Simulating a Multi-Target Acoustic Array on the Intel Paragon.
Abstract
This thesis is built around a parallel programming project which simulates the recordings of a linear array of hydrophones in the presence of several sources of acoustic signals (targets). Simulation of multi-target data with appropriate modeling of multipath, path refraction, and local and distant noise sources- is a useful and relatively sophisticated modeling chore. The signal suites include multiple frequency narrow band signals, broad-band flow noise, and randomly generated transient signals. The noise suites include coherent and incoherent Gaussian noise, impulsive noise, and continental shelf reverberation.This project is of interest from parallel programming viewpoint because it uses the Paragon a true Multiple Instruction Multiple Data (MIND) machine with three types of nodes: l)a Manager node, 2)tar- et nodes, and 3)Hydrophone nodes. The target nodes are responsible for generating the signal suites1 transient nose, and the coherent noise. The hydrophone nodes are responsible for calculation multipath, refraction and time delay as well as adding the local incoherent noise suite. While this simulation can be used as a stand-alone application, it also will from the foundation for a much larger, more sophisticated simulation, namely producing a virtual Command Information Center (CIC).
Document Details
- Document Type
- Technical Report
- Publication Date
- Oct 01, 1994
- Accession Number
- ADA288443
Entities
People
- Charles A. Jones
Organizations
- George Mason University