Security optical data storage in Fourier holograms
Wei-Chia Su,
1Yu-Wen Chen,
2,* Yu-Jen Chen,
1Shiuan-Huei Lin,
3and Li-Karn Wang
21Graduate Institute of Photonics, National Changhua University of Education, Changhua 500, Taiwan
2Institute of Photonics Technologies, National Tsing Hua University, Hsinchu 30013, Taiwan
3Department of electrophysics, National Chiao Tung University, Hsinchu 30050, Taiwan
*Corresponding author: [email protected]
Received 7 November 2011; revised 9 January 2012; accepted 9 January 2012; posted 17 January 2012 (Doc. ID 157738); published 15 March 2012
We have proposed and demonstrated a holographic security storage system that is implemented with a shift multiplexing technique. The security function of this storage system is achieved by using a micro-diffuser (MD) for random phase encoding of the reference beams. The apparatus of random phase en-coding in this system offers an additional and flexible function during the recording processes. The system can generate holographic security memory or nonsecurity holographic memory via using the MD or not. The storage capacity and the average signal-to-noise value of the security storage system are16 bits∕μm2and 3.5, respectively. Lateral shifting selectivity in this holographic security storage
system is theoretically analyzed and experimentally investigated. © 2012 Optical Society of America
OCIS codes: 210.2860, 210.0210, 090.0090.
1. Introduction
Volume-holographic storage has received increasing attention owing to its potential high storage capacity and fast access rate [1–3]. A novel technique for holo-graphic storage called collinear holography has been investigated in recent years [4]. This apparatus is implemented with a coaxial optical structure to re-cord holograms. The signal and reference beams in this system can be generated by using a spatial light modulator (SLM), and they are usually distributed on the inner and outer surface region of the SLM. These two beams propagate along the optical axis of the system and pass through a Fourier lens, and then finally they interfere at the focal area of the Fourier lens. A recording material is located at the focal area of the Fourier lens for the holographic re-cording. It has been proven that the collinear holo-graphic storage system performs high potential in miniaturization of the optical architecture and could be compatible with the existing disc storage systems. Shimura et al. have also shown that an additional
random phase encoding in the reference pixels im-proves the imaging performance in a collinear holo-graphic data storage system [5]. Meanwhile, random phase encoding in holographic storage is one attrac-tive and important issue for security storage due to the growing demand for protection of information [6–13]. Random phase encoding generated by using a microdiffuser (MD) has shown great advantages for holographic security data storage owing to their dif-ficulty in duplication [14].
The Fourier-architecture-based holographic sto-rage system with random phase encoding has been studied for a long time [15–18]. In this paper, the tra-ditional Fourier-architecture-based holographic sto-rage system is modified to become a coaxial optical structure. Accordingly, a holographic security sto-rage system based on random phase encoding is pre-sented, and the system is implemented with a shifting multiplexing technique.
As shown in Fig. 1, the MD was placed in the re-ference arm in order to generate a random phase wa-vefront for encryption of the holographic storage. The stored image can be retrieved from the medium if the original MD is placed in the original spatial position to generate the same random phase wavefront for
1559-128X/12/091297-07$15.00/0 © 2012 Optical Society of America
holographic reconstruction. However, the stored im-age can not be retrieved if users lack the original MD. Accordingly, security storage in this system is imple-mented by this MD. If the MD is moved away from the reference arm, the shift multiplexing mechanism for holographic storage can still be performed in this system, and the whole system becomes very similar to a system presented by Steckman et al. [19]. Con-sequently, the characteristic of the MD in this system offers an additional function during the recording process. We can generate a holographic security or nonsecurity memory optionally by moving the MD into the reference arm or moving it away in each ho-lographic recording. As a result, a hoho-lographic secur-ity and nonsecursecur-ity memory both can be generated individually by using our proposed system.
In our previous study, we developed similar techniques to implement encryption-selectable holo-graphic storage systems [20,21]. However, the study in this manuscript is different from our previous pub-lications. In comparison with the system proposed in [20], the holographic security storage proposed in this paper can be accessed without using phase-conjugated readout algorithms. In [20], we have shown that the fidelity of the phase-conjugated refer-ence beam is an important factor that significantly affects the signal-to-noise ratio (SNR) of the recon-structed holograms. In addition, a scheme of encryp-tion-selectable holographic storage in LiNbO3using angular multiplexing based on 90° geometry is de-scribed in [21]. Currently, we find that the study of encryption-selectable holographic storage based on Fourier holograms is still less discussed. Therefore, not only is the holographic security storage system demonstrated in this paper, but the encryption-se-lectable function of the proposed system is also dis-cussed. The effect of the MD on shifting selectivity in this holographic security storage system is analyzed theoretically and experimentally. The experimental results show that storage capacity with16 bits∕μm2 in the security storage system can be obtained and that the average SNR of the retrieved data image is 3.5.
2. Security Holographic Storage System
The experimental apparatus of the holographic se-curity storage system is shown in Fig. 2. We used a diode-pumped, solid-state laser at 532 nm as the light source. The laser beam was collimated and split into two parts, one being the signal beam and the other being the reference beam. The signal beam was incident upon the input image and was then passed through the beam splitter (BS) directly. The reference beam was also incident on the BS, but it reflected from the BS. These two beams propagated along the optical axis of the Fourier lens L2 and then passed through it. The focal length of lens L2 was 50 mm. In the reference arm, a MD was placed at 5 mm in front of the recording medium in order to generate random phase wavefronts, and the MD was fixed at the same position for all the following encryption processes. In this study, handmade ground glass was used as the required MD. The ground glass in the experiment was made by ground-ing a flat glass with Al2O3powders. Therefore, a ran-dom phase distribution on the surface of the ground glass diffuser was obtained. The detailed specifica-tions of the ground glass can be found in our previous work [6]. Based on the coaxial geometry, the signal beam would interfere with the random phase wave-fronts of reference beam in the focal area of the lens L2. A PQ-PMMA material [22] with dimensions of 20 mm × 20 mm × 2 mm was used as the holographic recording medium, and it was located at the focal area of the lens L2. The recording material was mounted upon a two-dimensional translation stage (piezo-motor driven linear stage, Newport) for imple-menting shift multiplexing of holographic storage. Based on the shift multiplexing algorithm, holo-grams were recorded track by track in the horizontal and vertical directions of PQ-PMMA during the recording processes. These holograms between adja-cent tracks should be stored by shifting the recording material with a distance larger than or at least equal to the shifting selectivity of this system. The analysis of the shifting selectivity will be presented in the next section. In the reading processes, a Fourier lens L3 located between recording medium and CCD plane performed an inverse Fourier transform of
Fig. 1. (Color online) Security holographic storage implemented
with a MD. CCD M1 M2 M3 L1 L2 L3 SF HP1 PBS HP2 BS Input image MD H
Fig. 2. Experiment setup for a security collinear holographic sys-tem: SF, spatial filter; L, lens; M, mirror; PBS, polarization beam splitter; HP, half-wave plate; MD, microdiffuser; H, hologram.
the diffraction beam. When the signal beam was blocked, the original reference beam was used as the reading beam and the medium was shifted to the corresponding recording position; the stored im-age could then be perfectly reconstructed on the CCD plane.
3. Lateral Selectivity
Shifting selectivity is an important parameter for af-fecting storage capacity. The lateral shifting selectiv-ity can be analyzed theoretically by using the VOHIL model (the volume hologram being an integrator of the lights emitted from elementary light sources) [23]. As shown in Fig.3, a convergent spherical wave was incident on the MD to generate the random phase wavefront. Thus, we can write the composite wavefront on the hologram plane as
Rwx3; y3 Z d∕2 −d∕2 Zd∕2 −d∕2Aw expjϕx1; y1 × expjkr1dx1dy1; (1)
where d is the dimension of the illumination region of the MD, Awis the amplitude of each spherical wave,
ϕx1; y1 is the initial random phase of each point
source of MD, and r1 x3− x12 y3− y12
z3− z121∕2 is the distance between the MD and
the hologram. The object beam was a plane wave in-cident upon the input image, which was placed aΔxo distance from optical axis to the image center and then passed through the Fourier lens L. Therefore, the interference fringes of the hologram records can be written as
Hx3; y3 jRwx3; y3j2 jIfSxo Δxo; yogj2
R
wx3; y3 · IfSxo Δxo; yog
Rwx3; y3 · IfSxo Δxo; yog; (2)
where If•g represents the Fourier transform func-tion and Sxo; yo is the input image function. During
the reading process, we used a random phase wave-front to read the hologram. The reading wave on the hologram plane can be expressed as
Rrx3; y3 Z d∕2 −d∕2 Zd∕2 −d∕2Ar expjϕx2; y2 × expjkr2dx2dy2; (3)
where Aris the amplitude of reading wave,ϕx2; y2
is the initial random phase of each point source of MD used for encoding the reading wave, and r2
is the distance between the MD and the holo-gram, which can be expressed as r2 x3− x22
y3− y22 z3− z221∕2. For the paraxial condition,
r1≈ z0 and r2≈ z0, where z0 is the distance between
MD and the hologram. When the recording material is only laterally shifted and causes a relative dis-placement of speckle wave with a distance of Δx x2− x1, and Δy y2− y1, we can express the
diffraction as D ∝ ZT∕2 −T∕2Rr· H · exp jk T 2− r dr ∝ ZT∕2 −T∕2 Zl y∕2 −ly∕2 Zl x∕2 −lx∕2 Zd y∕2 −dy∕2 Z d x∕2 −dx∕2
×IfSxo Δxo; yogjArjjAwj × exp j k 2z0x3− x1− Δx 2 y 3− y1− Δy2 x3− x12 y3− y12 × exp jk T 2− r dx1dy1dx3dy3dr; (4) where r z3∕ cos α, T t∕ cos α, α is the propagation
angle of the signal beam, and t is the thickness of hologram. Therefore, the diffraction intensity can be expressed as z0 x3 z3 T y3 d t α ∆ xo Input image L MD x1 z1 y1 Plane wave Plane wave H x2 z2 y2 Writing Reading xo yo r
I ∝ jDj2 ∝ Zl x∕2 −lx∕2 e−j2π x0 λfΔxλz0 x3 dx3 Z d x∕2 −dx∕2 e−j2π −Δx λz0 x1 dx1 Zl y∕2 −ly∕2 e−j2π y0 λfλz0Δy x3 dy3 Zd y∕2 −dy∕2 e−j2π −Δy λz0 y1 dy1 × Zw∕2 −w∕2 Z Sx0 Δx0; y0dx0dy0 Z t 2 cos α −t 2 cos α 1 cosαe jk t 2 cos α e−j2π 1 λ cos α z3 dz3 2 ∝ sinc2dxΔx λz0 sinc2 dyΔy λz0 sinc2 t λ cos2α SΔx0−f Δxz 0 ; − f Δy z0 2; (5)
where f is the focal length of lens L, and w is the di-mension of the illumination region of the pattern. Here we can define the lateral selectivity as the shift deviation for the first zero diffraction obtained from the sinc functions in Eq. (5). The theoretical simula-tion and experimental results are shown in Fig. 4. The parameters used in the theoretical simulation are λ 532 nm, Z0 5 mm, dx 2.5 mm, dy 2 mm, t 2 mm, and f 50 mm. As shown in Fig.5, a chessboard pattern with20 × 20 pixels was used as the input pattern. From our experimental measure-ment results, the lateral selectivity of collinear holo-graphic storage by shifting a single hologram was about only 1.3 and 1.5μm in the horizontal (x) and vertical (y) directions, respectively.
4. Optical Implementation
In this section, a practical holographic security sto-rage system using the shift multiplexing technique was demonstrated. In order to keep a higher SNR of the reconstructed data image, the shift distance for storing next adjacent hologram in this practical demonstration system was set as 5μm in both lateral directions [24]. According to the [24], we chose more than the third Bragg null in order to make sure that the reconstructed data image could keep a higher SNR. In this demonstration, 100 image pages were recorded within this material with 10 rows. Each
row contained 10 holograms, and these 10 holograms were partially spatially overlapped, but the shift se-paration of each adjacent hologram was 5μm. After the recording in the first row was completed, the ma-terial was horizontally shifted to its original position and additionally shifted in a vertical direction with 5 μm. We repeated the row recording process de-scribed above until the whole 100 holograms were stored within these 10 rows. The dimension of each pixel of the chessboard pattern with 20 × 20 pixels was 500 μm × 500 μm. Therefore, the DC term di-mension of the signal beam located at the focal plane was around100 μm × 100 μm. We can find these 100 holograms were partially spatially overlapped. The spatial distributions of these 100 holograms are illu-strated in Fig.6. Figure7(a)to(e)are the 1st, 45th, 50th, 55th, and 100th retrieved images obtained from this system, respectively. Figure 7(f) shows the retrieved images of encryption storage without using the correct MD. The storage capacity of our practical system can be expressed as
N M
Δx · Δy; (6)
where M is the pixel number of the data pages. Ac-cording to Eq. (6), the security storage capacity of our experimental system will reach16 bits∕μm2.
The SNR of each retrieve image that was en-crypted by MD based on the coaxial geometry was also calculated. In our experimental setup, the chess-board pattern with 20 × 20 pixels (500 μm in size) was used as a binary pattern. The magnification of
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Lateral displacement (micrometer)
Normalized diffraction intensity
simulation results(X direction) simulation results(Y direction) measurement results(X direction) measurement results(Y direction)
Fig. 4. (Color online) Theoretical and experimental results of lat-eral selectivity for security collinear holograms.
10 mm
10 mm
our retrieved image system was about 0.19. There-fore, the magnification of the retrieved image on the CCD plane was modulated such that one pixel on the retrieved image corresponded to one superpix-el on the CCD image sensor. One superpixsuperpix-el was com-posed of 20 × 20 camera pixels. For each retrieved image, there were 20 × 20 superpixels used for SNR calculation. Although 20 × 20 camera pixels on the retrieved image represent 1 bit, the edges were left out, so only the central10 × 10 pixels effec-tively within a superpixel was used for calculation of SNR. The SNR formula can be expressed as
SNR μ1− μ0 σ2
1 σ20;
(7) whereμ1,μ0andσ1,σ0are the mean value and devia-tion of the detected energy for one and zero bits, re-spectively [25]. As shown in Fig.8, the average SNR value of 100 pages data of retrieved images was about 3.5.
5. Security-Selectable Storage
In this paper, we extend the concept mentioned above to configure a security-selectable holographic storage
system. We can generate a holographic security or nonsecurity memory optionally by using the same system. A holographic security memory means that the whole stored images in the memory are en-crypted during the recording processes, and a holo-graphic nonsecurity memory means that whole stored images in the memory are not encrypted dur-ing the recorddur-ing processes. If users choose none-ncryption recording for all the stored images in this system, the MD should be moved away from the reference arm and the multiplexing storage me-chanism in our storage system becomes similar to the conventional shift-multiplexed holographic storage system [19]. If users choose encryption recording for all the stored images in this system, we still per-form shift multiplexing for recording, but the MD is moved into the reference arm for each exposure. Thus, a holographic security memory described in Section4can be generated.
In our system design, each security-selectable sto-rage system of volume holography has a unique MD for encryption and decryption. The MD in each sys-tem only can be shifted in horizontal direction but it
Fig. 6. (Color online) Spatial distribution of 100 stored holograms within the recording material. There are a total of 100 holograms in 10 tracks. Each track contains 10 shift-multiplexed holograms.
(a)
(d) (e) (f)
(c) (b)
Fig. 7. (Color online) Retrieved images of the stored patterns in this collinear holographic storage system. (a)–(e) Retrieved images by using the same MD locating at the correct position. (f) Diffrac-tion results of encrypDiffrac-tion storage without using the correct MD.
cannot be removed from the system without destruc-tion. Holographic nonsecurity memory generated in an arbitrarily recording system can be used as a re-movable memory. Stored data in holographic nonse-curity memory can be easily retrieved in another system because only spherical reference wave is re-quired for the readout. Contrary to nonsecurity mem-ory, the readout of holographic security memory generated by an arbitrary recording system in another system is impossible because lack of the ori-ginal encryption key to decrypt the security informa-tion. Encrypted holographic memory only can be accessed in the original recording system. Accord-ingly, each system can generate its own encrypted memory and therefore offers high security for con-tent protection.
In our demonstration of nonsecurity storage, the shift selectivity of shift-multiplexed holographic sto-rage without using a MD in this system is 6μm and 90μm in x and y directions, respectively. However, we have noticed the shifting selectivity in horizontal and vertical directions for nonsecurity storage are quiet different for this conventional shift-multiplexed holo-grams owing to the Bragg degeneration in the verti-cal direction [6,21,26,27]. Accordingly, 15 image pages were recorded within this material with 3 rows. Each row contained 5 holograms, and these 5 holograms were partially spatially overlapped, but the shift separation of each adjacent hologram was 6 μm. In addition, the shift separation of each adjacent row was 90 μm. After the recording in the first row was completed, we repeated the row record-ing process described above until all 15 holograms were stored within these 3 rows. Figures9(a)and(b)
are the retrieved images of the nonencryption sto-rage. The average SNR value of 15 pages data of re-trieved images was about 4.2.
6. Conclusion
In this paper, we have proposed a holographic secur-ity storage system by using a shift multiplexing tech-nique. A MD played an important role in encryption and decryption of holographic storage and recon-struction. Shift selectivity of the proposed security holographic storage system was analyzed. The ex-periments demonstrate a security holographic sto-rage of 100 holograms within this system, and the storage capacity and the average SNR value of the presented security collinear holographic storage sys-tem were16 bits∕μm2and 3.5, respectively. Selective
encryption storage also can be achieved by moving the MD into the reference arm or moving it away dur-ing the recorddur-ing process. Nonsecurity holographic storage using the same system was also presented. This work was supported by the National Science Council of Taiwan under contract no. NSC 97-2221-E-108-002-MY3
Reference
1. H. J. Caulfield, D. Psaltis, and G. Sincerbox, Holographic Data Storage (Springer-Verlag, 2000).
2. S. S. Orlov, W. Phillips, E. Bjornson, Y. Takashima, P. Sundaram, L. Hesselink, R. Okas, D. Kwan, and R. Snyder, “High-transfer-rate high-capacity holographic disk data-storage system,” Appl. Opt. 43, 4902–4914 (2004).
3. G. W. Burr, C. M. Jefferson, H. Coufal, M. Jurich, J. A. Hoffnagle, R. M. Macfarlane, and R. M. Shelby, “Volume holographic data storage at an areal density of 250 gigapixels∕in:2,” Opt. Lett. 26, 444–446 (2001).
4. H. Horimai, X. Tan, and J. Li,“Collinear holography,” Appl. Opt.44, 2575–2579 (2005).
5. T. Shimura, S. Ichimura, R. Fujimura, K. Kuroda, X. D. Tan, and H. Horimai,“Analysis of a collinear holographic storage
system: introduction of pixel spread function,” Opt. Lett.
31, 1208–1210 (2006).
6. C. C. Sun and W. C. Su,“Three-dimensional shifting
selec-tively of random phase encoding in volume holograms,” Appl. Opt.40, 1253–1260 (2001).
7. C. Denz, K.-O. Muller, F. Visinka, and T. Tschudi,“Digital volume holographic data storage using phase-coded
holo-graphic memory system,” Proc. SPIE 3802, 142–147
(1999).
8. J. F. Heanue, M. C. Bashaw, and L. Hesselink,“Encrypted ho-lographic data storage based on orthogonal-phase-code multi-plexing,” Appl. Opt. 34, 6012–6015 (1995).
9. W. C. Su and C. H. Lin, “Three-dimensional shifting
selectivity of decryption phase mask in double random phase
encoding holographic memory,” Opt. Commun. 241, 29–41
(2004).
10. O. Matoba and B. Javidi,“Encrypted optical memory system
using three-dimentional keys in the Fresnel domain,” Opt.
Lett.24, 762–764 (1999).
11. O. Matoba and B. Javidi, “Encrypted optical storage with
angular multiplexing,” Appl. Opt. 38, 7288–7293
(1999).
12. G. Unnikrishnan, J. Joseph, and K. Singh,“Optical encryption system that uses phase conjugation in a photorefractive crys-tal,” Appl. Opt. 37, 8181–8186 (1998).
13. O. Matoba, Y. Yokohama, M. Miura, K. Nitta, and T.
Yoshi-mura,“Reflection-type holographic disk memory with random
phase shift multiplexing,” Appl. Opt. 45, 3270–3274 (2006). 14. W. C. Su, C. C. Sun, Y. C. Chen, and Y. Ouyang,“Duplication of
phase key for random-phase-encrypted volume holograms,”
Appl. Opt.43, 1728–1733 (2004).
15. Y. Takeda, Y. Oshida, and Y. Miyamura, “Random phase
shifters for Fourier transformed holograms,” Appl. Opt. 11, 818–822 (1972).
16. C. B. Burckhardt,“Use of a random phase mask for the record-ing of Fourier transform holograms of data masks,” Appl. Opt. 9, 695–700 (1970).
17. O. Matoba and B. Javidi, “Secure holographic memory by
double-random polarization encryption,” Appl. Opt. 43,
2915–2919 (2004).
18. M. Yamazaki and J. Ohtsubo, “Optimization of encrypted
holograms in optical security systems,” Opt. Eng. 40,
132–137 (2001).
19. G. J. Steckman, A. Pu, and D. Psaltis,“Storage density of
shift-multiplexed holographic memory,” Appl. Opt. 40, 3387–3394
(2001).
20. C. C. Sun, W. C. Su, B. Wang, and A. E. T. Chiou,“Lateral shift-ing sensitivity of a ground glass for holographic encryption (b)
(a)
Fig. 9. (Color online) (a), (b) Retrieved images of nonencryption storage implemented with the same system.
and multiplexing using phase conjugate readout algorithm,”
Opt. Commun.191, 209–224 (2001).
21. W. C. Su and C. C. Sun,“Encryption-selectable holographic
storage in LiNbO3 with angle multiplexing,” Microw. Opt.
Technol. Lett.42, 227–230 (2004).
22. S. H. Lin, K. Y. Hsu, W. Z. Chen, and W. T. Whang,
“Experi-mental characterization of phenanthrenequinone-doped poly (methyl methacrylate) photopolymer for volume holographic storage,” Opt. Eng. 42, 1390–1396 (2003).
23. C. C. Sun, “A simplified model for diffraction analysis of
volume holograms,” Opt. Eng. 42, 1184–1185 (2003).
24. G. Barbastathis, M. Levene, and D. Psaltis,“Shift multiplex-ing with spherical reference waves,” Appl. Opt. 35, 2403–2417 (1996).
25. C. Moser and D. Psaltis,“Holographic memory with localized
recording,” Appl. Opt. 40, 3909–3914 (2001).
26. C. C. Sun, M. S. Tsaur, W. C. Su, B. Wang, and E. T. Chiou, “Two-dimensional shifting tolerance of a volume-holographic correlator,” Appl. Opt. 38, 4316–4324 (1999).
27. C. C. Sun, W. C. Su, B. Wang, and Y. OuYang,“Diffraction
se-lectivity of holograms with random phase encoding,” Opt.
