1. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.1 Computer Networks and Internets with Internet Applications, 4e By Douglas E. Comer Lecture PowerPoints By Lami Kaya, LKaya@ieee.org

2. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.2 Chapter 15 Network Characteristics Ownership, Service Paradigm, And Performance

3. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.3 15.1 Introduction 15.2 Network Ownership –15.2.1 Private Networks –15.2.2 Public Networks 15.3 Privacy And Public Networks 15.4 Advantages And Disadvantages 15.5 Virtual Private Networks 15.6 Guaranteeing Absolute Privacy 15.7 Service Paradigm Topics Covered

4. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.4 15.8 Connection-Oriented Service Paradigm –15.8.1 Continuous And Bursty Traffic –15.8.2 Simplex And Full Duplex Connections –15.8.3 Connection Duration And Persistence –15.8.4 Service Guarantees –15.8.5 Stream Or Message Interface 15.9 Connectionless Service Paradigm 15.10 Interior And Exterior Service Paradigms 15.11 Comparison Of Service Paradigms 15.12 Examples Of Service Paradigms Topics Covered (cont)

5. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.5 Topics Covered (cont) 15.13 Addresses And Connection Identifiers 15.14 NW Performance Characteristics –15.14.1 Delay –15.14.2 Throughput –15.14.3 The Relationship Between Delay And Throughput –15.14.4 Delay-Throughput Product 15.15 Jitter –15.15.1 Isochronous Networks –15.15.2 Asynchronous Networks

6. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.6 15.1 Introduction Each technology is classified into one of three categories: LAN, MAN, or WAN, Three additional characteristics of networks: –network ownership, –the type of service (both the service that the network provides to attached computers and the service it uses internally), –and the performance of the resulting system. These characteristics permit more accurate comparisons among technologies –because they provide detail about similarities and differences

7. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.7 15.2 Network Ownership NW HW and SW can be owned by a company or individual A NW owned and used by a single company or an individual is –said to be “private network” A NW owned by common carriers are –called “public network”

8. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.8 15.2.1 Private Networks (1) LAN technologies comprise the most common form of private network. To run a private network, a corporation –hires employees who create and operate the network –The necessary HW and software are purchased outright –And employees install the wiring, connect computers –and manage the resulting system A private corporation can install cables only on property that the corporation owns.

9. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.9 15.2.1 Private Networks (2) A large corporation may also use private WAN technologies to connect computers at multiple sites: –The corporation purchases WAN HW such as packet switches –Hires employees to operate the NW –The employees design network interconnections –Attach computers to NW, assign addresses, and control routing To form a private WAN, a corporation must lease connections between its sites from public carriers –Such WAN is still considered private because the leased connections carry data directly between the corporation's sites –No other corporations have access to the wires or the data

10. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.10 15.2.2 Public Networks A public network is analogous to a telephone system – NW is run as a service available to subscribers Any one can subscribe to the service A feature of a public NW is universal communication –A given subscriber's computer can communicate with any other A public NW that is available to many subscribers in many locations is more attractive than one that only serves a small geographic area

11. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.11 15.3 Privacy And Public Networks The term “public” refers to availability of the service –not to the data transferred Public networks provide private communication Some public networks permit a group of computers to communicate –analogous to a telephone conference call Public network does not use broadcast technology

12. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.12 15.4 Advantages And Disadvantages Advantage of a private NW is that the owner has complete control: –Chooses HW to use, the capacity, redundancy/backup –Sets policies –Can guarantee that NW is isolated from computers outside Isolation helps enforce security Disadvantages: –A large private NW can be expensive to install and maintain. –Purchasing the NW HW –Hire/train a staff to plan/install/manage/operate –Special tools for installation/maintenance –Keeping up with rapid change

13. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.13 15.4 Advantages And Disadvantages (cont) The chief advantages of a public NW –are flexibility and the ability to use state-of-the-art networking without maintaining technical expertise –A subscriber at an arbitrary location can connect to the NW at any time Connections between a host owned by one organization and a computer owned by another can be made

14. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.14 15.5 Virtual Private Networks (VPN) VPN technology allows a company with multiple sites to have a private NW –but use a public NW as a carrier VPN technology restricts traffic –so that packets can travel only between the company's sites If an outsider accidentally receives a copy of a packet – VPN technology ensures that they cannot understand the contents

15. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.15 Building a VPN Special HW and SW system for each site The system is placed between private/public NW. Each VPN system configured with the addresses of the VPN systems at other sites Additional details can be found in Chapter 40

16. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.16 Operation of a VPN Once VPN configured, the company's sites can only communicate with one another –they are cut off from the rest of the NW. The VPN system restricts packets: –VPN system at each site restricts incoming/outgoing packets –No packet can leave the site unless it is traveling to another one of the company's sites When two computers belong to same VPN exchange messages, routing takes place: –VPN sends the packet across the public NW to the destination –When the packet arrives the receiving VPN system verifies that it came from a valid peer site and then forwards the packet to its destination

17. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.17 15.6 Guaranteeing Absolute Privacy Some corporations are unwilling to send sensitive data across a public NW Absolute privacy –preserving individual anonymity, or –preserving the confidentiality of data being sent VPN system encrypts each packet before sending the packet across the public NW –An outsider cannot understand the contents of packets The receiving VPN system decrypts each incoming packet before sending it on to the destination computer –packet remains encrypted during its trip across the public NW

18. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.18 15.7 Service Paradigm NW systems offer a variety of services The goal is to provide a higher-level interface –Allows the computer to specify a remote destination, and –Transfer data without worrying about packets. Exact details of interface mechanisms may vary General type of interface is known as –an interface paradigm or –a service paradigm NWs are placed in one of two broad categories: –Connection-oriented (CO) service –Connectionless (CL) service

19. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.19 15.8 Connection-Oriented Service Paradigm Analogous to telephone communication: –A connection must be established between two computers –After connection data can be sent across –The connection must be terminated CO NW is an abstract idea –Circuit switching is a specific mechanism that provides a CO interface The term CO is generic –it applies broadly to a class of technologies –The class encompasses many technologies –The designs and details differ

20. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.20 15.8.1 Continuous And Bursty Traffic CO NW designed to handle voice or video –to accept and deliver continuous data at a fixed rate Some other CO NW are designed to handle burst traffic –A computer can send data for a while, –stop sending data –and then resume sending The connection does not disappear because no data is being sent

21. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.21 15.8.2 Simplex And Full Duplex Connections Some CO technologies provide full duplex connections (two-way) Some CO technologies provide simplex (one-way) To communicate using a simplex design, a pair of computers must establish two connections: –one from computer A to computer B and –another from computer B to computer A

22. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.22 15.8.3 Connection Duration And Persistence Some CO NWs are designed to use permanent connections that persist over months or years Some CO technologies permit switched connections –that can be established or terminated quickly and automatically

23. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.23 15.8.4 Service Guarantees Some CO NWs provide guarantees about the service that computers will receive. For example, –Guarantee a throughput rate or –Maximum packet loss rate. Other connection-oriented technologies do not provide guarantees. For example, –ATM provides statistical guarantees about performance, but does not absolutely guarantee delivery (i.e., cells can be lost).

24. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.24 15.8.5 Stream Or Message Interface Some CO NWs provide a stream interface –Once the connection is open, the computer can send a stream of data octets that are delivered to the other end –With a stream interface, no boundaries are recorded Receiver may receive a single block of 60 characters even though the sender generates three blocks of 20 characters. Other CO technologies provide a message interface –in which the NW guarantees to deliver data in the same size chunks that the sender transmitted

25. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.25 15.9 Connectionless (CL) Service Paradigm CL NWs operate analogous to the postal mail system Whenever it has data to send, a computer must –place the data in the appropriate frame format, –attach the address of receiver and –then pass the frame to the NW for delivery CL system transports the frame to the destination Details differ among connectionless technologies –Addressing scheme; Length address Method of assigning address –Imposing an upper bound on the size of a frame –Impose a minimum packet size

26. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.26 15.10 Interior And Exterior Service Paradigms A NW that provides one service paradigm to attached computers can use an entirely different service paradigm internally –For example, although it provided connectionless service to attached computers, the ARPANET used a connection-oriented paradigm internally We will see another example of mixed paradigms when we examine the TCP/IP protocols.

27. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.27 15.11 Comparison Of Service Paradigms Each service type has advantages and disadvantages. CO paradigm has ease of accounting and the ability to inform communicating computers immediately when a connection breaks Public NW that charge customers for NW use favor CO –because less effort is required to charge for time Learning about NW failure immediately can help applications that are using the NW. A failure in a CL may go unnoticed and unreported CL service paradigm has less initial overhead – a CL NW allows a computer to send data immediately, without waiting for a connection

28. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.28 15.12 Examples Of Service Paradigms LAN technologies such as Ethernet use CL; –Although a computer needs to wait for access to a shared medium before sending a packet the computer does not need to establish a connection WANs can use CO and CL service paradigms –Frame Relay uses CO –ATM switched virtual circuit PVC ( SVCs ) and permanent virtual circuit (PVC) –Switched Multi-megabit Data Service (SMDS) is also used in public WANs and offers CL Most LAN technologies are CL, but some LANs take a CO approach –ATM was designed for both WANs and LANs and uses CO Figure 15.1 summarizes the service paradigms offered

29. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.29

30. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.30 15.13 Addresses And Connection Identifiers In a CL, each packet must contain the address of the recipient. CO service often uses abbreviations for connections For a connection, –Computer sends a message to the NW specifying the address of destination –After connection, the NW responds with a message that verifies the connection and specifies a connection identifier Usually, a connection identifier is a small integer, much shorter than the full destination address –When sending/receiving data the computer uses the connection identifier instead of a destination address –Using connection identifiers reduces overhead because it makes the header of data packets smaller

31. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.31 15.14 NW Performance Characteristics NWs can be classified as low speed or high speed Such definitions are inadequate –because NW technologies change rapidly Scientists or engineers need to specify NW speeds precisely, –They do not use informal, qualitative terms –Instead, they use quantitative metrics Quantitative measures are important –because they make it possible to compare any two NWs

32. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.32 15.14.1 Delay The delay of a NW specifies how long it takes for a bit of data to travel across the NW Delays may differ slightly –depending on the location of the specific pair of computers that communicate Users only care about the total delay of a NW Engineers need to make more precise measurements –The maximum –The average delay

33. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.33 Delay Types Propagation Delay –A signal requires a some time to travel, which is proportional to the distance Switching Delay –an electronic device waits until all bits of a packet have arrived –and then takes a small amount of time to choose the next hop before sending a packet Access Delay –Most LANs use shared media, computers must delay until the medium is available Queueing Delay –Each packet switch enqueues incoming packets as part of the store-and-forward process If the queue already contains packets, the new packet may wait

34. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.34 15.14.2 Throughput Throughput is a measure of the rate at which data can be sent through the NW, in bits per second (bps) Throughput capability of the underlying HW is called BW –sometimes BW = throughput. –Programmers and users do not care about the capability of the underlying HW, they are interested in the actual data rate. A frame contains a header, which means that the effective throughput is less than the HW bandwidth. NWing professionals often use the term speed = throughput. –This can be confusing because delay and throughput are separate –Throughput is a measure of capacity, not speed. imagine a NW to be a road between two places and packets traveling across the NW to be analogous to cars traveling down the road. NW throughput specifies how many bits can enter NW per unit time

35. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.35 15.14.3 The Relationship Between Delay And Throughput In theory, the delay and throughput of a NW are independent. –In practice, however, they can be related –Think of the road analogy If a PS has a queue of packets waiting when a new packet arrives –the new packet will be placed on the tail of the queue –and will need to wait while the switch forwards the previous ones Excessive traffic in a NW is called congestion Data entering a congested NW will experience longer delays than data entering an idle NW Throughput and delay are not completely independent.

36. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.36 Estimating Delay The expected delay can be estimated from the current percentage of the NW capacity being used, if –D 0 delay when the NW is idle –U, utilization ( 0  1) –D, effective delay When a NW is completely idle, U is zero, D = D 0 When a NW operates at 1 / 2 of its capacity D = 2D 0 As traffic approaches the NW capacity (U  1 ), the delay  ∞ As traffic increases, delays increase –a NW that operates at close to 100% of its throughput capacity experiences severe delay

37. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.37 15.14.4 Delay-Throughput Product When used as a measure of the underlying HW –Delay x Throughput (D x T) –Delay x Bandwidth (BW) The (D x T) measures the volume of data that can be present on NW –the total # of bits in transit at any time This product is important for any NW with long delay –Means a NW can generate a large volume of data before the destination receives the first bit.

38. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.38 15.15 Jitter NW jitter represent the variance in delay –used for transmission of real-time voice and video Consider sending voice over a NW. –On the sending side, the analog signal is digitized –The samples are collected into packets or cells then transferred across the NW –At the receiving side the digital values are extracted and converted back to analog output –If the NW has zero jitter (exactly the same time to transit the NW) the audio output will exactly match the original input –Otherwise, the output will be flawed Telephone system and data NW handle jitter differently

39. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.39 15.15.1 Isochronous Networks To ensure digitized telephone calls are transmitted and played back correctly –the telephone NW is designed so that all transmissions have exactly the same delay. –Ex: if digitized data from a phone call is transmitted over two paths HW is configured so that both paths have exactly the same delay Term “isochronous”, pronounced as ``eye-sock-re-nus'‘, –to characterize a NW that has a jitter of zero

40. © 2007 Pearson Education Inc., Upper Saddle River, NJ. All rights reserved.40 15.15.2 Asynchronous Networks Asynchronous is alternative to isochronous NWs is a NW in which delay among packets can vary Most current data NWs are asynchronous –but still they can used for audio and video transmission Although audio and video work best when jitter is low –additional protocols have been designed to ensure correct playback