10MSST-2002, April 15-18 1Irvine, CaliforniaMetroLaser
HighHigh--density Holographic density Holographic Data Storage with Random Data Storage with Random Encoded Reference BeamEncoded Reference Beam
Vladimir B. MarkovMetroLaser, Inc.
18010 Skypark Circle, Irvine, CA [email protected]
10MSST-2002, April 15-18 2Irvine, CaliforniaMetroLaser
Outline
• Motivation• Outline of Theory• System Design• Results from a Shift Selectivity• Conclusions
10MSST-2002, April 15-18 3Irvine, CaliforniaMetroLaser
Motivation
Holographic memory offers:
• bit storage density of the order of 1012/cm3
• parallel access and parallel data processing
• high retrieval rate
• solid-state configuration
10MSST-2002, April 15-18 4Irvine, CaliforniaMetroLaser
PrinciplesPrinciples
• Selective properties of volume hologram• Volume holograms with amplitude-phase
modulated reference beam and theirselective properties
• Solid-state configuration with randomreference beam
• Selective properties of volume hologram• Volume holograms with amplitude-phase
modulated reference beam and theirselective properties
• Solid-state configuration with randomreference beam
10MSST-2002, April 15-18 5Irvine, CaliforniaMetroLaser
Angular Bragg SelectivityAngular Bragg Selectivity
Non-Dispersion Plane Selectivity
y
x
VolumeGrating
oRθ
ϕ
1e-005
0.0001
0.001
0.01
0.1
1
Dif
frac
tion
effi
cien
cy
-8 -4 0 4 8Angular Dev. (a.u.)
Dispersion PlaneSelectivity
10MSST-2002, April 15-18 6Irvine, CaliforniaMetroLaser
ØReference Beam RandomAmplitude-Phase Encoding:
ünew type of Spatial & Angular(isotropic) selectivity;üsolid-state architecture - no moving partsü secure data access
ØReference Beam RandomAmplitude-Phase Encoding:
ünew type of Spatial & Angular(isotropic) selectivity;üsolid-state architecture - no moving partsü secure data access
ØAngular and Spectral Braggselectivity results in:ünon-isotropic diffraction at off-Bragg tuning üincremental noiseüinsecure data accessürequire moving parts.
ØAngular and Spectral Braggselectivity results in:ünon-isotropic diffraction at off-Bragg tuning üincremental noiseüinsecure data accessürequire moving parts.
Angular-spectral selectivity ofvolume hologram and random encodingof reference beam are used as basic mechanisms for data multiplexing
10MSST-2002, April 15-18 7Irvine, CaliforniaMetroLaser
Random APM volume hologram - RecordingRandom APM volume Random APM volume hologram hologram -- RecordingRecording
y
x
zT
∆⊥xRo(r)
<σ⊥> = 1.22 λD/dL
K SSo(r)
S(r)
Initial position of speckle
10MSST-2002, April 15-18 8Irvine, CaliforniaMetroLaser
y
x
zT
Initial position of speckle
Ro(r)
<σ⊥> = 1.22 λD/dL
K S
S(r)
Random APM volume hologram - Reconstruction
Random APM volume Random APM volume hologram hologram -- ReconstructionReconstruction
10MSST-2002, April 15-18 9Irvine, CaliforniaMetroLaser
Shiftdirection
y
∆⊥y
x
zT
∆⊥
Displaced speckleposition
Initial speckleposition
∆⊥xRo(r)
<σ⊥> = 1.22 λD/dL
S
S(r)
K
Random APM volume hologram - Reconstruction
Random APM volume Random APM volume hologram hologram -- ReconstructionReconstruction
10MSST-2002, April 15-18 10Irvine, CaliforniaMetroLaser
Basic results of the Basic results of the analysisanalysis
Where Ro(q,z)R*(q,z) is spatial correlation function of a random amplitude-phase modulated (speckle) field:Where RWhere Roo(q,z)R(q,z)R*(q,z) *(q,z) is spatial correlation function of is spatial correlation function of a random amplitudea random amplitude--phase modulated (speckle) field:phase modulated (speckle) field:
∫ ∫∞+
∞−⊥
⊥⊥⊥
∆−×
∆=∆ ,qdq
znik
exp)q(Pz2nik
exp)'z,(C 2
eff
o2L
eff
2o rrr
rr
The diffracted field amplitude:The diffracted field amplitude:The diffracted field amplitude:
∫∫ ∫=S
T
oso dqdzzqRzqRikzqS ')',()',(]sinexp[)',( * rrr
0θ
10MSST-2002, April 15-18 11Irvine, CaliforniaMetroLaser
Mirror
Computer/Data Base
Controller
yx
z
TranslationStage
Iris
PolarizingFilter
NDFilter
Lens
PhotoDetector
CMOS
Modulator CondenserLensPre-modulator
BeamSplitter
Shutter1
Shutter2
SLM ImagingLens1
ImagingLens2
SpatialFilter
Lens
Experimental Setup for Random APE Holographic Memory
10MSST-2002, April 15-18 12Irvine, CaliforniaMetroLaser
Laboratory setup for Laboratory setup for APM hologram APM hologram
10MSST-2002, April 15-18 13Irvine, CaliforniaMetroLaser
SPECKLE SHIFT SPECKLE SHIFT SELECTIVITYSELECTIVITY
0.0 4.0 8.0 12.0 16.0 20.0X-SHIFT (µm)
0.00
0.20
0.40
0.60
0.80
1.00In
ten
sity
IN
D
<ε⊥> ~ 12.0 µm<ε⊥> ~ 8.0 µm<ε⊥> ~ 6.0 µm
XX--SHIFT SELECTIVITYSHIFT SELECTIVITY(experiment)(experiment)
<ε⊥> - average speckle size
10MSST-2002, April 15-18 14Irvine, CaliforniaMetroLaser
X-Y Speckle-Shift SelectivitySpeckle-Shift Selectivity is perfectly symmetric in both X and Y directions and the retrieved signal intensity decreases with ∆⊥ in almost 3 orders of the magnitude with no side-lobes. This promises low cross-talk and a high level of security.
TheoryTheory
0.00 1.00 2.00 3.00 4.00SHIFT ∆⊥x,y(µm)
0.001
0.010
0.100
1.000
DIF
FRA
CTE
D B
EA
M I
NTE
NSI
TY I ∆
⊥(a
.u.)
∆⊥X
∆⊥Y
ExperimentExperiment
10MSST-2002, April 15-18 15Irvine, CaliforniaMetroLaser
Image Characteristics@ Spatial Shift
Spatial Shift doesn’t introduce any side effects on the reconstructed image quality beside the intensity decrease. The phase distribution remains invariable with ∆⊥
∆⊥ = − 0.5 µm
∆⊥ = 0.0 µm
∆⊥ = 0.5 µm
∆⊥ = 1.5 µm
10MSST-2002, April 15-18 16Irvine, CaliforniaMetroLaser
0.00 5.00 10.00 15.00 20.00 25.00
Shift ∆⊥(µm)
0.00
0.20
0.40
0.60
0.80
1.00
Nor
mal
ized
Inte
nsity
I DRecordingmedium
Recordingmedium
OpticsOptics
Random APMRandom APM
Pre-encoderPre-encoder
DetectorDetector
Realization of Solid-StateData Storage ConfigurationRealization of SolidRealization of Solid--StateState
Data Storage ConfigurationData Storage Configuration
Principal setupPrincipal setup
10MSST-2002, April 15-18 17Irvine, CaliforniaMetroLaser
Solid-State technique validation
Solid-State technique validation
0 0.2 0.4 0.6 0.8 1
Relative shif t ∆⊥/<σ⊥>
0
0.2
0.4
0.6
0.8
1
Inte
nsity
I NDDecorrelation with:
• Pre-encoder spatial variation(shift or rotation)
• Reference beam spatial steering• Beam angular steering with deflector
• Encoder (or pre-encoder)rotation
Decorrelation with:
• Pre-encoder spatial variation(shift or rotation)
• Reference beam spatial steering• Beam angular steering with deflector
• Encoder (or pre-encoder)rotation
10MSST-2002, April 15-18 18Irvine, CaliforniaMetroLaser
Page encoding and data recallPage encoding and data recall
Recording
Reconstruction
10MSST-2002, April 15-18 19Irvine, CaliforniaMetroLaser
Data Recall SequenceData Recall Sequence
10MSST-2002, April 15-18 20Irvine, CaliforniaMetroLaser
Holographic Memory Module architecture with solid-state
configuration
APE-SLM
RecordingMedium
Data-SLM
Receiver
Laser
MemoryErasure
Anticipated HMMparameters:
¬Capacity - 1011 b¬Trans. rate - 1 Gb/sec¬Size - < 0.4 ft3
¬Weight - <1.5 kg¬Power cons.- < 50 W
Anticipated HMMparameters:
¬Capacity - 1011 b¬Trans. rate - 1 Gb/sec¬Size - < 0.4 ft3
¬Weight - <1.5 kg¬Power cons.- < 50 W
10MSST-2002, April 15-18 21Irvine, CaliforniaMetroLaser
Conclusion
• High- density holographic data storage is demonstrated with random encoded reference beam
• Parallel recording and retrieval• Optical memory in solid-state configuring
Acknowledgment This work was supported in part through the SBIR projects with NASA and DOE
• High- density holographic data storage is demonstrated with random encoded reference beam
• Parallel recording and retrieval• Optical memory in solid-state configuring
Acknowledgment This work was supported in part through the SBIR projects with NASA and DOE