2013/07/24 by Sangnam Nam, Nam, Sangnam
Computer Science · Mathematics · #Digital Filter Design and Implementation #FOS: Computer and information sciences #Image and Signal Denoising Methods #Information Theory (cs.IT) #Mathematical Analysis and Transform Methods #cs.IT #math.IT
paper · pdf · doi:10.48550/arxiv.1307.6321
SampTA, Bremen : Germany (2013)
arxiv created 2013/07/24 · openalex publication_date 2013/07/24 · arxiv updated 2013/07/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
By use of window functions, time-frequency analysis tools like Short Time Fourier Transform overcome a shortcoming of the Fourier Transform and enable us to study the time- frequency characteristics of signals which exhibit transient os- cillatory behavior. Since the resulting representations depend on the choice of the window functions, it is important to know how they influence the analyses. One crucial question on a window function is how accurate it permits us to analyze the signals in the time and frequency domains. In the continuous domain (for functions defined on the real line), the limit on the accuracy is well-established by the Heisenberg's uncertainty principle when the time-frequency spread is measured in terms of the variance measures. However, for the finite discrete signals (where we consider the Discrete Fourier Transform), the uncertainty relation is not as well understood. Our work fills in some of the gap in the understanding and states uncertainty relation for a subclass of finite discrete signals. Interestingly, the result is a close parallel to that of the continuous domain: the time-frequency spread measure is, in some sense, natural generalization of the variance measure in the continuous domain, the lower bound for the uncertainty is close to that of the continuous domain, and the lower bound is achieved approximately by the 'discrete Gaussians'.