Karthik Nagarajan

RAT: a methodology for predicting performance in application design migration to FPGAs

Before any application is migrated to a reconfigurable computer (RC), it is important to consider its amenability to the hardware paradigm. In order to maximize the probability of success for an applications migration to an FPGA, one must quickly and with a reasonable degree of accuracy analyze not only the performance of the system but also the required precision and necessary resources to support a particular design. This extra preparation is meant to reduce the risk of failure to achieve the applications design requirements (e.g. speed or area) by quantitatively predicting the expected performance and system utilization. This paper presents the RC Amenability Test (RAT), a methodology for rapidly analyzing an applications design compatibility to a specific FPGA platform.

RAT: RC Amenability Test for Rapid Performance Prediction

While the promise of achieving speedup and additional benefits such as high performance per watt with FPGAs continues to expand, chief among the challenges with the emerging paradigm of reconfigurable computing is the complexity in application design and implementation. Before a lengthy development effort is undertaken to map a given application to hardware, it is important that a high-level parallel algorithm crafted for that application first be analyzed relative to the target platform, so as to ascertain the likelihood of success in terms of potential speedup. This article presents the RC Amenability Test, or RAT, a methodology and model developed for this purpose, supporting rapid exploration and prediction of strategic design tradeoffs during the formulation stage of application development.

Downscaling Satellite-Based Soil Moisture in Heterogeneous Regions Using High-Resolution Remote Sensing Products and Information Theory: A Synthetic Study

In this study, a novel methodology based upon the information-theoretic measures of entropy and mutual information was implemented to downscale soil moisture (SM) observations from 10 km to 1 km. It included a transformation function that related auxiliary remotely sensed (RS) products at high resolution to in situ SM observations to obtain first estimates of SM at 1 km and merging this estimate with SM at coarse resolutions through Principle of Relevant Information (PRI). The PRI-based estimates were evaluated using synthetic observations in NC Florida for heterogeneous agricultural land covers (LC), with two growing seasons of sweet corn and one of cotton, annually. The cumulative density function showed an overall error in SM of <; 0.03 cubic meter/cubic meter in the region, with a confidence interval of 95% during the simulation period. The PRI estimates at 1 km were also compared with those from the method based upon Universal Triangle (UT). The spatially averaged root mean square error (RMSE) aggregated over the vegetative LC were 0.01 cubic meter/cubic meter and 0.15 cubic meter/cubic meter using the PRI and UT methods, respectively. The RMSE for downscaled estimates using the UT method increased to 0.28 cubic meter/cubic meter when Laplacian errors are used, while the corresponding RMSE for the PRI remains the same for both Laplacian or Gaussian errors. The Kullback-Liebler divergence (KLD) for estimates using PRI is about 50% lower than those using the method based upon UT indicating that the probability density function (PDF) of the PRI estimate is closer to PDF of the true SM, than the UT method.

Accelerating Machine-Learning Algorithms on FPGAs using Pattern-Based Decomposition

Machine-learning algorithms are employed in a wide variety of applications to extract useful information from data sets, and many are known to suffer from super-linear increases in computational time with increasing data size and number of signals being processed (data dimension). Certain principal machine-learning algorithms are commonly found embedded in larger detection, estimation, or classification operations. Three such principal algorithms are the Parzen window-based, non-parametric estimation of Probability Density Functions (PDFs), K-means clustering and correlation. Because they form an integral part of numerous machine-learning applications, fast and efficient execution of these algorithms is extremely desirable. FPGA-based reconfigurable computing (RC) has been successfully used to accelerate computationally intensive problems in a wide variety of scientific domains to achieve speedup over traditional software implementations. However, this potential benefit is quite often not fully realized because creating efficient FPGA designs is generally carried out in a laborious, case-specific manner requiring a great amount of redundant time and effort. In this paper, an approach using pattern-based decomposition for algorithm acceleration on FPGAs is proposed that offers significant increases in productivity via design reusability. Using this approach, we design, analyze, and implement a multi-dimensional PDF estimation algorithm using Gaussian kernels on FPGAs. First, the algorithm’s amenability to a hardware paradigm and expected speedups are predicted. After implementation, actual speedup and performance metrics are compared to the predictions, showing speedup on the order of 20× over a 3.2 GHz processor. Multi-core architectures are developed to further improve performance by scaling the design. Portability of the hardware design across multiple FPGA platforms is also analyzed. After implementing the PDF algorithm, the value of pattern-based decomposition to support reuse is demonstrated by rapid development of the K-means and correlation algorithms.

Integration of Business Event and Rule Management With the Web Services Model

Web services technology is emerging as a promising infrastructure to support loosely coupled, Internet-based applications that are distributed, heterogeneous and dynamic. It provides a standards-based, process-centric framework for achieving the sharing of distributed heterogeneous applications. While Web services technology provides a promising foundation for developing distributed applications for e-business, additional features are required to make this paradigm truly useful in the real world. In particular, interactions among business organizations need to follow the policies, regulations, security and other business rules of the organizations. An effective way to control, restrict and enforce business rules in the use of Web services is to integrate business event and rule management concepts and techniques into the Web services model. In this paper, we focus on incorporating the business event and rule-management concepts into the Web services model at the service provider side. Based on a code-generation approach, we have developed techniques and implemented tools to generate Web service “wrappers†and other objects required to integrate an Event-Trigger-Rule (ETR) technology with the Web services technology.

Impact of Assimilating Passive Microwave Observations on Root-Zone Soil Moisture Under Dynamic Vegetation Conditions

In this paper, L-band microwave observations were assimilated using the ensemble Kalman filter to improve root-zone soil moisture (RZSM) estimates from a coupled soil vegetation atmosphere transfer (SVAT)-vegetation model linked to a forward microwave model. Simultaneous state-parameter updates were performed by assimilating both synthetic and field observations during a growing season of sweet corn every three days, matching the temporal coverage of observations from the Soil Moisture and Ocean Salinity and Soil Moisture Active Passive missions. The sensitivities of parameters to the states were investigated using the information-theoretic measure of conditional entropy. Among the soil parameters, the pore-size index (λ) was the most sensitive to brightness temperatures (TB) during the early and midgrowth stages, while porosity (φ) was the most sensitive to TB during the reproductive stage. In the microwave model, the soil roughness parameters, root mean square (RMS) height (r), and correlation length (l) were the most sensitive during the early and mid stages, while the vegetation regression parameter (b) was the most sensitive during the reproductive stage. In the synthetic experiment, assimilation of TB provided RMS error reductions in RZSM estimates of 70% compared to open loop estimates. Minimal variations in performance were observed across different stages of the season during the synthetic experiment. However, when field observations of TB were assimilated, significant differences in RZSM estimates were observed during different growth stages. Maximum RMS difference (RMSD) reductions in RZSM estimates of 33.3% were observed compared to open loop estimates during the early stages, while improvements of 4.8% and 16.7% were observed in the mid- and reproductive stages, respectively. Further analyses of assimilation with field observations also suggest some improvements in the SVAT model are needed for moisture transport immediately following the precipitation/irrigation events. In the microwave model, the linear vegetation formulation for estimating canopy opacity, parameterized by b, was inadequate in capturing the complexities in TB during stages of high vegetative and reproductive growth rates.

A Scalable Approach to Fusing Spatiotemporal Data to Estimate Streamflow via a Bayesian Network

Flow of water through stream networks directly impacts flooding and transport of sediments and pollutants in watershed systems. Hence, knowledge of streamflow is critical for water management and mitigation of flooding and drought events. Unfortunately, spatially dense networks of in situ streamflow measurements are generally unavailable and would be prohibitively expensive to deploy and maintain. Thus, a data fusion framework is needed that utilizes available data to predict streamflow. Observed data in spatial (e.g., topography and land cover), temporal (e.g., streamflow and groundwater levels), and spatiotemporal domains (e.g., rainfall) impact streamflow. Some of these quantities can be obtained from remote sensing imagery; however, combining such disparate data types using traditional data fusion methods is problematic. Physically based hydrologic models have been used to predict streamflow but often with significant uncertainty because numerous assumptions are made for many unmeasured input and parameter values. Traditional Bayesian inference approaches suffer from superlinear increases in computational complexity as the number of data sets to be fused grows. In this paper, a scalable spatiotemporal approach based on Bayesian networks (BNs) is presented for estimating streamflow. An information-theoretic methodology based on conditional entropy is employed to quantify the impact of adding nodes in the BN in terms of information gained. The framework offers the flexibility of embedding knowledge from hydrologic models calibrated for the study area by introducing them as additional nodes in the network, thereby improving prediction accuracy. Posterior probabilities of estimates and the associated entropy provide valuable information on the quality of predictions and also offer directions for future watershed instrumentation.

Spatial Scaling and Variability of Soil Moisture Over Heterogeneous Land Cover and Dynamic Vegetation Conditions

In this letter, near-surface and root-zone soil moisture (RZSM), land surface temperature (LST), leaf area index, and vegetation water content were simulated at different spatial scales for three land cover types in North Central Florida under dynamic vegetation conditions. Insights into expected retrieval errors in soil moisture (SM) due to assumptions of static landscape were obtained from differences in the estimates using the static and dynamic land covers. Maximum differences of about 0.04 m3/m3 in near-surface SM and RZSM, and 5.1 K in LST were observed between estimates obtained over the vegetated and bare-soil regions during dry-soil conditions. During wet conditions, the maximum differences in near-surface SM and RZSM increased to about 0.05 m3/m3, while those in LST decreased to 3.6 K. The RZSM simulations generated at the two resolutions of 200 m and 10 km were used to implement an upscaling algorithm based on averaging, to illustrate the use of the synthetic data set for upscaling studies. This letter highlights the importance of simulating land surface states at multiple scales for heterogeneous landscapes under dynamic vegetation conditions and for developing accurate SM retrieval and scaling algorithms.

Automated L-Band Radar System for Sensing Soil Moisture at High Temporal Resolution

The ground-based University of Florida L-band Automated Radar System (UF-LARS) was developed to obtain observations of normalized radar backscatter (\mmbσ0) at high temporal resolution for soil moisture applications. The system was mounted on a 25 m manlift with capabilities of antenna positioning for multi-angle data acquisition and ranging. The RF subsystem of UF-LARS was based upon the established designs for ground-based scatterometers employing a vector network analyzer with simultaneous acquisition of V- and H-polarized returns. System integration and automated data acquisition were enabled using a software control system. Fifteen-minute observations of \mmb σ0 collected over a growing season of sweet-corn and bare soil conditions in North Central Florida, were used to study the sensitivity of \mmbσ0 to growing vegetation and near-surface (0-5 cm) soil moisture (\mmbSM0 - 5). On average, \mmb σ\mmbVV0 were observed to be 23% higher than \mmbσ\mmbHH0 during the mid- and late-stages of crop growth due to the vertical structure of stems. The correlation between 3-day observations of \mmbSM0 - 5 and \mmbσ\mmbVV0 reduced by 55% compared to those obtained for ≤ 30-min observations. These findings suggested that data set at high temporal frequencies can be used to develop more realistic and robust forward backscattering models.

Scalable and Portable Architecture for Probability Density Function Estimation on FPGAs

This paper describes the design, development, and analysis of a scalable and portable architecture for multi-dimensional, non-parametric PDF estimation using Gaussian kernels on FPGAs.