1-59

Wireless access networks shared wireless access network connects end system to router via base station aka “access point” wireless LANs: 802.11b/g (WiFi): 11 or 54 Mbps Tens of meters from access point wider-area wireless access provided by telco operator ~1Mbps over cellular system (3G – packet-switched wide-area wireless internet access) Tens of kilometers from access point next up (?): WiMAX – IEEE 802.16 (10’s Mbps) over wide area router base station WiMax (Worldwide Interoperability for Microwave Access) mobile hosts Introduction 1-60

Home networks Typical home network components: DSL or cable modem router/firewall/NAT Ethernet wireless access point wireless laptops to/from cable headend cable modem router/ firewall wireless access point Ethernet Introduction 1-61

Physical Media Twisted Pair (TP) two insulated copper wires Category 3: traditional phone wires, 10 Mbps Ethernet Category 5: 100Mbps Ethernet bit: propagates between transmitter/rcvr pairs physical link: what lies between transmitter & receiver guided media: signals propagate in solid media: copper, fiber, coax unguided media: signals propagate freely, e.g., radio Introduction 1-62

Physical Media: coax, fiber Fiber optic cable: glass fiber carrying light pulses, each pulse a bit high-speed operation: high-speed point-to-point transmission (e.g., 10’s-100’s Gpbs) low error rate: repeaters spaced far apart ; immune to electromagnetic noise Coaxial cable: two concentric copper conductors bidirectional baseband: single channel on cable legacy Ethernet broadband: multiple channels on cable HFC Introduction 1-63

Physical media: radio Radio link types: terrestrial microwave e.g. up to 45 Mbps channels LAN (e.g., WiFi) 11Mbps, 54 Mbps wide-area (e.g., cellular) 3G cellular: ~ 1 Mbps satellite Kbps to 45Mbps channel (or multiple smaller channels) 280 msec end-end delay geosynchronous versus low altitude (Low-Earth-Orbiting satellites) – future Internet access signal carried in electromagnetic spectrum no physical “wire” bidirectional propagation environment effects: reflection obstruction by objects interference Introduction 1-64

Question ? An image of 1024x768 pixels with 3 bytes/pixel. Assume the image is uncompressed. How long does it take to transmit over a 56-kbps modem channel? Over a 1-Mbps cable modem? Over a 10-Mbps Ethernet? Over 100-Mbps Ethernet? Clue: Mega = 1 x 106 Kilo = 1 x 103 1024 768

Question: transmission delay Answer How long does it take to transmit over a 56-kbps modem channel? Over a 1-Mbps cable modem? Over a 10-Mbps Ethernet? Over 100-Mbps Ethernet? 1024 768 SOLUTION: The image is 1024 x 768 x 3 bytes or 2,359,296 bytes. This is 18,874,368 bits. At 56,000 bits/sec., it takes about 337.042 sec. At 1,000,000 bits/sec, it takes about 18.874 sec. At 10,000,000 bits/sec., it takes about 1.887 sec. At 100,000,000 bits/sec., it takes about 0.189 sec.

How do loss and delay occur? packets queue in router buffers packet arrival rate to link exceeds output link capacity packets queue, wait for turn packet being transmitted (delay) A free (available) buffers: arriving packets dropped (loss) if no free buffers packets queueing (delay) B Introduction 1-67

Loss in Packet-Switched Networks - Length of Queue is finite Queue Packets are lost when queue is full Incoming packet is dropped Queue packet in queue is dropped Lost packet - Retransmitted by application or transport layer protocol

Four sources of packet delay B propagation transmission nodal processing queueing dnodal = dproc + dqueue + dtrans + dprop dproc: nodal processing check bit errors determine output link typically < msec dqueue: queueing delay time waiting at output link for transmission depends on congestion level of router Introduction 1-69

Four sources of packet delay B propagation transmission nodal processing queueing dnodal = dproc + dqueue + dtrans + dprop dtrans: transmission delay: L: packet length (bits) R: link bandwidth (bps) dtrans = L/R dprop: propagation delay: d: length of physical link s: propagation speed of medium (~2x108 m/sec) dprop = d/s dtrans and dprop very different Introduction 1-70 70

Queueing delay (revisited) R=link bandwidth (bits/sec) L=packet length (bits) a=average packet arrival rate (packets/sec) traffic intensity = La/R La/R ~ 0: average queueing delay small La/R -> 1: delays become large La/R > 1: more “work” arriving than can be serviced, average delay infinite! There is a formula that helps us gauge the traffic intensity, in order for us to determine whether the average queuing delay is tolerable or not. The graph depicts an exponential curve telling us that as the value of traffic intensity approaches 1, queuing delay becomes large. If we will examine the formula, as the average packet arrival rate picks up, and multiply that by the Packet length, the link’s bandwidth R may not be able to cope up in transmitting the packets. This estimates the extent of queuing delay. Design your system so that traffic intensity is not greater than 1. Let’s look at a demo!

Throughput throughput: rate (bits/time unit) at which bits transferred between sender/receiver instantaneous: rate at given point in time average: rate over longer period of time Average rate of successful message delivery over a communications channel link capacity Rs bits/sec link capacity Rc bits/sec server sends bits (fluid) into pipe server, with file of F bits to send to client pipe that can carry fluid at rate (Rs bits/sec) pipe that can carry fluid at rate (Rc bits/sec) Introduction 1-72

Throughput (more) Rs < Rc What is average end-end throughput? Rc bits/sec Rs bits/sec Rs > Rc What is average end-end throughput? Rs bits/sec Rc bits/sec link on end-end path that constrains end-end throughput bottleneck link Introduction 1-73

Throughput: Internet scenario per-connection end-end throughput: min(Rc,Rs,R/10) in practice: Rc or Rs is often bottleneck Rs Rs Rs R Rc Rc Rc e.g. 10 clients downloading with 10 servers Rc=1 Mbps, Rs=2Mbps, R=5Mbps 10 connections (fairly) share backbone bottleneck link R bits/sec Introduction 1-74

Delay and Routes in the Internet TraceRoute(diagnostic program) -defined in RFC 1393 SOURCE HOST DESTINATION HOST Program Program SOURCE: records time elapsed (time received- time packet sent) determines the round-trip delays to all intervening routers Round-trip delays include: Router processing delay Queuing delay (varies with time) Transmission delay Propagation delay records name & address of router (or destination HOST) that returns the message reconstructs the route taken by the packets (source-to-destination) If there are (N-1) routers, then SOURCE sends N special packets Each packet is addressed to the ultimate destination marked 1 to N For tracking down the route taken by packets, we can use a program like TraceRoute provided by www.TraceRoute.org Traceroute sends a sequence of Internet Control Message Protocol (ICMP) packets addressed to a destination host. Tracing the intermediate routers traversed involves control of the time-to-live (TTL) Internet Protocol parameter. Routers decrement this parameter and discard a packet when the TTL value has reached zero, returning an ICMP error message (ICMP Time Exceeded) to the sender. Traceroute works by increasing the TTL value of each successive batch of packets sent. The first three packets sent have a time-to-live (TTL) value of one, expecting that they are not forwarded by the first router. The next three packets have a TTL value of 2, so that the second router will send the error reply. This continues until the destination host receives the packets and returns an ICMP Echo Reply message. The traceroute utility uses the returning ICMP messages to produce a list of hosts that the packets have traversed in transit to the destination. The three timestamp values returned for each host along the path are the delay (aka latency) values, typically measured in milliseconds for each packet in the batch. When DESTINATION host receives the Nth packet: DESTINATION destroys the packet, then returns the message back to the source When the ith router receives the ith packet marked i: router destroys the packet Sends a message containing name and address of router back to the source

www.TraceRoute.org Trace:3x * - indicates packet loss Route trace: From MIT to Massey University Three delay measurements Trace Route from MIT IMPORTANT: This tool works by sending a series of UDP packets with different port numbers and TTL (Time To Live). If you are running firewall software, your software may interpret the incoming packets as a hostile "port scan" originating from this server (jis.mit.edu). Rest assured, your system is not being attacked. 1 W92-RTR-1-W92SRV21.MIT.EDU (18.7.21.1) 0.425 ms 0.287 ms 0.259 ms 2 EXTERNAL-RTR-1-BACKBONE.MIT.EDU (18.168.0.18) 21.179 ms 244.069 ms 223.625 ms 3 leg-208-30-223-5-CHE.sprinthome.com (208.30.223.5) 0.589 ms 0.459 ms 0.542 ms 4 144.232.21.50 (144.232.21.50) 2.951 ms 3.146 ms 2.966 ms 5 sl-bb21-chi-6-2.sprintlink.net (144.232.19.205) 21.073 ms 48.427 ms 20.784 ms 6 sl-bb24-chi-9-0.sprintlink.net (144.232.26.77) 141.917 ms 229.305 ms 219.150 ms 7 sl-bb21-sj-8-0.sprintlink.net (144.232.20.161) 68.260 ms 68.102 ms 68.044 ms 8 sl-bb22-sj-15-0.sprintlink.net (144.232.3.162) 68.016 ms 68.036 ms 68.608 ms 9 144.232.20.47 (144.232.20.47) 73.346 ms 73.617 ms 73.508 ms 10 sl-newzeal-1-0.sprintlink.net (144.223.243.18) 70.804 ms 71.082 ms 70.787 ms 11 p5-2.sjbr1.global-gateway.net.nz (203.96.120.213) 71.132 ms 70.990 ms 70.903 ms 12 203.96.120.118 (203.96.120.118) 195.054 ms 195.579 ms 196.648 ms 13 203.96.120.201 (203.96.120.201) 198.228 ms 211.397 ms 197.358 ms 14 massey-uni-ak-int.tkbr4.global-gateway.net.nz (202.49.163.230) 202.604 ms 218.925 ms 199.836 ms 15 * * * 16 * * * 17 * * * 18 * * * 19 * * * 20 * * * 21 * * * 23 * * * 22 * * * 24 * * * 25 * * * 26 * * * 28 * * * 27 * * * 29 * * * 30 * * * Trans-oceanic link * means no response (probe lost, router not replying) 6 columns: n, name of router, address of router, trip delay1,trip delay2,trip delay3 * - indicates packet loss

The round-trip delay decreased between the two routers! www.TraceRoute.org Route trace: From MIT to Massey University Trace Route from MIT IMPORTANT: This tool works by sending a series of UDP packets with different port numbers and TTL (Time To Live). If you are running firewall software, your software may interpret the incoming packets as a hostile "port scan" originating from this server (jis.mit.edu). Rest assured, your system is not being attacked. 1 W92-RTR-1-W92SRV21.MIT.EDU (18.7.21.1) 0.425 ms 0.287 ms 0.259 ms 2 EXTERNAL-RTR-1-BACKBONE.MIT.EDU (18.168.0.18) 21.179 ms 244.069 ms 223.625 ms 3 leg-208-30-223-5-CHE.sprinthome.com (208.30.223.5) 0.589 ms 0.459 ms 0.542 ms 4 144.232.21.50 (144.232.21.50) 2.951 ms 3.146 ms 2.966 ms 5 sl-bb21-chi-6-2.sprintlink.net (144.232.19.205) 21.073 ms 48.427 ms 20.784 ms 6 sl-bb24-chi-9-0.sprintlink.net (144.232.26.77) 141.917 ms 229.305 ms 219.150 ms 7 sl-bb21-sj-8-0.sprintlink.net (144.232.20.161) 68.260 ms 68.102 ms 68.044 ms 8 sl-bb22-sj-15-0.sprintlink.net (144.232.3.162) 68.016 ms 68.036 ms 68.608 ms 9 144.232.20.47 (144.232.20.47) 73.346 ms 73.617 ms 73.508 ms 10 sl-newzeal-1-0.sprintlink.net (144.223.243.18) 70.804 ms 71.082 ms 70.787 ms 11 p5-2.sjbr1.global-gateway.net.nz (203.96.120.213) 71.132 ms 70.990 ms 70.903 ms 12 203.96.120.118 (203.96.120.118) 195.054 ms 195.579 ms 196.648 ms 13 203.96.120.201 (203.96.120.201) 198.228 ms 211.397 ms 197.358 ms 14 massey-uni-ak-int.tkbr4.global-gateway.net.nz (202.49.163.230) 202.604 ms 218.925 ms 199.836 ms 15 * * * 16 * * * 17 * * * 18 * * * 19 * * * 20 * * * 21 * * * 23 * * * 22 * * * 24 * * * 25 * * * 26 * * * 28 * * * 27 * * * 29 * * * 30 * * * The round-trip delay decreased between the two routers! Can you explain why the delays sometimes decrease from one router to the next? 6 columns: n, name of router, address of router, trip delay1,trip delay2,trip delay3 * - indicates packet loss

Tracert (from xtra to mit) C:\>tracert web.mit.edu Tracing route to web.mit.edu [18.7.22.69] over a maximum of 30 hops: 1 1 ms 1 ms 1 ms 192.168.1.1 2 2 ms 2 ms 2 ms 192.168.8.1 3 56 ms 59 ms 55 ms 219-89-32-1.dialup.xtra.co.nz [219.89.32.1] 4 * 53 ms 54 ms 222.152.127.169 5 * 66 ms * 202.50.236.105 6 * * * Request timed out. 7 482 ms * * so-0-2-0.labr3.global-gateway.net.nz [202.50.232.26] 8 * * * Request timed out. 9 * 341 ms 290 ms g11-2-107.core01.lax05.atlas.cogentco.com [154.54.11.145] 10 243 ms 213 ms * t3-4.mpd01.lax01.atlas.cogentco.com [154.54.6.189] 11 217 ms 280 ms * g9-0-0.core01.lax01.atlas.cogentco.com [154.54.2.117] 12 * 344 ms 325 ms p2-0.core01.dfw01.atlas.cogentco.com [154.54.5.93] 13 * * 282 ms p15-0.core02.dfw01.atlas.cogentco.com [66.28.4.26] 14 250 ms * * p15-0.core01.mci01.atlas.cogentco.com [66.28.4.38] 15 * * * Request timed out. 16 * 367 ms * p15-0.core01.ord01.atlas.cogentco.com [66.28.4.61] 17 * 386 ms 434 ms p14-0.core01.alb02.atlas.cogentco.com [154.54.1.57] 18 * 345 ms 448 ms p6-0.core01.bos01.atlas.cogentco.com [154.54.7.42] 19 * 282 ms 285 ms g8.ba21.b002250-1.bos01.atlas.cogentco.com [66.250.14.210] 20 * * 408 ms MIT.demarc.cogentco.com [38.112.2.214] 21 342 ms * * W92-RTR-1-BACKBONE.MIT.EDU [18.168.0.25] 22 * 344 ms * WEB.MIT.EDU [18.7.22.69] 23 * 342 ms 380 ms WEB.MIT.EDU [18.7.22.69] Trace complete. C:\> Tracert (also known as traceroute) is a Windows based tool that allows you to help test your network infrastructure.

Tracert (from Massey to MIT) D:\Massey Papers\159334\Codes\Game Protocol v3.6>tracert web.mit.edu Tracing route to web.mit.edu [18.7.22.69] over a maximum of 30 hops: 1 <1 ms <1 ms <1 ms it023453-vlan205.massey.ac.nz [130.123.246.129] 2 <1 ms <1 ms <1 ms it028100-vlan801.massey.ac.nz [10.100.254.3] 3 1 ms <1 ms <1 ms 210.7.32.1 4 1 ms <1 ms <1 ms 210.7.36.67 5 142 ms 142 ms 142 ms 210.7.47.22 6 142 ms 142 ms 142 ms abilene-1-lo-jmb-706.sttlwa.pacificwave.net [207.231.240.8] 7 179 ms 187 ms 180 ms dnvrng-sttlng.abilene.ucaid.edu [198.32.8.50] 8 189 ms 189 ms 202 ms kscyng-dnvrng.abilene.ucaid.edu [198.32.8.14] 9 201 ms 214 ms 201 ms iplsng-kscyng.abilene.ucaid.edu [198.32.8.80] 10 202 ms 215 ms 202 ms chinng-iplsng.abilene.ucaid.edu [198.32.8.76] 11 202 ms 206 ms 207 ms ge-0-0-0.10.rtr.chic.net.internet2.edu [64.57.28.1] 12 219 ms 230 ms 220 ms so-3-0-0.0.rtr.wash.net.internet2.edu [64.57.28.13] 13 225 ms 224 ms 224 ms ge-1-0-0.418.rtr.chic.net.internet2.edu [64.57.28.10] 14 229 ms 229 ms 229 ms nox300gw1-Vl-110-NoX-ABILENE.nox.org [192.5.89.221] 15 229 ms 229 ms 229 ms nox230gw1-Vl-802-NoX.nox.org [192.5.89.254] 16 481 ms 230 ms 230 ms nox230gw1-PEER-NoX-MIT-192-5-89-90.nox.org [192.5.89.90] 17 230 ms 230 ms 230 ms W92-RTR-1-BACKBONE.MIT.EDU [18.168.0.25] 18 230 ms 230 ms 230 ms WEB.MIT.EDU [18.7.22.69] Trace complete. D:\Massey Papers\159334\Codes\Game Protocol v3.6>

Roadmap 1.1 What is the Internet? 1.2 Network edge 1.3 Network core end systems, access networks, links 1.3 Network core circuit switching, packet switching, network structure 1.4 Delay, loss and throughput in packet-switched networks 1.5 Protocol layers, service models Introduction 1-80

Protocol “Layers” Question: Networks are complex, with many “pieces”: hosts routers links of various media applications protocols hardware, software Question: Is there any hope of organizing structure of network? Or at least our discussion of networks? Introduction 1-81

An analogy: Organization of air travel ticket (purchase) baggage (check) gates (load) runway takeoff airplane routing ticket (complain) baggage (claim) gates (unload) runway landing Ticketed passengers Baggage-checked, Ticketed passengers Baggage-checked, Ticketed, passed through the gate passengers Passenger in-flight a series of steps Introduction 1-82

Layering of airline functionality ticket (purchase) baggage (check) gates (load) runway (takeoff) airplane routing departure airport arrival intermediate air-traffic control centers ticket (complain) baggage (claim gates (unload) runway (land) ticket baggage gate takeoff/landing Layers: each layer implements a service via its own internal-layer actions relying on services provided by layer below Introduction 1-83

Why layering? Dealing with complex systems: explicit structure allows identification, relationship of complex system’s pieces layered reference model for discussion modularization eases maintenance, updating of system change of implementation of layer’s service transparent to rest of system e.g., In the air travel analogy, a change in gate procedure doesn’t affect rest of system layering considered harmful? Introduction 1-84

Tasks of Layers Each layer may perform one or more of the following tasks: Error Control Flow control Segmentation and Reassembly Multiplexing Connection Set-up Potential Drawbacks of Layering: Duplication of services Possible violation of layer dependency (conflicting information dependency among layers)

Communication in a Layered Architecture Concept of Protocol Layering Let’s consider 2 Network Entities (e.g. End Systems, Packet Switches) Sending side Receiving side Layer 4 M M Layer 3 H3 M1 H3 M2 H3 M1 H3 M2 Layer 2 H2 H3 M1 H2 H3 M2 H2 H3 M1 H2 H3 M2 Layer 1 H1 H2 H3 M1 H1 H2 H3 M2 H1 H2 H3 M1 H1 H2 H3 M2 To get a good grasp of how network entities communicate in a layered architecture, let’s consider 2 network entities, which may represent either end systems or packet switches. Let’s also assume that there are 4 layers in each network entity, and that the communication between each layer is made possible by passing layer-n messages called n-PDUs. When the SOURCE HOST creates a message M (as defined by its application) at the highest layer (Layer 4), aimed to be delivered to the DESTINATION HOST, the message M is first divided into two parts by the next lower layer (Layer 3); that is, M becomes M1 and M2. Next, Layer 3 in the SOURCE HOST appends the H3 header to M1 and M2 to indicate additional information needed by the sending and receiving sides of Layer 3 to provide the services for Layer 4. A similar process is taken by the succeeding lower layers (appending necessary headers), until the messages are sent out of the SOURCE and onto a physical link. At the receiving end system, the messages are directed up the protocol stack. During each climb on the layers, the corresponding header for that layer is removed. Eventually, M is reassembled from M1 and M2 and then passed on to the application. SOURCE DESTINATION What happens when the SOURCE wants to send a message to the DESTINATION? Comprised of 4 Layers; where each layer n is governed by a protocol. Layers communicate by exchanging layer-n messages called (n-PDUs) Protocol Data Units. The contents, format, and procedure for exchanging PDUs are defined by Layer-n Protocol

Encapsulation destination source application transport network link message M application transport network link physical segment Ht M Ht datagram Ht Hn M Hn frame Ht Hn Hl M link physical switch destination network link physical Ht Hn M Ht Hn Hl M M application transport network link physical Ht Hn M Ht M Ht Hn M router Ht Hn Hl M Introduction 1-88

Internet protocol stack application: supporting network applications FTP, SMTP, HTTP transport: process-process data transfer TCP, UDP network: routing of datagrams from source to destination IP, routing protocols link: data transfer between neighboring network elements Ethernet, 802.11 (WiFi), PPP physical: bits “on the wire” application Transport network link physical Introduction 1-89

Internet protocol stack application: supporting network applications ftp, smtp, http transport: host-host data transfer tcp, udp network: routing of datagrams from source to destination ip, routing protocols (Hardware+Software) link: data transfer between neighbouring network elements ppp, ethernet physical: bits “on the wire” application transport network link physical Mostly software implemented Guaranteed delivery of application layer messages Defines fields in IP datagrams (destination address), how end systems and routers act on them Moves packets from one node (host or packet switch) to the next node Ethernet & ATM cards implement both link and Physical Layers Move individual bits within frame from one node to the next

Introduction 1-91

Tier-1 ISP: e.g., Sprint … …. to/from backbone peering to/from customers peering to/from backbone …. POP: point-of-presence Introduction 1-93

Internet structure: network of networks roughly hierarchical at center: small # of well-connected large networks “tier-1” commercial ISPs (e.g., Verizon, Sprint, AT&T, Qwest, Level3), national & international coverage large content distributors (Google, Akamai, Microsoft) treat each other as equals (no charges) IXP Tier 1 ISP Tier-1 ISPs & Content Distributors, interconnect (peer) privately Large Content Distributor (e.g., Google) Large Content Distributor (e.g., Akamai) Tier 1 ISP Tier 1 ISP … or at Internet Exchange Points IXPs http://www.akamai.com/html/industry/index.html Introduction 1-94

Chapter 1 – True or False Questions Exercises Chapter 1 – True or False Questions

Exercises We are sending a 30 Mbit MP3 file from a source host to a destination host. All links in the path between source and destination have a transmission rate of 10 Mbps. Assume that the propagation speed is 2 * 108 meters/sec, and the distance between source and destination is 10,000 km. 1 . Initially suppose there is only one link between source and destination. Also suppose that message switching is used, with the message consisting of the entire MP3 file. The transmission delay 3 seconds 3.05 seconds 50 milliseconds none of the above. 3 SEC

Exercises We are sending a 30 Mbit MP3 file from a source host to a destination host. All links in the path between source and destination have a transmission rate of 10 Mbps. Assume that the propagation speed is 2 * 108 meters/sec, and the distance between source and destination is 10,000 km. 2 . Referring to the above question, the end-to-end delay (transmission delay plus propagation delay) is 3.05 seconds 3 seconds 6 seconds none of the above A) 3.05

Exercises We are sending a 30 Mbit MP3 file from a source host to a destination host. All links in the path between source and destination have a transmission rate of 10 Mbps. Assume that the propagation speed is 2 * 108 meters/sec, and the distance between source and destination is 10,000 km. 3 . Referring to the above question, how many bits will the source have transmitted when the first bit arrives at the destination. 1 bit 30,000,000 bits 500,000 bits none of the above C) 500,000 BITS

Exercises We are sending a 30 Mbit MP3 file from a source host to a destination host. All links in the path between source and destination have a transmission rate of 10 Mbps. Assume that the propagation speed is 2 * 108 meters/sec, and the distance between source and destination is 10,000 km. 4 . Now suppose there are two links between source and destination, with one router connecting the two links. Each link is 5,000 km long. Again suppose the MP3 file is sent as one message. Suppose there is no congestion, so that the message is transmitted onto the second link as soon as the router receives the entire message. The end-to-end delay is 3.05 seconds 6.1 seconds 6.05 seconds none of the above C) 6.05 seconds

End of Session 