Michael Arntzen

Glinda: a framework for accelerating imbalanced applications on heterogeneous platforms

Heterogeneous platforms integrating different processors like GPUs and multi-core CPUs become popular in high performance computing. While most applications are currently using the homogeneous parts of these platforms, we argue that there is a large class of applications that can benefit from their heterogeneity: massively parallel imbalanced applications. Such applications emerge, for example, from variable time step based numerical methods and simulations. In this paper, we present Glinda, a framework for accelerating imbalanced applications on heterogeneous computing platforms. Our framework is able to correctly detect the application workload characteristics, make choices based on the available parallel solutions and hardware configuration, and automatically obtain the optimal workload decomposition and distribution. Our experiments on parallelizing a heavily imbalanced acoustic ray tracing application show that Glinda improves application performance in multiple scenarios, achieving up to 12x speedup against manually configured parallel solutions.

A framework for simulation of aircraft flyover noise through a non-standard atmosphere

This paper describes a new framework for the simulation of aircraft flyover noise through a non-standard atmosphere. Central to the framework is a ray-tracing algorithm which defines multiple curved propagation paths, if the atmosphere allows, between the moving source and listener. Because each path has a different emission angle, synthesis of the sound at the source must be performed independently for each path. The time delay, spreading loss and absorption (ground and atmosphere) are integrated along each path, and applied to each synthesized aircraft noise source to simulate a flyover. A final step assigns each resulting signal to its corresponding receiver angle for the simulation of a flyover in a virtual reality environment. Spectrograms of the results from a straight path and a curved path modeling assumption are shown. When the aircraft is at close range, the straight path results are valid. Differences appear especially when the source is relatively far away at shallow elevation angles. These differences, however, are not significant in common sound metrics. While the framework used in this work performs off-line processing, it is conducive to real-time implementation.

Framework for Simulating Aircraft Flyover Noise Through Nonstandard Atmospheres

This paper describes a new framework for the synthesis of aircraft flyover noise through a nonstandard atmosphere. Central to the framework is a ray-tracing algorithm that defines multiple curved propagation paths, if the atmosphere allows, between the moving source and listener. Because each path has a different emission angle, synthesis of the sound at the source must be performed independently for each path. The time delay, spreading loss, and absorption (ground and atmosphere) are integrated along each path and applied to each synthesized aircraft noise source to simulate a flyover. A final step assigns each resulting signal to its corresponding receiver angle for the simulation of a flyover in a virtual reality environment. Spectrograms of the results from a straight path and a curved path modeling assumption are shown. When the aircraft is at close range, the straight path results are valid. Differences appear especially when the source is relatively far away at shallow elevation angles. These diffe...

Aircraft Noise Simulation for a Virtual Reality Environment

Aircraft noise annoyance predictions based on traditional tools lack fidelity when modelling a single aircraft flyover. For evaluating annoyance of new procedures, different aircraft types or changing atmospheric conditions, a new modelling approach is necessary. A research effort has been initiated to develop a toolchain that links each of the relevant components. This allows the use of synthesized sound in a virtual reality simulator and enhances follow-up annoyance investigations. The toolchain components and current capabilities are presented in this paper. Audible results are illustrated by spectrograms to show the difference between different atmospheric conditions. Through the development of the toolchain, it becomes possible to demonstrate audible results of changing aircraft procedures, type and atmosphere in a virtual reality environment. This opens new ways to investigate how people value aircraft noise in different atmospheres in combination with smart planning of aircraft procedures.

Weather-Dependent Airport Noise Contour Prediction Concept Based on Ray Tracing

The most frequently used excess attenuation method for noise contour models is called the lateral attenuation correction. This empirical method disregards varying atmospheric conditions. It is known that varying atmospheric conditions can lead to large discrepancies for single-event cases. This paper studies these atmospheric effects for a longer period, thus involving multi-event calculations. A European Civil Aviation Conference Document 29 compliant noise model is used with the extended functionality to apply results from a ray-tracing excess attenuation calculation. Results are shown for monthly and yearly noise contours around an example airport using three different modeling options: the standard Document 29 methodology, the standard methodology augmented by varying atmospheric absorption, and the ray-tracing excess attenuation method. The differences found in excess attenuation between the standard Document 29 method and ray tracing are small. It is argued that the main differences are related to v...

A weather dependent noise contour prediction concept: Calculating multi-event noise contours with ray-tracing

This paper expands recent work on a standard aircraft noise model with an advanced excess attenuation method. The most frequently used excess attenuation method for noise contour models is called the lateral attenuation correction and this empirical method disregards varying atmospheric conditions. However, it is known that for single-event cases varying atmospheric conditions can lead to large discrepancies in results. This paper studies these effects for a longer period, involving multi-event calculations. An ECAC Doc.29 compliant noise model is utilized which is extended with the functionality to apply results from a ray tracing excess attenuation calculation. Results are shown for monthly and yearly noise contours around a fictive airport using different modeling options. In the end the differences in results between the three modeling options are small. However, some differences can be distinguished for both the 58 LDEN and 48 LDEN contour. The most prominent difference is a smaller 58 LDEN contour area and a larger 48 LDEN contour area. Based on the results, it is argued that the effects of refraction, ground attenuation and atmospheric absorption each play a distinctive role in the found differences. In conclusion, the lateral attenuation model, used to estimate the average excess attenuation in a varying atmosphere, provides a practical and realistic estimate for a yearly and monthly noise contour.