Cache modeling and optimization for portable devices running MPEG-4 video decoder

There are increasing demands on portable communication devices to run multimedia applications. ISO (an International Organization for Standardization) standard MPEG-4 is an important and demanding multimedia application. To satisfy the growing consumer demands, more functions are added to support MPEG-4 video applications. With improved CPU speed, memory sub-system deficiency is the major barrier to improving the system performance. Studies show that there is sufficient reuse of values for caching that significantly reduce the memory bandwidth requirement for video data. Software decoding of MPEG-4 video data generates much more cache-memory traffic than required. Proper understanding of the decoding algorithm and the composition of its data set is obvious to improve the performance of such a system. The focus of this paper is cache modeling and optimization for portable communication devices running MPEG-4 video decoding algorithm. The architecture we simulate includes a digital signal processor (DSP) for running the MPEG-4 decoding algorithm and a memory system with two levels of caches. We use VisualSim and Cachegrind simulation tools to optimize cache sizes, levels of associativity, and cache levels for a portable device decoding MPEG-4 video. Abu Asaduzzaman is, currently, a PhD candidate in the department of Computer Science and Engineering (CSE), Florida Atlantic University (FAU), Boca Raton, Florida. He received his MS degree in computer engineering from FAU in 1997. Mr. Asaduzzaman worked for ECI Telecom as a software engineer from 1998 to 2001. From 2001 to 2003, he worked for BlueCross and BlueShield of Florida and SunPass (FDoT) as an IT Consultant. Currently, he is working as a research assistant at CSE Dept, FAU. His research interests include cache optimization, architecture exploration, embedded system evaluation, and networks-on-a-chip (NoC). He has published several research papers in these areas. Abu is a member of the honor society of Phi Kappa Phi, Tau Beta Pi, Upsilon Phi Epsilon, and the Association for Computing Machinery (ACM) FAU Chapter. Imad Mahgoub received the MS degree in applied mathematics and MS degree in electrical and computer engineering, both from North Carolina State University, Raleigh in 1983 and 1986 respectively and the PhD degree in computer engineering from the Pennsylvania State University, University Park, PA in 1989. Dr. Mahgoub joined Florida Atlantic University (FAU), Boca Raton, Florida in 1989. Currently he is a full professor of Computer Science and Engineering department and the director of the Mobile Computing Laboratory. His research interests include performance evaluation, mobile computing, sensor networks, and parallel and distributed processing. He has published over 80 research papers in these areas. He is the co-editor of the Mobile Computing Handbook and the Handbook of Sensor Networks. Dr. Mahgoub has served on the program committees of numerous conferences. He has been the vice-chair for the Symposium on Performance Evaluation of Computer and Telecommunication Systems (SPECTS) since 2003. He is a senior member of the IEEE. He is also a member of Tau Beta Pi, Upsilon Pi Epsilon, the IEEE Computer Society, and the ACM.     

Incorporation of linear multipoint constraints in domain-decomposition-based iterative solvers - Part II: Blending FETI-DP and mortar methods and assembling floating substructures

Two approaches are presented for extending the FETI-DP domain-decomposition-based iterative method to the solution of problems with Linear Multipoint Constraints (LMPCs). Both cases of arbitrary LMPCs and LMPCs generated by the mortar method for enforcing a weak continuity of the solution across non-matching finite element interfaces are considered. In the first approach, the LMPCs are addressed during the preconditioning step but enforced only at convergence. In the second approach, the LMPCs are enforced at each iteration through the solution of an auxiliary coarse problem. Both extended FETI-DP solvers are benchmarked on a parallel processor for a wide range of constrained structural mechanics problems. The obtained results reveal that both solvers are numerically scalable and complement each other from the memory and CPU performance viewpoints.     

Influence of the cyclic plastic zone size on the propagation of the fatigue crack in case of 12NC6 steel

Under cyclic loading, the plasticized zone becomes complicated; it contains in particular a second plasticized zone, resulting from the local compression which occurs at the time of the closing of the crack to each cycle. The two plastic zones, monotonous (r_m) and cyclic (r_c), are proportional to (K_max/R_e)² et (ΔK/R_e)², respectively. The objective of this work is to study the evolution of the fatigue crack grown rate (FCGR) and the influence of the plastic zone size (r_c), which represents the seat of the residual stresses, on this evolution in the case of 12NC6 steel. Generally, the plastic zone size increases with the crack advance. The FCGR can be correlated with the energy absorptive in these plastic zones.     

Range profiles and images of a loaded straight wire

The range profiles and images of a straight wire without and with lumped impedance loading are discussed and demonstrated. The scattering mechanisms of a straight wire are qualitatively analyzed using the method of moments. Plots of range profiles at different aspect angles show that surface traveling waves are important scattering mechanisms of a straight wire. The presence of traveling wave makes ringlike artifacts appear in the reconstructed images. It was found that lumped impedance loading can effectively distort the range profiles and microwave images of a wire scatterer. In addition, randomly varied reactive loading can generate random peaks in the range profiles     

Chemical desulfurization of petroleum fractions for ultra-low sulfur fuels

An improved desulfurization process for removing sulfur from hydro treated diesel oil based on the oxidation of thiophenic type sulfur-containing compounds with H₂O₂ and acetic acid (AcOH) using H₂SO₄ as catalyst has been studied. The experimental results show that the sulfone content in the oxidation product increased rapidly with an increase in acetic acid and sulfuric acid ratios from 1:0 to 2:1 mole ratios. The maximum DBT conversion (wt.%) was at 2:1 mole ratio of acetic acid/sulfuric acid. This oxidation process is found to be capable of removing up to 90% of the sulfur compounds in hydro treated real fuels and can provide an alternative way to meet the future sulfur environmental requirements.     

Optimizing substrate surface and catalyst conditions for high yield chemical vapor deposition grown epitaxially aligned single-walled carbon nanotubes

Single-crystal stable-temperature (ST)-cut quartz substrates, which have a (0 1 1 1) crystallographic plane with their surface normal lying close to 38° from the y axis ([0 1 0]), were annealed in air prior to use as a support for aligned carbon nanotube growth by chemical vapor deposition. Very smooth substrate surfaces were obtained with annealing times in the vicinity of 15 h at a temperature of 750 °C. These smooth surfaces are ideal for the growth of horizontally aligned SWCNTs with high spatial density, while less dense SWCNTs were obtained with less smooth surfaces. Under optimized growth conditions, only SWCNT are observed and they can grow to lengths in excess of 100 μm. Our findings suggest structural defects interfere with the growth process. A binary Fe/Co catalyst was employed to grow the nanotubes. No obvious dependence on the Fe:Co ratio is observed.     

Image interpretation and prediction in microwave diversity imaging

The microwave image of a metallic object is interpreted from a point of view based on the understanding of the interconnection between the scattering mechanisms, the data acquisition system, and the image reconstruction algorithm. From this understanding it is possible to interpret and predict microwave images reconstructed from data collected over specified and angular windows. The connection between a special scattering mechanism, edge diffraction, and its reconstructed image is established. The microwave image of an edge is two bright points whose locations correspond to the end points of the edge if the normal aspect angle is not included in the angular windows; otherwise a line joining the two end points and representing the edge will appear in the image. Experimental images of a trihedral reflector constructed from data collected over different angular windows support this approach to image interpretation and prediction     

STRUCTURAL CHARACTERIZATION OF ARABIAN HEAVY CRUDE OIL RESIDUE

The residue (370°C+) from Arabian Heavy Crude Oil was separated into four fractions, asphaltenes, resins, aromatcis and saturates. The four fractions were found to be free of artifacts and analytically significant in themselves. Each fraction was further characterized by elemental analysis, infrared spectroscopy, n.m.r. spectroscopy and mass spectroscopy. The aromatics are the major constituent of the residue and the ratio of asphaltenes, resins, aromatics and saturates is about 2:3:8:3. The strucutral characterization study led to the conclusion that asphaltene fraction is maximum hydrogen deficient followed by resins, aromatics and saturates thus suggesting larger degree of ring condensation in the structure of asphaltenes than resins and aromatics.     

Selection of Methodology for Neural Network Modeling of Constitutive Hystereses Behavior of Soils

Classic constitutive modeling of geomaterials based on the elasticity and plasticity theories suffers from limitations pertaining to formulation complexity, idealization of behavior, and excessive empirical parameters. This article capitalizes on the modeling capabilities of neural networks as substitutes for the classic approaches. The neural network–based modeling overcomes the difficulties encountered in understanding the underlying microscopic processes governing the material's behavior by redirecting the efforts into learning the cause-effect relations from behavioral examples. Several methodologies are presented and cross-compared for effectiveness in approximating a theoretical hysteresis model resembling stress-strain behavior. The most effective methodology was used in modeling the constitutive behavior of an experimentally tested soil and produced models that simulated the real behavior of the soil with high accuracy. Although these models are empirical, they are retrainable and thus, unlike classic constitutive modeling techniques, can be revised and generalized easily when new data become available.