Wavelet Time Shift Properties Integration with Support Vector Machines

Abstract

Abstract. This paper presents a short evaluation about the integration of information derived from wavelet non-linear-time-invariant (non-LTI) projection properties using Support Vector Machines (SVM). These properties may give additional information for a classifier trying to detect known patterns hidden by noise. In the experiments we present a simple electromagnetic pulsed signal recognition scheme, where some improve is achieved with respect to previous work. SVMs are used as a tool for information integration, exploiting some unique properties not found in neural networks.

Discrete wavelet cycle integration for pulse detection

“…a linear classifier). The use of SVM on this paper has not been deemed relevant, because the special properties we used form this state-of-the-art algorithm were already discussed in [5]. Figure 3: The mean difference axis transformation.…”

“…There is a Nsamples pulse, N≤H, hidden inside white Gaussian noise all over the H-samples window. Our objective is to detect and locate the pulse within the window (see [5] for a detailed description). This is a very common problem for radar as well as communications signal processing.…”

“…On previous work [4,5,11] the authors have developed an algorithm for pulsed signal detection. This algorithm could be applied to several real-life problems like communications or radar, obtaining a process gain better than using previous approaches with wavelets [14].…”