Jan M' Rabaey, Adam Wolisz, Kannan Ramchandran, Paul Wright and the NAMP group

1
Topics In Sensor Networks

• Jan M. Rabaey, Adam Wolisz, Kannan Ramchandran,
  Paul Wright and the NAMP group
• EECS Dept.
• Univ. of California, Berkeley

2
The Sensor Network Implementation Challenge

• Meso-scale low-cost wireless transceivers for
• ubiquitous wireless data acquisition that
• are fully integrated
• Size smaller than 1 cm3
• minimize power/energy dissipation
• Limiting power dissipation to 100 mW
  enables energy scavenging
• support low data-rates (lt 100 kBit/sec)
• and form self-configuring, robust, ad-hoc
  networks containing 100s to 1000s of nodes

Berkeley PicoRadio Project
3
The road to low-energy, low-cost, small-sizeNot
wireless as usual!

• Simplicity rules!
• Advanced techniques used in traditional wireless
  links are not necessarily relevant
• Standby power the greatest enemy
• Monitoring connectivity dominates overall power
• Leakage dominates digital power
• Redundancy as a means to create robustness
• Elements and links can and will fail
• The environment and its conditions change rapidly

4
Simplicity Rules
Thin-Film Bulk Acoustic Resonator
OSC1
Preamp
PA
MOD1
Matching Network
OSC2
MOD2
FBAR-based

• Minimizes use of active components exploits new
  technologies
• Uses simple modulation scheme (OOK)
• Allows efficient non-linear PA
• Down-conversion through non-linearity (Envelope
  Detector)
• Tx and Rx in 3-4 mW range (when on)

(PicoRadio RF)
5
We indeed can make it small
PicoBeacon An Energy-Scavenging Radio
Regulator
RF Transmitter
Energy StorageCapacitor (10 mF)
Single solar cell
Antenna (ceramic)
6
Standby Power The Greatest Enemy
1 packet/sec

• Parameters
• 200 bits/packet
• 20 bit pre-amble
• 5 neighbors
• Synchronization using cycled receiver

0.1 packet/sec
7
Cycled Receiver Maximizing the Sleep Time
TICER (Transmitter Initiated CyclE Receiver)
RICER (Receiver Initiated CyclE Receiver)
(En-yi Lin and Charlie Zhong)
8
Reactive Radio Maximizing the Sleep Time
TX
RX
Shifts Burden to Transmitter Reduces monitoring
power to lt 50 mW
9
Maximizing the Sleep Time

• When does wakeup radio to perform better than
  TRICER?

1 pckts/sec, Pidle0.05mW ok
3 pckts/sec, Pidle0.1mW ok
The lighter the traffic, the lower Pidle has to
be to beat TRICER!!
(En-yi Lin)
10
Sleep Discipline

• SLEEP IF YOU CAN
• Maintain connectivity
• If the node is not necessary, goes to sleep and
  saves power
• For how long should the node be allowed to sleep
  ?

• Given
• Loss rate
• Delay constraint
• Data generation requirement

• Our Solution
• Adaptive
• Traffic node density
• Random
• Exponentially distributed sleeping times.
• Avoid phase synchronization.

Sleep discipline as a meansof controlling
network density
(J. Van Gruenen, A. Benivento, D. Petrovic )
11
Comparison Results
Adaptive sleeping time wins !!!
Exponentially distributed sleeping discipline has
a better load distribution !!
Exponentially distributed sleeping time
Deterministic sleeping time
12
The Disadvantage of Simple Radios
Factor 105 in error rate

• Small Change in Path Loss Has Dramatic Impact
  on Transmission Quality
• Channel is either good or bad

6 db
Simulated response of PicoNode radio
(U. Schuster)
13
The Variability of Link Quality
Broadcast quality over time as measured at
theBWRC round-table on a Friday
Free Lunch
NAMP Meeting
PicoRadio Meeting
14
Providing Robustness

• Traditional radios provide robustness through
  diversity
• Frequency e.g. wide-band solutions (hopping)
• Time e.g. spreading
• Spatial e.g. multiple antennas
• All these approaches either come with complexity,
  synchronization, or acquisition overhead, or
  might not even be applicable
• Data traffic irregular, and in very short bursts
• A better approach utilize the system properties!

15
Ad-hoc Multi-Hop Networks
Use redundancy of wireless transceiver nodes to
limit Tx power of individual radio to 0 dBm (10
m)
In additionMulti-hop ad-hoc providesreliability
and robustness
16
The Impact of Spatial Diversity
Adding a single node already changes broadcast
reliabilitydramatically spatial diversity is
the preferred way toprovide robustness in sensor
networks
17
Routing exploiting Spatial Diversity(or
Opportunistic Routing)
Path-based routing Network specifies next hop
One-hop neighbors

• List-based routing
• Network specifies forwarding region
• MAC chooses next-hop based on connectivity

Forwarding region
Potentially performs better than average fading
channel
(R. Shah)
18
Routing exploiting spatial diversity
(R. Shah)
19
How to obtain geographic information?
anchor
radio range
Rough topological informationis sufficient for
routingIn fact, only sense of direction is needed
intermediate node
Hop-terrain use number ofhops from anchors as
distance measure
Locationing engine 0.33 mm2in 130 nm CMOS
(T. Karalar)
20
Using UWB for Precise Ranging

• A sub-nanosecond UWB pulse bears timing
  information by detecting the time of arrival. 1ns
  corresponds to 1 foot of ranging accuracy.
• We are now proposing a different approach to
  enhance the ranging capability of UWB. The UWB
  pulse will occupy between 3GHz and 10GHz as FCC
  regulated for indoor communication systems.
• This timing information can be extracted by
  subsampling combined with analytic filtering UWB
  architecture.

(M. Chen)
21
Sub-sampling UWB Radio

• Ranging Application
• By making use of the center frequency of the
  passband UWB signal, we can extract even finer
  timing information than baseband UWB pulse.

22
Making the Network Work
Statistics and Management Service Allows to
capture network statistics at physical, link,
network, and application level
Used PicoNode 1 testbed to analyze and improve
performance
(J. Reason, R. Stutz)
23
A Very Rich Set of Data
Channel characteristics
Channel characteristics
Routing analysis
Interarrival times
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
(No Transcript)
29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
Everyday Computing
Addressing the everydayproblem (e.g. Where are
my keys?)
Courtesy of Mik Lamming and Jim Rowson, HP Labs
Pushes the synchronization issue! No predictive
traffic model.
35
Targeted Application The Energy-Aware Home
(P. Wright, E. Arensin collaboration with CEC)
36
Collaboration with GSRC
Capturing, optimizing, and quantifying network
protocols using the Metropolis Framework
(R. Chen, ASV)
37
Some New Awards (last 6 months)

• Reto Stutz received the EPFL Maillefer Price for
  his Diploma Thesis
• Its aim is to reward a diploma thesis which
  differs from the others in term of its
  originality, its potential for industrial
  application and its economic realism.
• The award of 4000 was presented in the presence
  of the president of Switzerland, the skipper of
  Alinghi (winner of America Cup), the French
  minister of research and new technologies (and
  first French female astronaut) and the president
  of Stanford.
• PicoRadio II chip set was awarded the second
  price at the DAC University Design Contest (out
  of more than 40 entries). Congratulations Mike,
  Josie, Tufan and Mika!
• PicoRadio received the ISSCC 2002 Jack Raper
  Award for Excellence in Technology Directions.

38
The Sensor Network Session

• Demand Response Enabling Technologies, Paul
  Wright
• Exploiting Spatial Diversity for Routing in
  Sensor Nets, Rahul Shah 
• Analytical Power Model For a Wireless Sensor
  Node, En-Yi Lin
• A Bandpass Radio Architecture for 3-10GHz UWB,
  Mike Chen