Network%20Research%20at%20CDM

1
Network Research at CDM

• James Yu, Ph.D.
• Associate Professor
• School of Computing
• jyu_at_cdm.depaul.edu

2
James Yu, Ph.D.

• Ph.D. in Computer Science from Purdue University,
  1985
• Distinguished Member of Technical Staff, Bell
  Labs, Lucent Technologies, 1985 2000
• Director of Network Technologies, ARBROS
  Communications, 2000 2001
• Assistant/Associate Professor, DePaul University,
  2002 - now

3
Research Interests

• Wireless Security
• VoIP Traffic Engineering
• Network Management
• Fault Tolerant Network Design
• Network education hands-on exercises for
  distance learning students

4
Framework of Network Research (I)
Problem Statement
Empirical Framework
Theoretical Model
References
Solutions to The problem
Revision
Data Analysis
Comparative Study
Conclusion
5
Framework of Network Research (II)
Problem Statement
Theoretical Model
Solutions to The problem
Empirical Framework
Data Analysis
Data Analysis
Comparative Study
Your solution must be theoretical sound along
with empirical support.
Conclusion
6
Problem Statement

• Is the problem significant and interesting?
• References, References, References
• Can you describe the problem in one or two
  sentences?
• Can you describe the problems with numbers?
• Is the problem solvable?
• Is the problem already solved?
• Do not confuse the problem with the solution.

7
Measurement
When you can measure what you are speaking
about, and express it in numbers, you know
something about it but when you cannot measure
it, when you cannot express it in numbers, your
knowledge is of a meager and unsatisfactory kind
it may be the beginning of knowledge, but you
have scarcely in your thoughts advanced to the
state of Science, whatever the matter may be."
Lord Kelvin, PLA, vol. 1, "Electrical Units of
Measurement", 1883-05-03
8
Dependent Variable(s) vs.Independent Variables
Y F (X1, X2, X3, ..)
9
Review of Research Paper

• 1-2 page of each paper you read
• Problem Statement
• Results and Major Conclusions
• Research Methodology (empirical or theoretical)
• Comparison with other research work
• Your assessment
• Can you reproduce the research results?
• Can you improve the research results?

10
Network Research Projects
11
Network the BIG Picture
PSTN
Network Services and Management
NNI
Switched Network
Internet
Access Network
Security Layer
Transport Network
ISP Network
UNI
carrier network
Security Layer
Security Layer
Data LAN
Voice Network
2G/3G Network
Data Center
Call Center
cellular network
converged enterprise network
12
Project I Performance Analysis
Key measurement throughput (bps) and delay (ms
or ?s)
Ethernet Switch
IP Router
Linux Bridge/Router
Windows Bridge/Router
13
Switch Performance Comparisonon Gigabit Ethernet
Question could you develop a problem statement
from this chart?
14
Wireless Performance
Question why is TCP performance is only 50 of
the line rate for wireless communications?
15
Project II Wireless Denial of Service (DoS)
Research
16
TCP performance under DoS
Stop
Start
17
Project III VoIP Research
PSTN Switch
PSTN Switch
SoftSwitch
SoftSwitch
Carrier VoIP Network
Trunk MG
Trunk MG
Call Manager (SIP Proxy)
Q.931
Access MG
Call Manager (Enterprise)
MG Media Gateway
18
VoIP Traffic Engineering

• Design the network with sufficient capacity to
  meet the traffic demand with satisfactory
  performance
• Requirement assurance of Voice Quality.
• If call quality cannot be assured, reject the
  call request.
• Demand (A) - Traffic Intensity
• number of calls duration of average calls
• Erlang
• Resources (N) Network bandwidth (bps) and
  capacity (pps)
• Grade of Service (GoS) blocking probability

19
Research Areas of VoIP

• IP Multimedia Subsystem (IMS)
• How is a VoIP call from one service provider to
  another service provider today?
• VoIP Security
• Voice Quality
• Quality of Service (QoS) over internet and
  intranet
• Supplemental and Intelligent Network (IN)
  services
• API and VoXML

20
Project IV Fault Tolerant Network
RSL
Dual Homing
VRRP
RSTP (802.1w)
ISP-01
ISP-02

• Network Design Goal
• No single point of failure
• Failover time ltlt 1 sec

21
Project V Remote Lab Access
Linux Gateway
Public Internet
Private Networking Labs
Windows Terminal Server
Devices w/ Console interface
Devices w/ IP interface
Telecom Terminal Server
Console interface
Problem Statement supporting distance learning
students to perform hands-on network exercises
from any place with 247 access to the network lab
22
Research on Networking Education

• Does the DL environment support the same learning
  goals effectively as the in-class students?
• How do you define and measure the conformance to
  the learning goals and the effectiveness of the
  learning environment?
• Does multiple choice questions measure the
  learning as effective as other methods?

23
Project VI Network Management
Command-Oriented
Vendor specific
Variable-Oriented
SNMP/MIB
Object-Oriented
CORBA
Document-Oriented
XML-Based
Transaction-Oriented
NETCONF
24
NETCONF

• A new Protocol from IETF
• It is designed to address many current issues of
  network mangement.
• XML-based and Transaction-Based
• The XML standard is well defined
• The XML Schema allows flexible definition of the
  structured management information
• download/upload of complete device configuration
• human-readable
• cost-effective to develop new applications and
  services
• Integration with security infrastructure
• Open issue a lack of data models

25
Research Deliverables

• No deliverables no research
• Publications, Publications, Publications
• Journal Papers only if accepted
• Conference Papers
• Technical Reports (little weight)
• Grant proposals only if awarded
• Whitepapers? (does not count)
• Tools, Demo, etc good only if you have
  publications associated with them.

26
Technical Writings

• It is more important than your technical
  capabilities. Actually, it is a lot more
  important.
• Take a technical writing course if you have never
  taken one.
• use the DePaul Writing Labs
• William Strunk Jr. the Element of Style
• Yes, we are very picky (and extremely picky)

27
Recommendations

• There is no shortcut for research.
• Your approach may not work. Actually, it is not
  likely to work most of the time, but do not give
  up.
• You may find an elegant solution if you redefine
  the problem. Be flexible.
• If this is your 1st or 2nd year, you may try to
  reproduce with minor refinement of others
  research work.
• Be in a research team. If not, keep regular
  contact and discussion with other Ph.D. students.
  Build your community now (CSC426)
• It is easier to publish at SoCRS and you should
  do it.