15-744: Computer Networking

Intro to Computer Networks

Outline

• Administrivia

• Layering

2
Who’s Who?
• Professor: Srinivasan Seshan
• http://www.cs.cmu.edu/~srini
• [email protected]
• Office hours: by appt.
• TAs:
• Devdeep Ray
• http://www.cs.cmu.edu/~devdeepr/
• Devdeepr [at] cs [at] cmu [dot] edu
• Vamshi Konagari
• vkonagar [at] andrew.cmu.edu
• Course info
• http://www.cs.cmu.edu/~15744/

3
Objectives
• Understand the state-of-the-art in network
protocols, architectures and applications
• Understand how networking research is
done
• Teach the typical constraints and thought
processes used in networking research
• How is class different from undergraduate
networking (15-441)
• Training network programmers vs. training
network researchers

4
Web Page
• Check regularly!!

• Course schedule
• Reading list
• Lecture notes
• Announcements
• Assignments
• Project ideas
• Exams
5
Discussion Site
• On Piazza
• Please signup at
http://piazza.com/cmu/spring2018/15744

6
Discussion Site
• For each lecture, post a brief comment about each
paper:
• Since I would like to read the reviews before the lecture, you
should have this done by 5pm the day before the lecture.
• Learn to critique and appreciate systems papers

• Each student will present on the lecture topic once this

semester
• 10 min presentation
• Choose a relevant but uncovered paper
• HW0 for signup

7
How to read a paper (3-pass approach)
1. Skim abstract/intro + section headings +
references (5-10min)
• Make rough assessment of paper
• Many people will read your paper at this level

2. Read but ignore details (e.g. proofs) (1hr)

• Good general understanding of techniques
• Identify related work you need to look at

3. ”Virtual re-implementation” (1-3hrs)

• Identify hidden assumptions
• Identify issues with techniques used

8
How to read a paper
• Learn to be critical
• Many papers are part “marketing” – trying to show their design in
the best possible light
• Some papers may be “old”

• Learn to be positive
• Very easy to become overly critical especially once you know topic
area
• Focus on what you learned from the paper

• Why or why not keep this paper in syllabus?

• What issues are left open for future research?
• What are the important implications of the work?
• What would you have done differently or differently now?

9
Course Materials
• Research papers
• Links to ps or pdf on Web page
• Combination of classic and recent work
• ~40 papers
• Optional readings
• Recommended textbooks
• For students not familiar with networking
• Peterson & Davie, Kurose & Ross, Tanenbaum
& Wetherall

10
Grading
• Homework assignments (15%)
• 4 Problem sets & hands-on assignments
• Class + discussion site participation (10%)
• Midterm exam + final exam (40%)
• Closed book, in-class
• 2 or 3 person project (35%)
• Main focus of class work
• Make project productive for you!

11
Class Coverage
• Little coverage of physical and data link layer
• Little coverage of undergraduate material
• Students expected to know this or learn this along
the way
• Focus on network to application layer
• We will deal with:
• Protocol rules and algorithms
• Investigate protocol trade-offs
• Why this way and not another?

12
Lecture Topics

Traditional Recent Topics

• Layering • Data centers
• Internet architecture • Mobility/wireless
• Routing (IP) • Security
• Transport (TCP) • Content delivery
• Queue management • SDN
(FQ, RED) • P2P systems
• Naming (DNS) • Privacy

+ some TBD slots

13
Homework 0
• Fill Google Form
• 4 lecture choices for presentation
• 1 sentence version of project interest areas &
list of project partner
• E.g., I want to apply game theory to network routing
• 1 request for TBD lecture slot

14
Outline

• Administrivia

• Layering

15
This/Monday Lecture: Design Considerations

• How to determine split of functionality

• Across protocol layers
• Across network nodes
• Assigned Reading
• [SRC84] End-to-end Arguments in System
Design
• [Cla88] Design Philosophy of the DARPA
Internet Protocols
• Optional Reading
• [CT90] Architectural Considerations for a New
Generation of Protocols

16
What is the Objective of Networking?
• Communication between applications on
different computers
• Must understand application
needs/demands
• Traffic data rate
• Traffic pattern (bursty or constant bit rate)
• Traffic target (multipoint or single destination,
mobile or fixed)
• Delay sensitivity
• Loss sensitivity

17
Back in the Old Days…

18
Packet Switching (Internet)

Packets

19
Packet Switching
Positives Challenges

• Interleave packets from • Store and forward

different sources • Packets are self contained
• Efficient: resources used units
on demand • Can use alternate paths –
reordering
• Statistical multiplexing
• Contention
• General
• Congestion
• Multiple types of
• Delay
applications
• Allows for bursty traffic
• Addition of queues

20
Internet[work]
• A collection of Internet[work]
interconnected
networks
• Host: network
endpoints (computer,
PDA, light switch, …)

21
Internet[work]
• A collection of Internet[work]
interconnected
networks
• Host: network
endpoints (computer,
PDA, light switch, …)
• Router: node that
connects networks

22
Challenge
• Many differences between networks
• Address formats
• Performance – bandwidth/latency
• Packet size
• Loss rate/pattern/handling
• Routing
• How to translate between various network
technologies?

23
Challenge 1: Address Formats
• Map one address format to another?
• Bad idea à many translations needed
• Provide one common format
• Map lower level addresses to common format

24
Challenge 2: Different Packet Sizes
• Define a maximum packet size over all
networks?
• Either inefficient or high threshold to support
• Implement fragmentation/re-assembly
• Who is doing fragmentation?
• Who is doing re-assembly?

25
Gateway Alternatives
• Translation
• Difficulty in dealing with different features
supported by networks
• Scales poorly with number of network types
(N^2 conversions)
• Standardization
• “IP over everything” (Design Principle 1)
• Minimal assumptions about network

26
IP Standardization
• Minimum set of assumptions for underlying net
• Minimum packet size
• Reasonable delivery odds, but not 100%
• Some form of addressing unless point to point

• Important non-assumptions:
• Perfect reliability
• Broadcast, multicast
• Priority handling of traffic
• Internal knowledge of delays, speeds, failures, etc

27
How To Find Nodes?

Internet

Computer 1 Computer 2

Need naming and routing

28
Naming

What’s the IP address for www.cmu.edu?

It is 128.2.11.43

Computer 1 Local DNS Server

Translates human readable names to logical endpoints

29
Routing

Routers send
packet towards
destination H R

H
R R
H

R
H
R
R H: Hosts
H R: Routers

30
Meeting Application Demands
• Reliability
• Corruption
• Lost packets
• Flow and congestion control
• Fragmentation
• In-order delivery
• Etc…

31
What if the Data gets Corrupted?

Problem: Data Corruption

GET index.html GET windex.html

Internet

Solution: Add a checksum

0,9 9 6,7,8 21
X
4,5 7 1,2,3 6

32
What if Network is Overloaded?

Problem: Network Overload

Solution: Buffering and Congestion Control

• Short bursts: buffer
• What if buffer overflows?
• Packets dropped
• Sender adjusts rate until load = resources à “congestion control”

33
What if the Data gets Lost?

Problem: Lost Data

GET index.html
Internet

Solution: Timeout and Retransmit

GET index.html
GET index.html
Internet
GET index.html

34
What if the Data Doesn’t Fit?

Problem: Packet size

• On Ethernet, max IP packet is 1.5kbytes

• Typical web page is 10kbytes

Solution: Fragment data across packets

ml x.ht inde GET

GET index.html

35
What if the Data is Out of Order?

Problem: Out of Order

ml inde x.ht GET

GET x.htindeml

Solution: Add Sequence Numbers

ml 4 inde 2 x.ht 3 GET 1

GET index.html

36
Lots of Functions Needed
• Link
• Multiplexing
• Routing
• Addressing/naming (locating peers)
• Reliability
• Flow control
• Fragmentation
• Etc….

37
What is Layering?
• Modular approach to network functionality
• Example:

Application

Application-to-application channels

Host-to-host connectivity

Link hardware

38
Protocols
• Module in layered structure
• Set of rules governing communication
between network elements (applications,
hosts, routers)
• Protocols define:
• Interface to higher layers (API)
• Interface to peer
• Format and order of messages
• Actions taken on receipt of a message

39
Layering Characteristics
• Each layer relies on services from layer
below and exports services to layer above
• Interface defines interaction
• Hides implementation - layers can change
without disturbing other layers (black box)

40
Layering

User A User B

Application

Transport

Network

Link

Host Host

Layering: technique to simplify complex systems

41
E.g.: OSI Model: 7 Protocol Layers

• Physical: how to transmit bits

• Data link: how to transmit frames
• Network: how to route packets
• Transport: how to send packets end2end
• Session: how to tie flows together
• Presentation: byte ordering, security
• Application: everything else

42
OSI Layers and Locations

Application

Presentation

Session

Transport

Network

Data Link

Physical

Host Switch Router Host

43
IP Layering (Principle 2)
• Relatively simple
• Sometimes taken too far

Application

Transport

Network

Link

Host Router Router Host

44
IP Hourglass
• Need to interconnect many
existing networks
• Hide underlying technology
email WWW phone...

SMTP HTTP RTP... Applications

from applications TCP UDP…

IP

• Decisions: ethernet PPP…

CSMA async sonet... Technology

• Network provides minimal copper fiber radio...

functionality
• “Narrow waist”

Tradeoff: No assumptions, no guarantees.

45
Is Layering Harmful?
• Sometimes..
• Layer N may duplicate lower level functionality
(e.g., error recovery)
• Layers may need same info (timestamp, MTU)
• Strict adherence to layering may hurt
performance

46
Friday Lecture: Recitation on Routing
• Background material from 15-441/641
• Basics of routing on the Internet
• Link-state
• Distance-vector
• Forwarding
• Address allocation
• IP basics

47
Monday Lecture: Design Considerations
• How to determine split of functionality
• Across protocol layers
• Across network nodes
• Assigned Reading
• [SRC84] End-to-end Arguments in System
Design
• [Cla88] Design Philosophy of the DARPA
Internet Protocols

48
Homework 0
• (by Friday) Fill Google Form
• 4 lecture choices for critique/public review
• Note that lecture dates may shift
• 1 topic choice for first TBD lecture
• 1 sentence version of project interest & list of
project partner(s)
• E.g., I want to apply game theory to network routing.
• We will be posting some questions on Piazza
to answer about the two papers for Monday’s
lecture (post response by 5pm on Sunday)

49
Layer Encapsulation

User A User B

Get index.html

Connection ID

Source/Destination

Link Address

50
Protocol Demultiplexing
• Multiple choices at each layer

FTP HTTP NV TFTP

TCP UDP
Network IP TCP/UDP
IPX IP
Type Protocol Port
Field Field Number
NET1 NET2 … NETn

51