Opportunities for Ubiquitous networks

1
Opportunities forUbiquitous networks
Hans-Werner Braun NLANR (UCSD/SDSC) hwb_at_nlanr.net
http//www.nlanr.net
2
Goals and objectives for our network measurement
and analysis activities

• Creation of an infrastructure to support
  measurements and network analysis
• header traces, performance, statistics, routing
• high performance networking environments
• Network measurement and analysis research
  activities
• Support of outside researchers with data and
  other help
• Tool development for analysis and visualizations
• Result reporting for the high performance
  environment
• web interface
• newsletter (Network Analysis Times)
• published papers

3
Passive measurement deployment status
U. of Michigan Michigan State U.
STARTAP/APAN
Argonne Nat. Lab
Ohio State U.
NCAR U. Colorado, Boulder
NCSA
FIX-West
Old Dominion U.
AIX/MAE-West NASA-Ames
MCNC North Carolina State U. U. of North
Carolina Duke U.
SDSC, U. California, San Diego
Rice U. Baylor College of Medicine U. of
Houston Texas AM U.
Tel Aviv U. (I-2)
U. of Florida Miami U. Florida State U.
25 November 1999
4
Optical splitters
5
Status of passive measurements activities

• Current situation
• packet trace data in http//moat.nlanr.net/Traces
  
• pre-analyzed data in http//moat.nlanr.net/Datacu
  be
• 15 active OC3 ATM sites
• one active FDDI monitor
• 2 active OC12 ATM sites
• Near to medium term future
• deploying approximately 24 more passive
  machines
• using the DAG3 technology, OC3 or OC12
• POS and ATM capability
• in collaboration with Abilene/I2

6
Active measurementdeployment status
UAlaska
Washington State U.
UWashington
UVermont
Montana State
Oregon State
MTU
UWisc- Milwaukee
NDSU
UOregon
UWisc
MSU
Dartmouth
SDSMT
URochester
Startap
WSU
UIowa
NTNU
Iowa State
MIT
BU
UMass
NWU
UWyoming
UMich
Harvard
PSU
UIC
UConn
FNAL
Yale
Columbia
ColoState
NCSA
CMU/PSC
UDel
UPenn
Princeton
NCAR
UMBC
UIUC
IU
WVU
JHU
ACCESS
UCBoulder
UMd
Kansas State
NSF
UC
UMissouri
GMU
Georgetown
UCB
UofUtah
UVirginia
UKansas
ODU
Stanford
Oklahoma State
SLAC
WUSTL
UCSC
NCREN/NCSC
UTK
Vanderbilt
Duke
UNC-CH
UofOklahoma
UCLA
CSU-SB
NCSU
GATech
ASU
UC-Irvine
UNM
UAH
CSUPomona
Mississippi State
UofGeorgia
SDSC
UArizona
UCSD
Emory U.
SMU
UAB
SDSU
NMSU
UA
FSU
UFlorida
Rice
UHawaii
UCF
USF
UMiami
UWaikato
23 February 2000
7
Major goals for AMP

• Have a joint research/engineering agenda for the
  high performance community
• third party site-to-site measurements between
  NSF supported HPC sites
• to complement the intra-HPNSP measurements
• to provide a common point of understanding of
  performance issues between HPNSPs and users
• to work towards user-user and application-applica
  tion performance assessments
• to provide a base for further research

8
Active measurement environment
Dept/ user
Dept/ user
HPC backbone network
Site (campus)
Site (campus)
Dept/ user
Dept/ user
Dept/ user
Dept/ user
inner perimeter
site perimeter
user/application perimeter
9
IP use and plen matrices
10
Cichlid 2
11
Collaborations

• availability of network workload and performance
  data and software to gain more insight into the
  Internet fabric
• opportunity to involve more students and
  faculty thesis projects
• hosting of visiting researchers
• local (UCSD) student involvement
• faculty and student collaborations with other
  sites
• collaborators have to be self-guided to a large
  extend

12
Create high performance networks --where no
network has gone before.(well, where it is
uneconomical for commercial providers.)

• Understand and define actual needs
• Coopt needy constituents
• Try to coopt colleagues with prior experiences
• Look for funding
• Collaborate and cost/effort share
• Create multi-usage, multi-constituent
  environment

13
High Performance Wireless Research and Education
Networkcase study
14
The need for speed?

• What constitutes network performance
• raw bandwidth?
• availability?
• real-time communication, predictable response
  times?
• total ubiquity, quick setup on a need basis?
• affordability?
• What do YOU need? What are YOUR performance
  metrics?
• Earthquake sensors low bandwidth,
  predictability, ubiquity?
• Ecologists telemetry low bandwidth,
  predictability, ubiquity?
• Astronomer data high bandwidth, not necessarily
  continuously?
• Education availability, high bandwidth
  opportunity
• Field researchers . . . .
• Researchers living in rural areas . . . .

15
Participants of a local project

• National Laboratory for Applied Network Research
• measurement and network analysis group
• Scripps Institution of Oceanography
• San Diego Supercomputer Center
• School of Engineering
• Center for Wireless Communications
• San Diego State University
• Astronomy department

16
Needs for multi-level ubiquity, create structure

• backbone infrastructure
• e.g., creating a geographic corridor
• fixed in-zone services, e.g.
• reaching residential houses (high volume,
  bursty)
• remote telemetry sites (earthquake sensors, low
  volume, high priority)
• mobile services
• omni-directional, perhaps at reduced data rates
• in-building communications, e.g.
• residential house
• campus building

17
Topology considerations
18
Wireless environment

• meshed infrastructure as an enabler for
• systemic integration of wireless technologies
• inclusive and open-ended project
• multi-technology
• multi-vendor
• multi-objectives
• wide area
• local area (including in-building)
• multiple communities
• researchers (e.g., geophysics, astronomy,
  ecology)
• telemetry needs
• disadvantaged communities (e.g. native res
  locations)
• various applications

19
Some issues for an integrated system

• system performance assessments and metrics
• how to centrally control access (MAC filters?)
• how to prioritize traffic (at RF level, based on
  MAC addresses?)
• utilizing unlicensed spectrum
• how to integrate new and interoperable
  technologies
• access to microwave towers to create the
  backbone
• RF radiation concerns
• integration of capillary nodes (e.g., laptops)
• integration of data generators (e.g., seismic
  sensors)
• integration of a campus infrastructure (e.g.,
  UCSDs wireless trial)
• and of course -- funding (industrial
  partnerships, agencies)

20
Available unlicensed spread spectrum ranges
0.9 GHz (26 MHz allocated spectrum) 2.4
GHz (83.5 MHz allocated spectrum) 5.8 GHz
(300 MHz allocated spectrum) 24 GHz (250
MHz allocated spectrum) 60 GHz (5000 MHz
allocated spectrum)
21
Stephenson Peak
22
Initial prototype
backbone
level-1
level-2
23
Prototype system
Mt. Woodson backbone site
p-p radio
Amplifier
F i l t
Hans-Werner Braun location
UCSD/SDSC backbone site
measure- ment
Weather station
measure- ment
p-mp radio
measure- ment
Weather station
Weather station
NAT gateway
p-p radio
Amplifier
F i l t
p-p radio
Amplifier
SDSC router
Doug Bartlett location (SIO Microbiologist)
HPC
p-p radio
laptop
commodity
24
Proposedbackbonetopology
Towards Toro Peak
Palomar Mountain
Mt. Woodson
Mt. Laguna
Cuyamaca Peak
UCSD
Otay Peak
25
Backbone node layout
wide area component
IP router
local area component
measurement host (PMA AMP)
application host
environment telemetry
26
View towards UCSD
27
View towards Ramona
28
connection points of prototype network sites
Mt. Woodson commercial microwave tower
SDSC roof
Ramona
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
(No Transcript)