FFT-Based Spectrum Analysis Using a Digital Signal Processor

Abstract

A spectrum analyzer based on Fast Fourier Transform (FFT) techniques was implemented using the TMS320C6201 Digital Signal Processor device manufactured by Texas Instruments. Portable C programs demonstrated optimization of the FFT algorithm for maximum speed. Previously published algorithms were adapted to the unique features of this Very-Long Instruction Word (VLIW) parallel processor and application, taking into account fixed-point arithmetic, parallel operation of functional units, and a hierarchy of memory capacities and speeds. The effectiveness of the VLIW C compiler, with automatic optimization, is compared with an explicitly-scheduled assembly-language program. The resulting program was then used to demonstrate the crucial need to keep program data in the Internal Data Memory to preserve hard-won performance gains.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 19, 2001
Accession Number
ADA401369

Entities

People

  • Charles T. Dorcey

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Advanced Electronics
  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Arithmetic Units
  • Assembly Languages
  • Computer Architecture
  • Computer Programming
  • Computer Programs
  • Computers
  • Digital Signal Processing
  • Electrical Engineering
  • Floating Point Operations
  • Instruction Set Architecture
  • Operating Systems
  • Parallel Processors
  • Signal Processing
  • Software Development
  • Spectrum Analysis
  • Spectrum Analyzers
  • Spreadsheet Software

Fields of Study

  • Engineering

Readers

  • Approximation Theory.
  • Computational Linguistics
  • Integrated Circuit Design and Technology.