Jan M' Rabaey

1
Ubiquitous Wireless and Disappearing
Electronics (aka PicoRadio)

• Jan M. Rabaey
• EECS Dept.
• Univ. of California, Berkeley

2

The Research Agenda

• Towards easily deployable and robust
  self-configuring ubiquitous wireless networks
• Energy, cost and size awareness at all the levels
  of the abstraction chain

Towards a universal, scaleable and portable
application interface for ubiquitous wireless
Testbeds
Applications and Applications Support
Networking, Applications, Media Access, and
Positioning
Unified energy and power efficient protocol stacks
Computational Platforms
Bringing size and power/energy levels down to
unprecedented levels
RF and Baseband
Energy Generation
Integrating, prototyping and testing all of the
above in realistic environments
3

• RF
• Yuen Hu Chee
• Brian Otis
• Nathan Pletcher

• Mixed-Signal andBaseband
• Josie Ammer
• Yan-Mei Li
• Simone Gambini

The Team

• Application, Networkand Services
• Delynn Bettencourt
• Alvise Bonivento
• Stanley Chen
• Tufan Karalar
• Enyi Lin
• Dragan Petrovic
• Rahul Shah
• Jana Van Greunen

• Digital Processing
• and Design Tools
• Louis Alarcon
• Fred Burghardt
• Huifang Qin
• Mike Sheets
• Thuan Trinh
• Allan Tsao
• Ruth Wang
• Rong Zhang

• Energy Scaveningand Packaging
• Eric Carleton
• Elaine Lai
• Eli Leland
• Mike Montero
• Dan Odell
• Elizabeth Reilly

joined effort with NEST
4
A Service-based Universal Application Interface
How to ensure portability, scalability and
configurability inAd-hoc Sensor and Actuator
Networks in presence of multiple hardware
platforms and network protocols? RAISE THE
ABSTRACTION LEVEL
Application Space
Application Instance
Platform Mapping
AI Platform
Application Interface (AI)
Platform Design Export
Platform Instance
Architecture Space

• Defines the services needed to operate a sensor
  network
• Independent of network architecture and hardware
  platform
• Enabling interoperability and innovation

Adam Wolisz (TUB), Jana Van Greunen, Delynn
Bettencourt, Fred Burghardt, Alvise Bonivento,
Alberto Sangiovanni-Vincentelli
5
The Sensor Network as a Distributed Database
Typical application call Get the temperature in
the kitchen
(Equivalent Command Service interacts with
actuators)
Originated from TinyDB (UC Berkeley)
6
Prototype The Demand-Response Application
Distributed Control
Sensors
One or more controllers (advanced thermostats)
Actuators
Energy sensors Temp Sensors
Determine what and when to turn on/off based on
readings from sensors and do this using actuators
CALISO
WWW
Virtual sensors
Virtual actuators
7
A Formal Design Methodology
Alvise Bonivento, Alberto Sangiovanni-Vincentelli
8
Extending the scope to consumer multimedia
networks?

• New devices are entering the home environment at
  an ever increasing rate.
• Standards are proliferating interconnect,
  recording and playback, display
• Devices do not necessarily interconnect easily
• Configuration a disaster

9
How to Deal with the Myriad of Protocols and
Formats?
Put the Intelligence in the Network Smart
Home Routers Provide on-the-fly protocol
conversion and trans-coding based on properties
of source and destination devices
Home router Flexible, dynamically
reconfigurable, upgradeable
Stanley Chen, K. Lutz, J. Wawrzyniek - In
collaboration with GSRC
10
Ad-hoc Networking and Media Access Combining
Energy-Efficiency and Reliability Through
Redundancy and Randomness
Developed complete modeling and simulation
environment (Omnet, Matlab, BEE)
NAMP Group
11
Ad-hoc Networking and Media Access
Goal Providing an integrated energy efficient,
truly ad-hoc and absolutely reliable network
stack from physical to application layer
The answer is yes!

• The Grand Challenge
• Can individual nodes set their sleep time based
  on purely local observations so that network
  lifetime and latency constraints are met?

Actual total wakeup rate
Optimal wakeup rate
Change in traffic rate
Presentation by Rahul Shah
12
Locationing Essential Component
Accurate ranging the most challenging
part Propose wide-band (OFDM based) two-way
approach
Presentation by Tufan Karalar
13
Extending the Paradigm Dense NetworksStatistical
communication Strength in Numbers
Forwarding node
Not only extremely reliable, but also guaranteed
throughput (using distributed coding)

• Random frequency multi-hopping
• Information packet traverses from source to
  destination in a multi-hop fashion.
• Transmitter broadcasts signal to neighboring
  block on random channel (as determined by process
  variations).
• Receivers randomly select channel to listen to.

Dragan Petkovic and K. Ramchandran
14
Towards a Fully Integrated Self-Contained Node
Moving the bar ever higher (that is, lower)

• Active ultra low-power efforts on all components
  of the puzzle
• Aggressive use of alternative technologies
  (RF-MEMS, integrated passives, ) when
  appropriate
• Integration is key!

15
The PicoCube
Advanced 2.5 packaging the only real answer to
mm3 nodes
In collaboration with P. Wright, D.
Steingard,and S. Sanders
16
Some First Steps
17
Operational Super-Regenerative Receiver
2mm
1mm
1 mm
1.9 mm

• No External Components (inductors, crystals,
  capacitors)
• Full Digital control of analog/RF blocks
• -Rx Osc Frequency 4MHz x 500kHz-BB Filter
  BW/Gain-Tx Tank Frequency
• Power Consumption 400mW Receive, 1.8 mW
  Transmit

To be presented at ISSCC 2005
B. Otis and Y.H. Chee
18
Improving TX Efficiency
MICROS1 Direct Conversion TX (30mW, ?3.3)

• Reduced Complexity
• RF MEMS Oscillator
• Capacitive Transformer

• Active Antenna
• TX Chain co-design
• Lower Power Oscillator
• Low VDD

1P. Choi et. al, ISSCC 2003 2B. Otis, Y.H. Chee,
et. al, VLSI 2004
Y.H. Chee (tomorrow)
19
Active Antenna Transmitter
20
Moving the bar againUltra-Low voltage RF design
(wake-up radio)

• Demonstrated 1.4GHz LC oscillator (summer 2004)
• Operation below 300mV is possible
• 2 ns startup time
• Challenge Calibration/Tuning
• Self tuning oscillator

Just back from fab
Nate Pletcher
21
Baseband Processing
Digital wins for packets lt 400 bits Even assuming
analog synch circuits consume no power!!
Josie Ammer, Yanmei Li, and Simone Gambini
22
The Quest for Ultra Low-Voltage Digital Design
Goal Operate next generation PicoNode at 300 mV
or below
Presentation by Ruth Wang tomorrow

• Solution
• Trade-off performance for voltage and/or leakage
  management
• Challenges
• Keeping performance penalty bounded
• Dealing with process variations
• Keeping the memories operational (data retention)

Yoda group (http//bwrc.eecs.berkeley.edu/Research
/yoda ) Huifang Qin, Ruth Wang, Paul Friedman,
Andrei Vladimirescu, Costas Spanos
23
Managing leakage while reducing thresholds

• Stacked structures enable aggressive voltage
  scaling
• Ion/Ioff increases with increasing stack height
  (leakage suppression)
• More robust to correlated (tune or adapt) and
  random variations (self-cancel)
• Decreased short channel effect

• PLAs as an attractive fabric
• NAND/NAND configuration offers high stack heights
• Regular (good for variation control and OPC)
• tunable

Mircea Stan, Louis Alarcon, Bob Brayton
24
Ultra Low Voltage SRAM Design

• Explore techniques for robust, ultra low voltage
  (ULV) SRAM design.
• Developed an analytical model for SRAM Data
  Retention Voltage (DRV) with process variations.
  
• Optimize voltage bounds using
• ULV sizing techniques
• Error-tolerant SRAM design
• Results 80 leakage power saving in standby
  operation at DRV100mV by tuning the sizing, body
  bias voltage and bitline voltage.
• Ongoing work Error-tolerant design to further
  reduce SRAM DRV and leakage power.

• 90nm SRAM test chip submitted October 2004
• Studies impact of sizing on DRV

Huifang Qin, Thuan Trinh
25
Integrated Silicon clocks

• 16MHz micromachined resonators SiGe structural
  layer, low temp processing
• Two oscillators
• 1mW Low Power Osc
• 100mW Low Phase Noise Osc

2mm
Brian Otis, N. Pletcher, E, Quevy
26
Summary

• A Quite Productive Year
• A whole slew of silicon prototypes
• Numerous presentations at major conferences (see
  web-site)
• Best paper award at ISQED 2004
• Keynote presentations at ASPDAC 2005, CCNC 2005,
  MICS04, DesignCon 04
• Book on Ambient Intelligence in the press soon
• Pro-active collaboration with member companies
  (Agilent, Infineon, Philips, ST)
• Continuing along the same track, while widening
  the scope (towards other ad-hoc areas) and
  focusing on even more aggressive technology
  scaling and integration