In this paper we present the HiPAR-DSP 16, a parallel and programmable processor architecture which is adapted to the demands of SAR image processing. TO provide a high performance, the HiPAR-DSP 16 features an array of 16 parallel processing units. Each of these processing units can process up to 3 instructions per clock cycle. Efficient data exchange between the processing units can be done by a shared memory with concurrent access. The HiPAR-DSP 16 is able to perform a 4096 samples complex FFT in 154 microsecond(s) and a compete (omega) k SAR processing algorithm on 4k range line with a PRF of more than 200 Hz in real-time. This shows the high capability of the HiPAR-DSP 16 for onboard real-time SAR systems.

We have performed chirp measurements for four wave mixing (FWM) in a semiconductor optical amplifier using a CW pump signal and a pulsed probe signal. The FWM chirp, which depends on the chirp of the probe signal, pump power, and pump wavelength, has been measured. A calculation of the chirp under FWM has ben carried out. The results of the numerical simulation are in agreement with the experimental results.

A tunable laser operating near 944nm is important for micro LIDAR systems. In this work, we have carried out a systematic study of Nd and Yb doped fiber glass materials which can be used for this purpose. The emission cross- sections of these materials at 944nm have been calculated. The wavelength dependence of emission of these materials has been studied. Among the silica based material Nd doped silica fiber glass has the largest cross section for 944 nm laser emission. The co-dopants reduce this cross section. Some Nd doped non-silica materials have higher stimulated emission cross section than silica based materials.

Wide bandwidth and high-resolution analog to digital converters (ADCs) are required for the next generation of sensor systems. Progress at advancing the electronic ADC modules has been very slow due in large part to the difficulties in fabricating the electronic circuitry required for very high resolution and high sampling rate converters. It is anticipated that the use of photonic ADCs will far surpass the performance of electronic ADCs in terms of both sampling speed and resolution. We have recently designed a novel photonic ADC module that incorporates the use of semiconductor saturable absorbers to perform the data quantization at speeds in the tens of GHz regime. Experimental material characterization results including the nonlinear transmission and the recovery time of the semiconductor saturable absorbers use din the data conversion process will be presented. Different material parameters will also be analyzed including the effects of low temperature growth, band-edge position, and strain on these material properties.

This paper provides a comprehensive understanding of a novel photonic high-speed analog-to-digital converter (ADC) approach and its anticipated performance. Compelling analytic models, experimental dat, and simulation results will convince the reader that the system will achieve the goal of developing an ADC with a resolution on the order of 12-bits and with conversion speeds in excess of 10 Gsps.

Current sate-of-the-art electronic analog-to-digital converters (ADCs) operating at multi gigahertz sampling frequencies are known to exhibit fairly limited resolution. These high-frequency restrictions stem primarily from the response time of the constituent transistors that make up the ADCs comparators. In an effort to improve the resolution of ADCs operating at ultrahigh sampling frequencies, several areas of investigation are currently underway regarding the capabilities of hybrid optoelectronic systems. High-power optical pulses can be used as sampling windows and high- bandwidth electro-optic modulators as voltage-to-intensity transducers to provide a means for digitizing ultrafast voltage waveforms with much greater accuracy than conventional ADCs. When optical sampling is employed, the primary limiting factors determining ADC conversion accuracy becomes the noise in the sampling pulse train and the extent of the sampling time. Detrimental pulse train noise is associated with either phase modulation or amplitude modulation, and recent measurements of AM and residual PM noise on our 10 Ghz ring laser show the best results to date for an actively-mode-locked semiconductor diode system. Carrier offset integration bands extending form 10 Hz to 10 MHz exhibit RMS levels of AM and PM noise as low as 0.12 percent and 43 fs, respectively. In addition, linear dispersion compensation has successfully reduced the optical pulsewidth from 13 ps to 1.2 ps. Based on these experimental numbers, this laser could form the front end for an optoelectronic ADC capable of a theoretical resolution as high as 8.6 bits.

There is demand for high-sped all-optical networks for the next generation internet that can transport the data header and packet of information at rates of between 40-100 Gb/s. Such networks will require high bandwidth and high-speed data transport. DWFM has been proposed as a viable scheme to implement such networks. Recently we reported the generation of optical subcarrier frequencies having bandwidth of the order of 2.5-3 terahertz. We prose a scheme for the design of high-density optical networks, in which the header is carried over the subcarrier frequencies and the packets are carried over the optical wavelengths. This scheme has many advantages, for example, it can allow for separate processing of header and packet, as well as provide higher bandwidth and high-speed data transport. We shall discuss the generation scheme for the terahertz optical subcarriers, a modulation scheme for these carriers, and how they are multiplexed in an all-optical network architecture.

A novel approach to residual jitter measurement examines the intensity cross correlation generated by two optical pulses with various relative delays. A relative delay of 25 pulses produces a residual jitter value of 26 fsec RMS for a 10 GHZ actively mode-locked ring laser. The phase noise measurement carried out to the Nyquist frequency offset gives 47 fsec RMS pulse-to-pulse timing jitter. The field correlation measurement obtains a 10 asec RMS pulse-to-pulse optical carrier jitter.

We present a preliminary study of all-optical switching devices based on the photo and electrochromic behavior of sputtered and sol-gel based tungsten oxide films. While many questions remain unanswered these films show considerable promise for meeting a variety of switching applications. These switching applications complement the optical memory applications we have previously demonstrated. In the present study we have used 488 nm control light to switch 632 nm data light. The films range from about 100 nm to 900 nm in thickness and can be deposited on various substrates. The switching was characterized using single layer films. The switching time is currently submillisecond with incident control powers of just a few tens of milliwatts of control light using a spot size of about ten microns in diameter.

Colossal Storage Inc. has patents on new ways of non-contact reading and writing with non destructive reading of information to a ferroelectric molecule. These methods will be used to develop the worlds first 2D/3D area/ volume holographic mass storage device. US Patents, number 6,028,835 2/00 and number 6,046,973 4/00 for an integrated read/write head for ferroelectric optical media.

We demonstrate an optical transmission link capable of 50- Gb/s through 500 meters of multimode fiber. The aggregate bit rate is carried by 10 parallel wavelength channels, where the individual channel speed is 5-Gb/s. The source of the optical signal is an novel hybrid OTDM-WDM system based on a multi wavelength mode-locked semiconductor diode laser. The performance of the link is evaluated using bit-error rate (BER) analysis. Eye diagrams of the received optical signal yield a BER of 5 X 10^-10. Spectral and temporal traces show robust pulse transmission and the potential for further increase in the aggregate bit rate.

We describe a novel technique for actively mode-locking laser pulses with a repetition rate ranging between 3-5 GHz. The pulses generated are nearly transformed limited with a pulse width of 72ps. In our experiment, a synthesized sweep generator drove a pigtailed laser diode with a bandwidth of 2.5 GHz with a center frequency that was tunable between 6 and 13 GHz. The frequency span of the sweep was 2.5 GHz, which is exactly equal to the axial mode spacing of the laser diode. Active mode-locking of the laser was achieved with a start and end sweep frequency of 4740 MHz and 20 GHz, respectively. We found the mode-locked pulses to be highly stable with a repetition rate that is tunable. Tunability was accomplished by keeping the center frequency and span fixed at 11.87 GHz and 14.26 GHz, and tuning the start and end sweep frequencies over a range of 5GHz to 20GHz.

The paper is to introduce the principles of vector optical heterodyning principally enable for new aerospace sensor systems. This approach comes form a new adjoining coherent method of elastic light scattering, which was earlier experimentally provided for the specific small-size experiments of dynamics light scattering. The theory of vector optical heterodyning for remote sizing and velocimetry is presented. The method is available to advance within the technique for coherent signal processing, which appears to be available by beating several scattered vector waves within a square-low photo-detector by their registration. While the classical passive methods of optical beating deals with the problem to study a scalar optical signal that is due to elastic scattering with a Doppler shift and a single narrow-band laser beam, the new method deals with two vector waves simultaneously undergo scattering by a target. In case one of the vector waves has strong periodical modulation of direction of polarization. Coherent Light Beating Scattering occurs, which units elastic Light Scattering and Coherent non-linear Beating within the square-low photo-detector. As a result, the vector optical signal provides the information about both size and velocity of a remote scattering object, which is not available for well-known Doppler methods with scalar optical heterodyning.

We proposed a new spatio-angularly multiplexed holographic memory system using moving windows and double-focusing lens, which can eliminate crosstalk due to two neighboring moving windows in the vertical direction of the conventional moving window holographic memory system, and demonstrated its feasibility through optical experiments.

We relate quantum mechanics to the concept of an experimental trial, borrowed from statistics, in order to make clear that devices designed using quantum-mechanical equations require, in effect, management by a classical digital process-control computer (CPC). Energy dissipation essential for tolerance to component imperfections in classical digital computers is shown not to mix with quantum superposition expressed by the quantum dynamics of a Schrodinger equation. This precludes so-called quantum computers from producing super-linear speed-up at a scale large enough to be useful, and at the same clarifies a conceptual separation between digital computing and devices exhibiting quantum phase sensitivity. Quantum effects, such as random-number generation and photon correlations, are know to be useful and show further great potential for devices for computational and other purpose. Several such quantum-based devices are discussed that require or can be made more effective by bridging between these devices and CPCs that manage them. While the interdisciplinary promise of 'quantum computing' has faded, the area of quantum-to- classical bridging, once liberated from the misleading goal of super-linear speed-up, invites vigorous study and invention.

We study the Fourier harmonic analysis of a functions on discrete 1D and nD Heisenberg-Weyl groups HW₃ and HW_2n+1, where K equals GF(2), GF(2^m), GF(p), GF(p^m) are the Galois fields, and develop fast quantum Fourier- Heisenberg-Weyl transforms on this groups.

Fast Classical and quantum algorithms are introduced for a wide class of non-separable nD discrete unitary K- transforms(DKT)K_Nn. They require a number of 1D DKT K_n smaller than in the Cooley-Tukey radix-p FFT-type approach. The method utilizes a decomposition of the nDK- transform into a product of original nD discrete Radon Transform and of a family parallel/independ 1DK-transforms. If the nDK-transform has a separable kernel, that again in this case our approach leads to decrease of multiplicative complexity by factor of n compared to the tow/column separable Cooley-Tukey p-radix approach.

We derive a simple integral representation and the corresponding Rodrigues-type difference formula for the continuous q-Hermite polynomials of Rogers. As a consequence, this also yields the appropriate formula for the Rogers-Szego and Stieltjes-Wigert polynomials.

Many approaches have been proposed towards the realization of practical quantum information processing. They include laser manipulation of the energy states in atomic or ion traps, nuclear magnetic resonance in molecules, quantum computation in optical lattices, and in Bose-Einstein condensates. For any of these system to be viable quantum computation, it should have stable and long-lived internal states. In addition it should be possible to quantum- mechanically entangle two or more qubits and the entangled states should be capable of non-local realism phenomena. The ground state and metastable states of low Z atoms especially the single and two-electron atoms and their ions are the most commonly used qubits for quantum information processing systems. Efforts to exhibit these phenomena in high Z elements has been much more challenging. In part because of the high energy needed to ionize the high Z elements into effective one- or two-electron system suitable for quantum information processing. Furthermore the energy difference between two stable states of the High Z elements that are possible qubits states turn to be very high. Preferably the energy difference of two qubits levels should be in the meV to a few tens of electron volts for practical realization of a quantum computing system. In general any two stable terms of an atomic or molecular system with low to moderate energy separation, that do not violate term rules for selection and symmetry, may be viable qubits states.

After summarizing a recent calculation of the maximum Renyi information loss from a POVM quantum cryptographic receiver to a general unitary probe, we calculate the least effective inconclusive-rate monitoring threshold for the receiver.

In this paper a binary image encryption technique and decryption system based on a joint transform correlation are proposed. IN this method, an encrypted image is obtained by multiplying a phase encoded original binary image with a random phase. A Fourier transform of the encrypted, image is used as the encrypted data and a Fourier transform of the random phase is used as the key code. For decryption, the encrypted data is used for one half of the joint input plane, while the key code is used for the other half. After the joint input plane is inverse Fourier transformed, the original binary image can then be reconstructed on a square law device, such as a CCD camera. The proposed encryption technique does not suffer from strong auto-correlation terms appearing in the output plane. In addition, the reconstructed data can be directly transmitted to a digital system for real-time processing. Based on computer simulations, the proposed encryption technique and decoding system were demonstrated as adequate for optical security applications.

In this paper, we proposed a new visual cryptography scheme based on optical interference that can improve the contrast and signal to noise ratio of reconstructed images when compared to conventional visual cryptography methods. The binary image being encrypted is divided into any number of n slides. For encryption, randomly independent keys are generated along with another random key based on a XOR process of random keys. The XOR process between each divided image and each random key produces the encryption of n encrypted images. These encrypted images are then used to make encrypted binary phase masks. For decryption, the phase masks are placed on the paths of a Mach-Zehnder interferometer.

This paper considers the behavior of a model persistent current qubit in the presence of a time-dependent electromagnetic field. A semi-classical approximation for the electromagnetic field is used to solve the time- dependent Schrodinger equation (TDSE) for the qubit, which is treated as a macroscopic quantum object. The qubit is describe3d by a Hamiltonian involving the enclosed magnetic flux (Phi) and the electric displacement flux Q, which obey the quantum mechanical commutation relation. The paper includes a brief summary of recent work on quantum mechanical coherence in persistent current circuits, and the solution of the TDSE in superconducting rings. Of particular interest is the emergence of strongly non-perturbative behavior that corresponds to transitions between the energy levels of the qubit. These transitions are due to the strong coupling between the electromagnetic fields and the superconducting condensate and can appear at frequencies predicted by conventional methods based on perturbations around the energy eigenstate of the time-independent system. The relevance of these non-perturbative processes to the operation of quantum logic gates based on superconducting circuits and the effect of the resultant non linearities on the environmental degrees of freedom coupled to the qubit are considered.

The Fresnel zone plate lens was invented and developed for optical frequencies. However, fabrication difficulties at the short optical wavelengths have prevented obtain good efficiencies. At longer microwave or millimeter-wavelengths fabrication is easier and phase correcting zone plate antennas have been used to obtain good efficiencies. This paper describes a new type of phase correcting zone plate having even better efficiency, namely a diffraction efficiency of 99 percent compared to a true lens, and an overall efficiency much better than a true lens. For the usual zone plate antenna employed at microwave or millimeter wavelengths, path length adjustment is accomplished by cutting different depths in a dielectric plate or by using two or more dielectrics having different dielectric constants. The new design uses a tilted cut in a dielectric plate, which more accurately matches the shape of a true lens and produces much lower phase error. The construction is still near and can be made for example, by a milling machine with a tilted bit. For a circular zone plate, the lens is a stepped conical or tapered shape. Because the phase steps are small, the far-field antenna pattern is excellent and sidelobe-levels are very low. Analysis of typical configurations will be given, showing that phase errors are small, lower than those for an eighth-wave corrected phase zone plate.