Intelligent Signal Processing for Autonomic Computing

Michael Potter, University of Manitoba
(with Witold Kinsner and Aram Faghfouri)

Abstract

This paper presents two examples of very complicated signals in order to demonstrate an effective use of distinct techniques in the area of intelligent signal processing (ISP) for autonomic computing (AC) systems. The first example considers a mixture of two electrocardiogram (ECG) signals, and uses the independent component analysis (ICA) to demix them, without any prior knowledge of the sources, in order to establish a basis for an autonomic diagnostic consultant [1].

The second example considers two-dimensional lightning-strike maps of Manitoba, and uses a robust wavelet-based technique [2] to compute their multifractal singularity spectra, together with a fuzzy classifier to extract the essential classes of strike concentrations for a better planning of power transmission lines.

This work is viewed in the context of the emerging autonomic computing (AC) systems which are self-configuring, self-optimizing, self-organizing, self-healing, self-protecting, and self-telecommunicating, thus leading to increased reliability, robustness, and dynamic flexibility. Such systems also require new solutions that might be offered through ISP and cognitive informatics.

References

[1] M. Potter and W. Kinsner, "Surrogate ECG signals for the null-hypothesis analysis of heart monitoring," in Proc. IEEE Can. Conf. Electrical & Computer Eng., CCECE05 (Saskatoon, SK; May 2-4, 2005).
[2] A. Faghfouri and W. Kinsner, "Local and global analysis of
multifractal singularity spectrum through wavelets," in Proc. IEEE Can. Conf. Electrical & Computer Eng., CCECE05 (Saskatoon, SK; May 2-4, 2005).