Project 1

Power-Aware Digital Signal Processing Blocks for Mobile Health

Staff:

  • Srinivasan Krishnamurthy

Current work involves:

  • Designing the EEG Artefact Rejection algorithm in MATLAB (for functionality testing).
  • Converting MATLAB code in to a ‘C’ code under VC++2008 compiler.
  • Analyzing the Complexity of the algorithm for hardware implementation.
  • Converting the ‘C’ code in to a suitable fixed-point DSP code.
  • Simulating the fixed-point DSP code in to the selected DSP hardware.
  • Porting the fixed-point DSP code in to the selected DSP hardware.
  • Testing the performance of the fixed-point DSP code in the DSP hardware.
  • Getting the power measures of the algorithm in the DSP hardware.

Present status of work:

A 11-channel FastICA algorithm for source separation using MATLAB – completed Conversion of MATLAB to ‘C’ code – completed.DSP hardware selection – completed Setting up of the DSP hardware simulator software – completed Conversion of floating-point ‘C’ to fixed-point DSP code – in progress.

Figures and Graphs:

(1)    Independent Component Analysis (ICA) – detailed block representation

Top