Analysis and performance of a UPC implementation of a parallel longest common subsequence algorithm

An important problem in computational biology is finding the longest common subsequence (LCS) of two nucleotide sequences. This paper examines the correctness and performance of a recently proposed parallel LCS algorithm that uses successor tables and pruning rules to construct a list of sets from which an LCS can be easily reconstructed. Counterexamples are given for two pruning rules that were given with the original algorithm. Because of these errors, performance measurements originally reported cannot be validated. The work presented here shows that speedup can be reliably achieved by an implementation in Unified Parallel C that runs on an Infiniband cluster. This performance is partly facilitated by exploiting the software cache of the MuPC runtime system. In addition, this implementation achieved speedup without bulk memory copy operations and the associated programming complexity of message passing.

Parallel algorithm for solving integer linear programs

Linear programs, or LPs, are often used in optimization problems, such as improving manufacturing efficiency of maximizing the yield from limited resources. The most common method for solving LPs is the Simplex Method, which will yield a solution, if one exists, but over the real numbers. From a purely numerical standpoint, it will be an optimal solution, but quite often we desire an optimal integer solution. A linear program in which the variables are also constrained to be integers is called an integer linear program or ILP. It is the focus of this report to present a parallel algorithm for solving ILPs. We discuss a serial algorithm using a breadth-first branch-and-bound search to check the feasible solution space, and then extend it into a parallel algorithm using a client-server model. In the parallel mode, the search may not be truly breadth-first, depending on the solution time for each node in the solution tree. Our search takes advantage of pruning, often resulting in super-linear improvements in solution time. Finally, we present results from sample ILPs, describe a few modifications to enhance the algorithm and improve solution time, and offer suggestions for future work.

Control design and genetic algorithm optimization for electrostatic MEMS

This dissertation discusses structural-electrostatic modeling techniques, genetic algorithm based optimization and control design for electrostatic micro devices. First, an alternative modeling technique, the interpolated force model, for electrostatic micro devices is discussed. The method provides improved computational efficiency relative to a benchmark model, as well as improved accuracy for irregular electrode configurations relative to a common approximate model, the parallel plate approximation model. For the configuration most similar to two parallel plates, expected to be the best case scenario for the approximate model, both the parallel plate approximation model and the interpolated force model maintained less than 2.2% error in static deflection compared to the benchmark model. For the configuration expected to be the worst case scenario for the parallel plate approximation model, the interpolated force model maintained less than 2.9% error in static deflection while the parallel plate approximation model is incapable of handling the configuration. Second, genetic algorithm based optimization is shown to improve the design of an electrostatic micro sensor. The design space is enlarged from published design spaces to include the configuration of both sensing and actuation electrodes, material distribution, actuation voltage and other geometric dimensions. For a small population, the design was improved by approximately a factor of 6 over 15 generations to a fitness value of 3.2 fF. For a larger population seeded with the best configurations of the previous optimization, the design was improved by another 7% in 5 generations to a fitness value of 3.0 fF. Third, a learning control algorithm is presented that reduces the closing time of a radiofrequency microelectromechanical systems switch by minimizing bounce while maintaining robustness to fabrication variability. Electrostatic actuation of the plate causes pull-in with high impact velocities, which are difficult to control due to parameter variations from part to part. A single degree-of-freedom model was utilized to design a learning control algorithm that shapes the actuation voltage based on the open/closed state of the switch. Experiments on 3 test switches show that after 5-10 iterations, the learning algorithm lands the switch with an impact velocity not exceeding 0.2 m/s, eliminating bounce.

EXPLORATION OF THE MTSAT2 SATELLITE CAPABILITIES FOR REAL TIME DETECTION AND CHARACTERIZATION OF VOLCANIC EMISSIONS

In this report, we attempt to define the capabilities of the infrared satellite remote sensor, Multifunctional Transport Satellite-2 (MTSAT-2) (i.e. a geosynchronous instrument), in characterizing volcanic eruptive behavior in the highly active region of Indonesia. Sulfur dioxide data from NASA's Ozone Monitoring Instrument (OMI) (i.e. a polar orbiting instrument) are presented here for validation of the processes interpreted using the thermal infrared datasets. Data provided from two case studies are analyzed specifically for eruptive products producing large thermal anomalies (i.e. lava flows, lava domes, etc.), volcanic ash and SO2 clouds; three distinctly characteristic and abundant volcanic emissions. Two primary methods used for detection of heat signatures are used and compared in this report including, single-channel thermal radiance (4-µm) and the normalized thermal index (NTI) algorithm. For automated purposes, fixed thresholds must be determined for these methods. A base minimum detection limit (MDL) for single-channel thermal radiance of 2.30E+05 Wm- 2sr-1m-1 and -0.925 for NTI generate false alarm rates of 35.78% and 34.16%, respectively. A spatial comparison method, developed here specifically for use in Indonesia and used as a second parameter for detection, is implemented to address the high false alarm rate. For the single-channel thermal radiance method, the utilization of the spatial comparison method eliminated 100% of the false alarms while maintaining every true anomaly. The NTI algorithm showed similar results with only 2 false alarms remaining. No definitive difference is observed between the two thermal detection methods for automated use; however, the single-channel thermal radiance method coupled with the SO2 mass abundance data can be used to interpret volcanic processes including the identification of lava dome activity at Sinabung as well as the mechanism for the dome emplacement (i.e. endogenous or exogenous). Only one technique, the brightness temperature difference (BTD) method, is used for the detection of ash. Trends of ash area, water/ice area, and their respective concentrations yield interpretations of increased ice formation, aggregation, and sedimentation processes that only a high-temporal resolution instrument like the MTSAT-2 can analyze. A conceptual model of a secondary zone of aggregation occurring in the migrating Kelut ash cloud, which decreases the distal fine-ash component and hazards to flight paths, is presented in this report. Unfortunately, SO2 data was unable to definitively reinforce the concept of a secondary zone of aggregation due to the lack of a sufficient temporal resolution. However, a detailed study of the Kelut SO2 cloud is used to determine that there was no climatic impacts generated from this eruption due to the atmospheric residence times and e-folding rate of ~14 days for the SO2. This report applies the complementary assets offered by utilizing a high-temporal and a high-spatial resolution satellite, and it demonstrates that these two instruments can provide unparalleled observations of dynamic volcanic processes.

Variations of the FEAST Eigenvalue Algorithm

FEAST is a recently developed eigenvalue algorithm which computes selected interior eigenvalues of real symmetric matrices. It uses contour integral resolvent based projections. A weakness is that the existing algorithm relies on accurate reasoned estimates of the number of eigenvalues within the contour. Examining the singular values of the projections on moderately-sized, randomly-generated test problems motivates orthogonalization-based improvements to the algorithm. The singular value distributions provide experimentally robust estimates of the number of eigenvalues within the contour. The algorithm is modified to handle both Hermitian and general complex matrices. The original algorithm (based on circular contours and Gauss-Legendre quadrature) is extended to contours and quadrature schemes that are recursively subdividable. A general complex recursive algorithm is implemented on rectangular and diamond contours. The accuracy of different quadrature schemes for various contours is investigated.