Parallel geostatistics for sparse and dense datasets

Very large spatially-referenced datasets, for example, those derived from satellite-based sensors which sample across the globe or large monitoring networks of individual sensors, are becoming increasingly common and more widely available for use in environmental decision making. In large or dense sensor networks, huge quantities of data can be collected over small time periods. In many applications the generation of maps, or predictions at specific locations, from the data in (near) real-time is crucial. Geostatistical operations such as interpolation are vital in this map-generation process and in emergency situations, the resulting predictions need to be available almost instantly, so that decision makers can make informed decisions and define risk and evacuation zones. It is also helpful when analysing data in less time critical applications, for example when interacting directly with the data for exploratory analysis, that the algorithms are responsive within a reasonable time frame. Performing geostatistical analysis on such large spatial datasets can present a number of problems, particularly in the case where maximum likelihood. Although the storage requirements only scale linearly with the number of observations in the dataset, the computational complexity in terms of memory and speed, scale quadratically and cubically respectively. Most modern commodity hardware has at least 2 processor cores if not more. Other mechanisms for allowing parallel computation such as Grid based systems are also becoming increasingly commonly available. However, currently there seems to be little interest in exploiting this extra processing power within the context of geostatistics. In this paper we review the existing parallel approaches for geostatistics. By recognising that diffeerent natural parallelisms exist and can be exploited depending on whether the dataset is sparsely or densely sampled with respect to the range of variation, we introduce two contrasting novel implementations of parallel algorithms based on approximating the data likelihood extending the methods of Vecchia [1988] and Tresp [2000]. Using parallel maximum likelihood variogram estimation and parallel prediction algorithms we show that computational time can be significantly reduced. We demonstrate this with both sparsely sampled data and densely sampled data on a variety of architectures ranging from the common dual core processor, found in many modern desktop computers, to large multi-node super computers. To highlight the strengths and weaknesses of the diffeerent methods we employ synthetic data sets and go on to show how the methods allow maximum likelihood based inference on the exhaustive Walker Lake data set.

Simultaneous demultiplexing, data regeneration, and clock recovery with a single semiconductor optical amplifier-based nonlinear-optical loop mirror

We demonstrate simultaneous demultiplexing, data regeneration and clock recovery at 10Gbits/s, using a single semiconductor optical amplifier–based nonlinear-optical loop mirror in a phase-locked loop configuration.

Simultaneous clock recovery and data regeneration using a nonlinear optical loop mirror as an all-optical mixer

High-speed optical clock recovery, demultiplexing and data regeneration will be integral parts of any future photonic network based on high bit-rate OTDM. Much research has been conducted on devices that perform these functions, however to date each process has been demonstrated independently. A very promising method of all-optical switching is that of a semiconductor optical amplifier-based nonlinear optical loop mirror (SOA-NOLM). This has various advantages compared with the standard fiber NOLM, most notably low switching power, compact size and stability. We use the SOA-NOLM as an all-optical mixer in a classical phase-locked loop arrangement to achieve optical clock recovery, while at the same time achieving data regeneration in a single compact device

Dielectric and electro-optical properties of some cyanobiphenyl liquid-crystals

A variety of methods have been reviewed for obtaining parallel or perpendicular alignment in liquid-crystal cells. Some of these methods have been selected and developed and were used in polarised spectroscopy, dielectric and electro-optic studies. Also, novel dielectric and electro-optic cells were constructed for use over a range of temperature. Dielectric response of thin layers of E7 and E8 (eutectic mixture liquid-crystals) have been measured in the frequency range (12 Hz-100 kHz) and over a range of temperature (183-337K). Dielectric spectra were also obtained for supercooled E7 and E8 in the Hz and kHz range. When the measuring electric field was parallel to the nematic director, one loss peak (low-frequency relaxation process) was observed for E7 and for E8, that exhibits a Debye-type behaviour in the supercooled systems. When the measuring electric field was perpendicular to the nematic director, two resolved dielectric processes have been observed. The phase transitions, effective molecular polarisabilities, anisotropy of polarisabilities and order parameters of three liquid crystal homologs (5CB, 6CB, and 7CB), 60CB and three eutectic nematic mixtures E7, E8, and E607 were calculated using optical and density data measured at several temperatures. The order parameters calculated using the different methods of Vuks, Neugebauer, Saupe-Maier, and Palffy-Muhoray are nearly the same for the liquid crystals considered in the present study. Also, the interrelationship between density and refractive index and the molecular structure of these liquid crystals were established. Accurate dielectric and dipole results of a range of liquid-crystal forming molecules at several temperatures have reported. The role of the cyano-end group, biphenyl core, and flexible tail in molecular association, were investigated using the dielectric method for some molecules which have a structural relationship to the nematogens. Analysis of the dielectric data for solution of the liquid-crystals indicated a high molecular association, comparable to that observed in the nematic or isotropic phases. Electro-optic Kerr effect were investigated for some alkyl cyanobiphenyls, their nematic mixtures and the eutectic mixture liquid-crystals E7 and E8 in the isotropic phase and solution. The Kerr constant of these liquid crystals found to be very high at the nematic-isotropic transition temperatures as the molecules are expected to be highly ordered close to phase transition temperatures. Dynamic Kerr effect behaviour and transient molecular reorientation were also observed in thin layers of some alkyl cyanobiphenyls. Dichroic ratio R and order parameters of solutions containing some azo and anthraquinone dyes in the nematic solvent (E7 and E8), were investigated by the measurement of the intensity of the absorption bands in the visible region of parallel aligned samples. The effective factors on the dichroic ratio of the dyes dissolved in the nematic solvents were determined and discussed.

Quasi-lossless optical links for broad-band transmission and data processing

We present the first experimental implementation of a recently designed quasi-lossless fiber span with strongly reduced signal power excursion. The resulting fiber waveguide medium can be advantageously used both in lightwave communications and in all-optical nonlinear data processing.

Capture and storage of respiratory test data

DUE TO COPYRIGHT RESTRICTIONS ONLY AVAILABLE FOR CONSULTATION AT ASTON UNIVERSITY LIBRARY WITH PRIOR ARRANGEMENT

Processing and rendering of Fourier domain optical coherence tomography images at a line rate over 524 kHz using a graphics processing unit

In Fourier domain optical coherence tomography (FD-OCT), a large amount of interference data needs to be resampled from the wavelength domain to the wavenumber domain prior to Fourier transformation. We present an approach to optimize this data processing, using a graphics processing unit (GPU) and parallel processing algorithms. We demonstrate an increased processing and rendering rate over that previously reported by using GPU paged memory to render data in the GPU rather than copying back to the CPU. This avoids unnecessary and slow data transfer, enabling a processing and display rate of well over 524,000 A-scan/s for a single frame. To the best of our knowledge this is the fastest processing demonstrated to date and the first time that FD-OCT processing and rendering has been demonstrated entirely on a GPU.

All-optical passive 2R regeneration of dispersion-managed RZ data at 40 and 80 Gbit/s using in-line NOLMs

We demonstrate that the use of in-line nonlinear optical loop mirrors (NOLMs) in dispersion-managed (DM) transmission systems dominated by amplitude noise can achieve passive 2R regeneration of a 40 and 80 Gbit/s RZ data stream. This is an indication that the use of this approach could obviate the need for full-regeneration in high data rate, strong DM systems, when intra-channel four-wave mixing poses serious problems.

Study of the operating regime for all-optical passive 2R regeneration of dispersion-managed RZ data at 40 Gb/s using in-line NOLMs

In this letter, we numerically demonstrate that the use of inline nonlinear optical loop mirrors in strongly dispersion-managed transmission systems dominated by pulse distortion and amplitude noise can achieve all-optical passive 2R regeneration of a 40-Gb/s return-to-zero data stream. We define the tolerance limits of this result to the parameters of the input pulses.

All-optical passive regeneration of 40 Gbit/s soliton data stream using dispersion management and in-line NOLMs

It is numerically demonstrated, for the first time, that dispersion management and in-line nonlinear optical loop mirrors can achieve all-optical passive regeneration and distance-unlimited transmission of a soliton data stream at 40 Gbit/s over standard fibre.

All-optical passive regeneration of 40 Gbit/s soliton data stream using dispersion-management and in-line nonlinear optical loop mirrors

We numerically demonstrate for the first time that dispersion management and in-line nonlinear optical loop mirrors can achieve all-optical passive regeneration and distance-unlimited transmission of a soliton data stream at 40 Gbit/s over standard fibre.

All-optical passive regeneration of 40 Gbit/s soliton data stream using dispersion management and in-line NOLMs

The feasibility of stable soliton transmission system was demonstrated using a practical dispersion map in conjunction with in-line nonlinear optical loop mirrors (NOLMs). The system's performance was examined at 40 Gbit/s data rate in terms of maximum propagation distance corresponding to a bit error rate of more than 10-9. The bit error rate was estimated by means of the standard Q-factor.

Quasi-lossless optical links for broad-band transmission and data processing

We present the first experimental implementation of a recently designed quasi-lossless fiber span with strongly reduced signal power excursion. The resulting fiber waveguide medium can be advantageously used both in lightwave communications and in all-optical nonlinear data processing. © 2005 IEEE.

Simultaneous clock recovery and data regeneration using a nonlinear optical loop mirror as an all-optical mixer

High-speed optical clock recovery, demultiplexing and data regeneration will be integral parts of any future photonic network based on high bit-rate OTDM. Much research has been conducted on devices that perform these functions, however to date each process has been demonstrated independently. A very promising method of all-optical switching is that of a semiconductor optical amplifier-based nonlinear optical loop mirror (SOA-NOLM). This has various advantages compared with the standard fiber NOLM, most notably low switching power, compact size and stability. We use the SOA-NOLM as an all-optical mixer in a classical phase-locked loop arrangement to achieve optical clock recovery, while at the same time achieving data regeneration in a single compact device