IEEE SECON 2004

1
IEEE SECON 2004

• Andres Lagar Cavilla

2
Sensor and Ad Hoc Communications and Networks

• First edition
• IEEE Communications Society
• 4-7 October 2004
• Santa Clara, California

3
Miscellaneous Stats

• 68/358 papers accepted (19)
• Two keynote speeches
• Four Panels
• Four Tutorials

4
Overall Impresions

• The Industry is hyping about this
• Sensor nets preferred over Ad hoc nets
• More tangible right now
• But the Industry is also waiting

5
The Killer Application

• A panacea?
• Everybody is looking for it. No one is risking
  anything
• Most of the funding comes from the military
• No clear guidelines

6
The chicken egg problem

• Standards to support the development of
  applications are missing
• Applications to propel (and steer) the
  development of standards are missing

7
However

• Some startups are taking the first strides
• Crossbow
• Sensoria
• PacketHop
• Many industry envoys wre probing the field
  during the conference
• Exponential growth in investment predicted

8
Three things to think about at home

• 802.15.4
• IEEE standard to integrate multiple radio
  interfaces in a network
• ZigBee
• An industry alliance, equivalent to WiFi for
  sensor networks and 802.15.4
• Ultra Wide Band
• No spectrum constraints
• Quasi-unlimited propagation

9
A National Sensor Network

• John Strand, Oak Ridge National Labs
• A review of what US army has been doing on sensor
  networks
• Total Installation Awareness
• Scary!!!!
• But they claim to have it working

10
Commercial Wireless Sensor Networks

• Mike Horton, Crossbow Tech
• First field for deployment of sensor nets will be
  on monitoring tasks
• Last field will be ubiquitous computing
• Enabling technology is already here
• Integrating standards in order to provide
  solutions are missing
• Everything Crossbow does is Open Source

11
Greatest technological challenges in deploying
MANETS

• Multi network roaming
• Cross layer design between L2/L3
• Support for hybrid networking interfaces
• Integration of Internet gateways
• MANETs are hardly self-managing
• PacketHop people insisted on this
• Lack of service discovery

12
Mesh networks

• Provide community broadband access
• Multihop routing
• Nodes are neither mobile nor resource constrained
• Lots of impulse from Microsoft, via Viktor Bahl
• Skepticism from pretty much everyone else

13
Paper session Applications

• Controlled Mobility for Sustainable Wireless
  Sensor Netorks
• Controlled mobility enhancing sensor network
  coverage in the wild
• eBlocks Electronic Blocks for Basic
  Sensor-Based Systems
• Educational students would try to assemble
  systems from basic blocks
• Analysis of the performance of IEEE 802.15.4 for
  Medical Sensor Body Area Netowrking

14
Other interesting papers

• The Limits of Localization Using Signal Strength
  A Comparative Study
• A review of localization algorithms, concluding
  theyre all pretty much the same
• ATEMU A Fine-Grained Sensor Network Simulator
• A HUGE Mica Mote simulator at the CPU instruction
  level