Status of BWRC Jan 2000 Retreat

1
Status of BWRCJan 2000 Retreat
Bob Brodersen Dept. of EECS Univ. of
Calif. Berkeley

• http//bwrc.eecs.berkeley.edu

2
The Center Goal

• Develop methods for specifying, optimizing,
  simulating, verifying and implementing all
  aspects of wireless systems
• Application definition
• Communication algorithm and protocol design
• Analog and digital architectural optimization
• IC implementation and test

3
Center activities

• Two basic center application drivers
• Universal spectrum sharing
• PicoRadios
• Investigate tradeoffs between various
  implementation architectures with respect to
  flexibility, power and area
• Develop a rapid implementation design flow from
  the high level specifications to an integrated
  realization.

4
Application Drivers

• 1) Universal Spectrum Sharing
• An approach to channel utilization which allows
  uncoordinated use of spectra without loss in
  capacity
• Extensible over time to exploit advances in
  technology and support new applications
• 2) PicoRadio
• System on a chip implementation supporting all
  functions up to external interface (sensors,
  transducers)
• Total power dissipation in the 100s of
  microwatts achieved through optimization of
  protocols and architectures

5
Architectural Choices
Flexibility
Embedded Processor
DSP (e.g. TI 320CXX )
Reconfigurable Processor (Maia)
Embedded
FPGA
Direct Mapped
Inefficiency
Hardware (MUD)
6
The Automated Design Environment
Specification (C, Matlab, SDL)
Analog Data Processing
Protocols Control
Behavioral
Digital Data Processing
Behavioral/ Structural
VCC, Opnet, Telelogic, Stateflow
Stateflow Simulink
Matlab, Simulink
Structural
Unicad, Cadence, Synopsys
Spectre
ARMulator,ARM Compiler
Physical
HP EESoft ASITIC Cadence
ARM FPGA Express
Unicad Cadence, Power TimeMill
7
Research Contract Drivers

• Design Environment for Single Chip Radios (DARPA)
  - 1 M/Yr - 9/2000
• Intercom
• Design flow
• PicoNode (DARPA) - 1 M/Yr - 9/2002
• Power Aware Computing Program
• Communications for arrays of sensors
• Ultra Wideband Radios (ONR) 100k/Yr 1/2003

8
Center Activities
Universal Spectrum Sharing
PicoRadio
Applications
Behavioral/Architectural Specification,
Verification And Optimization
Design Tools
Automated Design
Implementation
PicoNode Testbed
BEE Testbed
IC Implementation
9
Outline

• Introduction
• Algorithms
• Design Methodology
• Baseband Analog and RF design

10
Approach to Spectrum Sharing

• Three types of dimensionality in signal space
• Time
• Frequency
• Physical Space
• Need to exploit all these degrees of freedom to
  maximize the number of users and to minimize
  their interference with each other

11
Time dimension

• TDMA and CDMA are ways to exploit this dimension
• Divide up Tsymbol into Nt Tsymbol/ Tchip
  segments
• Nt degrees of freedom in available bandwidth, fBW

fBW1/Tchip
0
Tchip
0
Tsymbol NtTchip
12
Direct Sequence Spread Spectrum
X
tsymbol
Data Input
Spread output data
Spreading code

• Receiver can distinguish between each code
  providing CDMA (Code Division Multiple Access)
• However there is interference if there is
  multipath, so .

13
CDMA with MUD

• Multiuser detection reduces the interference
  between codes due to multipath and thus improves
  the capacity of CDMA

• Transistor Count 0.4 million
• 1.2-2.4 GOP with 25 MHz clock

• Four adaptive pilot correlators
• Die size 3.38mm x 4.58mm

14
Frequency dimension

• How about using the frequency dimension?
• Simple method is FDM (Frequency Division
  Multiplexing), one frequency per user
• OFDM uses all frequencies for each user, like
  DSSS uses all time slots (Nf degrees of freedom)

fBW Nf/Tsymbol
Note
Tsymbol Nf/fBW
0
Tsymbol
0
15
Both together

• Obtain NfNt degrees of freedom
• Many options in combining the two dimensions

Nf/T
Frequency
1/T
T
NtT
Time
16
Now what about physical space?

• Another set of options that can be used in many
  ways
• Increase efficiency so that less signal space is
  used per user (BLAST like algorithms)
• Increase energy in useful directions so that
  signal to noise improves
• Provide isolation between users
• However requires multiple antennas to be flexible
  enough for mobile users

17
Another option Ultra Wideband

• Effort starting up to investigate the issues in
  transmitting data over extremely wide bandwidths
• Data modulated onto extremely fast transitions
  (CMOS is great for that)
• Wideband antenna design is critical antenna
  basically sets the bandwidth
• From information theory standpoint, inefficient
  use of spectrum is best for lowest energy
  transmission

18
Shannon likes UWB!
Bits/Hz
4
3
Energy Limited
2
Bandwidth Limited
1
-5db
Energy/Bit
5 db
10 db
15 db
1/2
UWB
1/4
1/8
1/16
19
This mornings algorithm session

• Spatial processing
• Describes the close relationship of various multi
  antenna array algorithms (Beamforming, BLAST and
  SVD)
• Frequency domain processing
• OFDM and its sensitivity to impairments
• Frequency/time domain options
• MC CDMA, COFDM,
• Proposed frequency/time/spatial design

20
Outline

• Introduction
• Algorithms
• Design Methodology
• BEE

21
The Automated Design Environment
Specification (C, Matlab, SDL)
Analog Data Processing
Protocols Control
Behavioral
Digital Data Processing
Behavioral/ Structural
VCC, Opnet, Telelogic, Stateflow
Stateflow Simulink
Matlab, Simulink
Structural
Unicad, Cadence, Synopsys
Spectre
ARMulator,ARM Compiler
Physical
HP EESoft ASITIC Cadence
ARM FPGA Express
Unicad Cadence, Power TimeMill
22
Simulink description of radio system
Rf modeling
Digital modeling
23
Baseband equivalent analog modeling
24
This afternoon The protocol/radio interface
Specification (C, Matlab, SDL)
Analog Data Processing
Protocols Control
Behavioral
Digital Data Processing
Behavioral/ Structural
VCC, Opnet, Telelogic, Stateflow
Stateflow Simulink
Matlab, Simulink
Structural
Unicad, Cadence, Synopsys
Spectre
ARMulator,ARM Compiler
Physical
HP EESoft ASITIC Cadence
ARM FPGA Express
Unicad Cadence, Power TimeMill
25
Tuesday morning digital design flow
Specification (C, Matlab, SDL)
Analog Data Processing
Protocols Control
Behavioral
Digital Data Processing
Behavioral/ Structural
VCC, Opnet, Telelogic, Stateflow
Stateflow Simulink
Matlab, Simulink
Structural
Unicad, Cadence, Synopsys
Spectre
ARMulator,ARM Compiler
Physical
HP EESoft ASITIC Cadence
ARM FPGA Express
Unicad Cadence, Power TimeMill
26
Analog Design (Tuesday Morning)
Analog
Digital
cos(wot)
RF input
I (200MS/s)
(
f
2GHz)
c
A/D
Digital
Baseband
Receiver
RF filter
LNA
A/D
Q (200MS/s)
chip boundary
sin(wot)
Crystal
27
Center Activities
Universal Spectrum Sharing
PicoRadio
Applications
Behavioral/Architectural Specification,
Verification And Optimization
Design Tools
Automated Design
Implementation
PicoNode Testbed
BEE Testbed
IC Implementation
28
The BEE - Bigascale Emulation Engine)

• Sufficient processing capability to support real
  time operation of complex baseband algorithms,
  with attached analog frontends
• Arrays of FPGAs and potentially DSPs
• Same input description as chip design
• The goal is to provide a realtime testbed for the
  advanced algorithm development

29
The BEE Hardware (G. Wright)
Programming maintenance I/F
PE
RX control
Radio RX
uP
User I/F
TX control
Radio TX
30
BEE and the BWRC Design Flow
Simulink/Stateflow description
Matlab .mdl files
Custom netlister (preserves hierarchy)
Custom EDIF files
Makefile driven technology-specific mapping
Synthesis, layout, design rule checking
BEE Field Programmable Logic Array
Library module instantiation, synthesis, partition
ing, fitting
Custom ASIC
Code generation, timing verification
DSP code
31
Accomplishments

• System defined which exploits all the signal
  space dimensions
• Time Direct Sequence
• Frequency OFDM
• Space Multi Element Arrays
• First test circuit almost through the automed
  design flow
• First test circuits of .25 micron direct
  conversion analog front end in measurement
• Multistandard radio test chip in testing