Foundation for Wide Area Networks

1
Foundation for Wide Area Networks

• CS2520 - TELCOM2321
• Wide Area Networks
• KyoungSoo Park
• University of Pittsburgh
• Many slides borrowed from Jen Rexford Walter
  Cerroni

2
Course Objectives

• Understand the design principles of high-speed
  computer networks
• Understand the key concepts
• How they are designed
• How they work in practice
• How they are likely to evolve in the future
• Understand the issues with todays deployment
• Read and discuss classic recent papers
• Difference from undergrad course?
• Focus more on research ideas (less on programming
  skills)
• Treat the subjects in depth

3
Tentative Course Topics

• Network Architectures
• Packing Switching and Routing
• Flow and Congestion Control
• Network/Transport Layer Protocols
• Active Queue Management
• Quality of Service (IntServ, DiffServ)
• Multi-Protocol Label Switching
• Naming and DNS
• Content Distribution Networks (CDNs), P2P

4
Course Highlights

• Lectures
• Homework / Lab assignments
• Project
• Mid-term and final exams
• Prerequisites CS 1652 or other background on
  computer networking fundamentals
• Grading
• Homework/Lab 30
• Project 30
• Mid-term exam 20
• Final exam 20

5
Course Project

• Design and evaluate an interesting networking
  idea or
• Literature review on a specific topic
• You can work in group, if you like
• Important due dates
• March 3 short proposal (1-2 pages)
• Must discuss with me ahead of time
• April 14 final presentation (20 minutes)
  report(10 pages)
• Can use a research testbed for evaluation
• PlanetLab a real-world testbed with 800
  machines
• Emulab a dedicated emulation testbed

6
Textbook and references

• Larry L. Peterson and Bruce S. Davie, Computer
  Networks A Systems Approach, 4th Edition,
  Morgan Kaufmann, 2007
• Basic concept exercise problems
• Content knowledge is required for preliminary
  exams
• Research papers on specific topics will be
  suggested
• Additional references
• James F. Kurose and Keith W. Ross, Computer
  Networking A Top-Down Approach, 4th Edition,
  Addison Wesley, 2007.
• Zheng Wang, Internet QoS Architectures and
  Mechanisms for Quality of Service, Morgan
  Kaufmann, 2001.
• Richard Stevens, TCP/IP Illustrated, Volume
  1The Protocols , Addison Wesley, 1993

7
Other information

• Lecture (KyoungSoo Park)
• When Tue 600-850PM in 5313 SENSQ
• Slides available online at course Web site
• Office hours to be announced, and by appointmnet
• Teaching Assistant(Lei Jin)
• Email jinlei_at_cs.pitt.edu
• Course Web site http//www.cs.pitt.edu/kyoung
  soo/cs2520/
• 1st assignment
• Email me your picture URL (not the picture
  itself)
• Send me what you want to learn from the class if
  any

8
Todays Topic

• The Internet architecture overview
• Todays largest internetwork
• Design choices, strength and weakness
• Basic building blocks
• Multiplexing
• Switching
• Addressing
• Flow and error control
• Congestion control
• Performance measures
• Applications

9
The Internet The Good, The Bad and The Ugly
10
What is the Internet?
The Internet is the worldwide, publicly
accessible network of interconnected computer
networks that transmit data by packet switching
using the standard Internet Protocol (IP). It is
a "network of networks" that consists of millions
of smaller domestic, academic, business, and
government networks, which together carry various
information and services, such as electronic
mail, online chat, file transfer, and the
interlinked Web pages and other documents of the
World Wide Web.
http//en.wikipedia.org/wiki/Internet
11
The Internet A Remarkable Story

• Tremendous success
• A research experiment that trulyescaped from the
  lab
• The brilliance of under-specifying
• Best-effort packet-delivery service
• Key functionality at programmable end hosts
• Enabled massive growth and innovation
• Ease of adding hosts links, new technologies
• Ease of adding new services (Web, P2P, VoIP, )

12
Idea 1 Functionality at the edge of Network
13
Telephone Network Dumb Edge, Smart Core

• Dumb phones
• Dial a number
• Speak and listen
• Smart switches
• Set up and tear down a circuit
• Forward audio along the path
• Limited services
• Audio
• Later, fax, caller-id,
• A monopoly for a long time

14
Internet Smart Edge, Dumb Core
End-to-End Principle Whenever possible,
communications protocol operations should be
defined to occur at the end-points of a
communications system.
Programmability With programmable end hosts, new
network services can be added at any time, by
anyone.
And then end hosts became powerful and
ubiquitous.
15
Programmability

• Architectural decision with profound effects
• Where you place programmability in the system
  determines who gets to innovate
• And what kinds of innovations can happen
• Todays Internet
• Programmable hosts ? innovation in applications
• Non-programmable routers ? more control by
  standards bodies, routers vendors, and carriers

16
Idea 2 Best-effort Packet Switching
17
Internet Protocol (IP) Packet Switching

• Like the postal system
• Divide information into letters
• Stick them in envelopes
• Deliver them independently
• And sometimes they get there

• Whats in an IP?
• The data you want to send
• A header with the from and to addresses

18
Why Packets?

• Packets can be delivered by most anything
• Serial link, fiber optic link, coaxial cable,
  wireless, birds
• Data traffic is bursty
• Logging in to remote machines, exchanging e-mail
• Dont waste bandwidth
• No traffic exchanged during idle periods
• Better to allow multiplexing
• Different transfers share access to same links

19
Best-Effort Packet-Delivery Service

• Best-effort delivery
• Packets may be lost
• Packets may be corrupted
• Packets may be delivered out of order

source
destination
IP network
20
Why Best-Effort?

• Simpler network
• No error detection and correction
• Dont remember from one packet to next
• Dont reserve bandwidth and memory
• Transient disruptions are okay during failover
• but, applications do want efficient, accurate
  transfer of data in order, in a timely fashion
• Fortunately, the end host take care of that!

21
End Host Can Take Care of Requirements

• No error detection or correction
• Higher-level protocol can provide error checking
• Successive packets may not follow same path
• No problem as long as packets reach destination
• Packets can be delivered out-of-order
• Receiver can put packets back in order (if
  needed)
• Packets may be lost or arbitrarily delayed
• Sender can send the packets again (if desired)
• No reaction to congestion, beyond drop
• Sender can slow down in response to loss or delay

22
Idea 3 Layering and the IP Hourglass Model
23
Layering A Modular Approach

• Sub-divide the problem
• Each layer relies on services from layer below
• Each layer exports services to layer above
• Interface between layers defines interaction
• Hides implementation details
• Layers can change without disturbing other layers

Application
Application-to-application channels
Host-to-host connectivity
Link hardware
24
The Narrow Waist of IP
Applications
FTP
HTTP
TFTP
DNS
TCP
UDP
Waist
IP
Data Link
NET1
NET2
NETn

Physical
The Hourglass Model
The waist facilitates interoperability
25
Above and Below the Waist

• IP over anything
• Internetworking protocol that runs on anything
• Accommodate innovation in link technology
• and heterogeneity throughout the network
• Anything over IP
• Variety of transport protocols can be built
• Though, in practice, mainly just TCP and UDP
• TCP ordered, reliable stream of bytes
• UDP simple (unreliable) message delivery
• And any applications on top of that

26
End-to-End IP
host
host
HTTP message
HTTP
HTTP
TCP segment
TCP
TCP
router
router
IP packet
IP packet
IP packet
IP
Ethernet interface
Ethernet interface
SONET interface
Ethernet interface
SONET interface
Ethernet frame
SONET frame
Ethernet frame
27
Idea 4 Decentralized Control
28
Benefits of Decentralization Scalability

• Scalability
• Limit amount of state, and frequency of updates
• Addressing
• Internet routers only need to know how to reach
  blocks of addresses (e.g., 12.0.0.0/8)
• Routing
• Link failure in one network is typically not
  visible in another
• Naming
• Look-up of www.cnn.com doesnt go to same server
  as look-up of www.pitt.edu

29
Benefits of Decentralization Autonomy

• Autonomy
• Allow different parties to manage different parts
  of the system, and apply their own policies
• Addressing
• ARIN delegates address space to ATT, who
  delegates smaller blocks to its customers
• Routing
• ATT controls flow of traffic through its
  backbone
• Naming
• CNN controls addresses for www.cnn.com

30
Problems Lurking
31
Challenges Tied to Early Design Decisions

• Power of programmable end hosts
• Easy to spoof IP addresses, e-mail addresses,
• Incentives for users to violate congestion
  control
• Malicious users launching Denial-of-Service
  attacks
• Best-effort packet-delivery service
• Inefficient in high-loss environments (wireless)
• Poor performance for interactive applications
• Expensive per-packet handling on high-speed links

32
Challenges Tied to Early Design Decisions

• Layering and the IP narrow waist
• Low efficiency due to many layers of headers
• Poor visibility into underlying shared risks
• Complex network management due to multiple
  interconnected protocols and systems
• Decentralized control
• Hierarchical addressing makes mobility difficult,
  and requires careful configuration
• Autonomy makes measurement (and troubleshooting
  and accountability) hard
• Autonomy makes protocol changes difficult

33
Recurring Challenges

• Security
• Weak notions of identity that are easy to spoof
• Protocols that rely on good behavior
• Mobility and disconnected operation
• Hierarchical addressing closely tied with routing
• Presumption that hosts are connected
• Network management
• Many coupled, decentralized control loops
• Limited visibility into across layers and
  networks
• Application performance requirements
• Real-time, interactive applications
• Throughput sensitive vs. delay-sensitive

34
Internet is Not Standing Still

• Partial solutions to these problems
• Often as add ons or extensions
• Hampered by need to be backwards compatible, and
  work when only partially deployed
• Rather than complete architectural solutions
• Solutions create problems of their own
• Violations of architectural assumptions
• Unexpected interactions with applications
• Adding complexity to an already complex system

35
Example Middleboxes

• Middleboxes are intermediaries
• Interposed in-between the communicating hosts
• Often without knowledge of one or both parties
• Examples
• Network address translators
• Firewalls
• Traffic shapers
• Intrusion detection systems
• Transparent Web proxy caches
• Application accelerators

36
Middleboxes Address Practical Challenges

• Host mobility
• Relaying traffic to a host in motion
• IP address depletion
• Allowing multiple hosts to share a single address
• Security concerns
• Discarding suspicious or unwanted packets
• Detecting suspicious traffic
• Performance concerns
• Controlling how link bandwidth is allocated
• Storing popular content near the clients

37
Middleboxes Violate Network-Layer Principles

• Globally unique identifiers
• Each node has a unique, fixed IP address
• reachable from everyone and everywhere
• Simple packet forwarding
• Network nodes simply forward packets
• rather than modifying or filtering them

source
destination
IP network
38
Two Views of Middleboxes

• An abomination
• Violation of layering
• Cause confusion in reasoning about the network
• Responsible for many subtle bugs
• A practical necessity
• Solving real and pressing problems
• Needs that are not likely to go away
• Would they arise in any edge-empowered network,
  even if redesigned from scratch?

39
Clean-Slate Network Architecture

• Clean-slate architecture
• Without constraints of todays artifacts
• To have a stronger intellectual foundation
• And move beyond the incremental fixes
• Still, some constraints inevitably remain
• Ignore todays artifacts, but not necessarily all
  reality
• Such as
• Resource limitations (CPU, memory, bandwidth)
• Time delays between nodes
• Independent economic entities
• Malicious parties
• The need to evolve over time

40
Summary Of the Internet Architecture

• Internet architecture is a huge success
• Functionality at programmable edge nodes
• Best-effort packet-delivery service
• Layering and the IP hourglass model
• Decentralized control of the global system
• These very features are causing problems
• Security, mobility, manage-ability, performance,
  reliability,
• Rethinking the network architecture
• For a strong intellectual foundation
• And long-term improvements to the Internet

41
Basic Building Blocks
42
Multiplexing

• Cost-effective resource sharing
• Full-mesh network is not scalable n(n-1) links
• Sharing links among multiple inputs/outputs
• Statistical Multiplexing
• Send from any available input sources
• No reservation needed for link use
• Utilize the pipe as much as possible
• How to prevent starvation?
• Smaller pieces packets

43
Switching Forwarding

• Send data from a src port to a dest port
• Bridge, switch, router, etc.
• Forwarding vs. routing?
• Forwarding a process of finding the right dest
  switch port by looking up the addr in the
  forwarding table
• Routing a process of building the forwarding
  table
• Switching modes
• Circuit switching
• Packet switching
• Virtual circuit switching

44
Circuit Switching

• Exclusive dedication of a portion of the
  available bandwidth to carry traffic between src
  dest
• Bandwidth is allocated using
• Frequency Division Multiplexing (FDM)
• Time Division Multiplexing (TDM)
• Call requires three phases
• Connection phase,
• A circuit is set up between source and
  destination
• Transmission phase,
• Traffic exchange takes place
• Termination phase,
• The call is disconnected
• Legacy telephone network, a typical example

45
Circuit Switching

• Advantages
• Guranateed bandwidth
• Simple abstraction in-order delivery
• Fast forwarding no need to look at header
• Low per-packet overhead
• Disadvantages
• Wasted bandwidths
• Blocked connections when all resources are
  reserved
• Connection setup delay (RTT)
• Network state per-connection state

46
Packet Switching

• No transmission capacity is dedicated along the
  routing path (efficient bandwidth utilization)
• No call setup needed before data transmission
• Data is transmitted in a sequence of packets
• Store and forward
• Best-effort packet delivery
• Each packet is independent of each other
• Out-of-order delivery is possible
• No delivery guarantee
• No state preserved at the network

47
Virtual Circuit Packet Switching

• An initial phase is used to setup a fixed route
• Similar to circuit switching, except that a delay
  occurs at each node
• Call request and Call accept must both wait their
  turns for transmission
• Upon path set-up, the virtual circuit appears to
  the user as a dedicated circuit
• The circuit is actually shared among multiple
  users
• Destination address no longer required
• Only a virtual circuit number is needed
• Packets have shorter headers
• Faster forwarding
• Packets follow the same routing path
• no order disruption

48
Timing comparison
49
Addressing

• Providing suitable identifiers to nodes
• So you can direct data to a node (dest)
• So you know which node sent the data (src)
• and how to send back to that node
• Key concepts in addressing
• Number of unique addresses
• Flat vs. hierarchical structure
• Persistent vs. temporary identifiers
• Handling diminishing address space
• Spoofing of source addresses

50
Comparing MAC and IP Addresses
E.g., social security number vs. postal address
51
Growth in the Number of IP Prefixes
Internet bust
Internet boom
recovery?
CIDR
pre-CIDR
52
Flow Control Error Control

• Flow control
• To prevent the receiver from overrunning
• Link layer vs. transport layer
• Stop-and-wait and sliding window algorithm
• TCP sliding window receiver says I can receive
  this much
• Error control
• Guards against loss or damage of data
• Mostly error detection and retransmission
• CRC-32 for Ethernet
• Internet checksum for IP header, UDP TCP
  payload
• Should each layer independently implement its own?

53
Congestion Control

• Congestion control
• To prevent the network from overrunning
• Reduce the rate if you see a sign of congestion(
  pkt loss)
• TCP Additive Increase Multiplicative Decrease
  (AIMD)
• Explicit congestion notification (ECN)
• Router sets a bit in the IP header
• Issues with congestion control
• How to infer congestion?
• How to react to it?
• Who should enforce it?
• How to prevent malicious users?
• What are the goals?

54
Performance Measures

• Bandwidth vs. throughput
• Bandwidth diameter of a pipe
• Throughput measured bandwidth
• Delay
• Propagation delay
• Transmission delay
• Queuing delay
• Keep the pipe full
• Bandwidth-delay product
• Transfer time 1RTT data size/bandwidth
• How long it takes for the last byte to be ACKed
• Ignores queuing delay, errors and retransmission

55
Typical Applications
55
56
Applications

• Practical goal of computer networks
• Communication between applications on two or more
  end nodes
• Popular application protocols
• HTTP, SMTP, SSH, Telnet, FTP, etc.
• Multimedia application gets more popular
• Video conferencing
• Video on demand (VoD)
• IPTV TV broadcast over IP networks
• Peer-to-peer application
• The most popular in terms of traffic usage
• Net neutrality?

57
Multimedia application

• Bandwidth needed for (low-quality) video playback
• Resolution 352x240
• 3 bytes per pixel, 30 frames per second
• 352 x 240 x 3 x 30 (bytes/sec) 60.8 Mbps!
• Compression to the rescue
• Difference between two adjacent frames is small
• Send the base frame diff
• Quick reality check play 2 hr movie in real time
• DVD quality (MPEG-2) 4GB/7200 4.7 Mbps
• Blu-ray quality 25GB/7200 29.8 Mbps

58
System Requirements delay and jitter

• Video and Audio processing need to be bounded by
  deadlines, defined by a time interval
• Data transmission must adhere to the demands
  concerning time domains
• End-to-end jitter (delay variation) need to be
  bounded
• Large jitter values cause degradation of
  interactive applications
• Large jitter values result in large buffer
  requirements and high end-to-end delay
• All guarantees necessary for achieving data
  transfer within the time bound must be met

59
Peer-to-peer

• Comprises 44 of the network traffic in 2008 in
  North America
• Web browsing(27.3), streaming(14.8), VPN(5.9),
  etc.
• Swarming technology
• Split the large content into pieces
• Exchange pieces with n different peers
• Democratic distribution of content
• How to find the peers?
• Can we trust the data from the peers?
• Quality of service?

60
Summary

• The Internet architecture
• The good, the bad and the ugly
• Simple network with open innovation opportunity
  at the edge
• Very successful but has room for improvement
• Basic building blocks for WAN
• Multiplexing, switching, addressing
• Flow error control, congestion control
• Networks should support various applications
• Multi-service networks not tied to a single
  service
• Support both latency sensitive and bandwidth
  intensive applications fair to different kinds
  of traffic

61
Reading assignment

• Two seminal papers on the Internet
• Design Philosophy of DARPA Internet by Clark
• End-to-end principles by Saltzer, Reed and Clark
• Will briefly discuss them next class
• Dont forget to send your picture URL!