Alternative Communication or Ignorance is Bliss

1
Alternative CommunicationorIgnorance is Bliss

• K. Pister
• EECS, BSAC
• UC Berkeley

2
Limits to RF Communication

• 8 GHz (3.5cm)
• 20 W
• 1.5x109 km
• 115 kbps
• -130dbm Rx
• 10-21 J/bit
• kT4x 10-21 J _at_300K
• 5000 3.5cm photons/bit

3
Maxell (Hitachi) RF ID Chip
4
1kbps, 100m, ground

• Sensitivity kBT Df Nf SNRmin
• kBT 1kHz 10x
  limit SNR
• S -174dBm 30 dB 10 dB 10 dB
• -124 dBm
• Path loss 16p2 (d/l)4 /Gant (min1)
• 22dB 40dB log10300 Gant
• 122 dB Gant
• Transmit 1mW, receive 122dBm ? OK
• 1uJ/bit fundamental Tx cost.

5
1kbps, 100m, ideal

• Sensitivity kBT Df Nf SNRmin
• kBT 1kHz at
  limit coding wizards
• S -174dBm 30 dB 0 dB 0 dB
• -134 dBm
• Path loss 16p2 (d/l)2 /Gant (UAV)
• 22dB 20dB log10300 6 (dipole)
• 66 dB
• Transmit 1nW, receive 126dBm ? OK
• 1pJ/bit fundamental Tx cost.

6
1kbps, 100m, Bluetooth

• Sensitivity -75 dBm (standard)
• kBT 1MHz lousy
  radios OK!
• S -174dBm 60 dB 39 dB
• Path loss 10 dB/desk? /wall?
• Transmit 1mW for 1ms
• 1nJ/bit fundamental Tx cost.
• actual Tx, Rx power drain 100mW
• 100nJ/bit, 10s of meters?

7
Hearing Aid Microphones
Knowles Electronics 3024

• Excellent sensitivity
• Low power consumption (20mA _at_ 1.3V)
• Simple electronic interface (on-board amplifier)
• Tiny (2.5mm)3

8
Acoustic Communication
Knowles Electronics 3024
Power in an acoustic wave P P02 S / r
c where P0 is the air pressure S is the surface
area r is the fluid density and c is the speed of
sound
For a 10m sphere at 30dB SPL, P 1.5mW (loud
whisper) For a 1km sphere at 30dB SPL, P 15mW
(crickets) For a 100m sphere at 110dB SPL, P
4kW (Rolling Stones)
9
Optical Communication w/ MEMS

• Passive communication
• Modulated corner cube reflectors
• RAND inspired (Bob Zwirn, R. Steeb, K. Brendley),
  1990/92
• Active communication
• Steered lasers
• DARPA/ISAT 93 (M. Horowitz)

10
Optical Communication
Path loss
0-25
25
Loss (Antenna Gain) Areceiver / (4p
d2) Antenna Gain 4p / q½2
11
Output Power Efficiency
Pout

• RF
• Slope Efficiency
• Linear mod. 10
• GMSK 50
• Poverhead 1-100mW
• Optical
• Slope Efficiency
• lasers 25
• LEDs 80
• Poverhead 1uW-100mW

True Efficiency
Slope Efficiency
Pin
Poverhead
12
Optical Comm Proof of Concept

• Receiver
• 1 lens, std. video CCD
• PCMCIA Frame grabber
• Software decoding

• Transmitter
• Std. Laser pointer
• 2 day life full duty
• 4bps, OOK

13
Video Semaphore Decoding
14
Steered Agile Laser Transceivers (SALT)

• Imaging receiver
• Multiple laser turrets
• 3 axis gyro-based feed-forward control

Co-PI Bernhard Boser
15
SALT Goal

• 1-10 km
• 1-10 Mbps
• ltlt 1W
• 1 cm3

16
1-10 Mbps CMOS imaging receiver
17
Optical Receiver Noise

• Thermal noise from amplifier
• Int2 4kTB/R
• Shot noise Ins2 2 q Id B
• Background light photocurrent
• Signal light photocurrent
• Diode leakage
• Bottom line
• Amplifier dominates for Npixel gt 10100
• 50uW /pixel gives 65dBm noise

18
2D beam scanning
19
6-bit DAC Driving Scanning Mirror

• Open loop control
• Insensitive to disturbance
• Potentially low power

20
8mm3 laser scanner
Two 4-bit mechanical DACs control mirror scan
angles. 6 degrees azimuth, 3 elevation
21
Acquisition

• SALT/SALT (w/ imaging receiver)
• Single scan, hello byte
• Logarithmic acquisition w/variable divergence
• 1ms for short range (100m)
• Mote/mote (omni receivers)
• Two scans
• hello, Im Bob pointing at x,y
• Bob again, at x,y. Ive heard Jan who was
  pointing at i,j and Kris who said
• 1 s ?

22
Maintenance

• SALT/SALT
• Feed-forward on beam steering
• 3 axis gyro (10-10kHz BW)
• Imaging receiver (DC-100kHz?)
• Dithering
• Whats loudest here? here? here?
• Mote/mote (omni receivers)
• Dithering only
• wmax q1/2duty/(tbNpacket)
• gtgt1 rad/sec for Mbps comm

23
Theoretical Performance
5km
Ptotal 50mW Pt 5mW q½ 1mrad ? Gant
71dB BR 5 Mbps
Areceiver 1cm2 Pr 10nW (-50dBm) Ptotal
50uW /pixel SNR 15 dB 10,000 photons/bit
10nJ/bit
24
Theoretical Performance
5m
Ptotal 100uW Pt 10uW q½ 1mrad BR 5 Mbps
Areceiver 0.1mm2 Pr 10nW (-50dBm) Ptotal
50uW SNR 15 dB
20pJ/bit!
25
Theoretical Performance
500km
Ptotal 50mW Pt 5mW q½ 1mrad BR 2 Mbps
Areceiver 1m2 Pr 10nW (-50dBm) Ptotal 50uW
/pixel SNR 17 dB
25nJ/bit!
26
Conclusion

• Grit your teeth and use the radio
• 50uJ/bit 1-10km
• 100nJ/bit 0-50m
• Unless youre lucky enough to have line of sight
• Use optical comm when possible
• 10nJ/bit 1-10km
• 20pJ/bit 0-50m

27
01 Goal
28
RF Sensitivity

• Pn kBT Df Nf
• Sensitivity Pn SNRmin
• e.g. GSM (European cell phone standard), 115kbps
• kBT 200kHz 8x SNR
• S -174dBm 53 dB 9 dB 10 dB
• -102 dBm
• RX power drain 200mW ? 2uJ/bit
• TX power drain 4W ? 40 uJ/bit

29
RF Path Loss

• Isotropic radiator, l/4 dipole
• PrPt / (16p2 (d/l)n)
• Free space n2
• Ground level n27, average 4

30
N4
From Mobile Cellular Telecommunications, W.C.Y.
Lee Pt 10-50W
31
Path Loss

• Like to choose longer wavelength
• Loss (l/d)n
• 916MHz, 30m, ? 92dB power loss
• ? need 92dBm receiver for 1mW xmitter
• ? power!
• Penetration of structures, foliage,
• But
• Antenna efficiency
• Size l/4 _at_ 1GHz 7.5cm

32
RF Sensor Future

• RF tags Sensors
• Ultra Wide Band
• 10ps? digital pulse trains
• LLNL
• 60 GHz
• Major path loss problems
• But oh, the bandwidth!
• MEMS RF components
• Mechanical filters already dominate RF
• Never ever bet against Pisano and Howe