• DocumentCode
    1850330
  • Title

    Output reconstruction in an OFB with instantaneous erasure using parity-check matrix

  • Author

    Akbari, Mohsen ; Labeau, Fabrice

  • Author_Institution
    Electr. & Comput. Eng. Dept., McGill Univ., Montreal, QC, Canada
  • fYear
    2012
  • fDate
    27-31 Aug. 2012
  • Firstpage
    1179
  • Lastpage
    1183
  • Abstract
    This paper introduces a new method based on parity-check matrix for reconstructing the output of an Oversampled Filter Bank (OFB) when sub-band vectors in different time instances are subject to erasures with arbitrary patterns (i.e. instantaneous erasure). Using the property of parity-check matrix, in this method, the erased samples of the sub-band vectors are recovered in the sub-band domain and then they are fed to the synthesis filters for output perfect reconstruction. In particular, we provide the theoretical sufficient conditions on OFB structure in order for this approach to be applicable for (maximum) erasure recovery. Furthermore, we modify the proposed reconstruction method in the presence of quantization noise and suggest an approach based on consistent reconstruction. We will also provide simulation results to evaluate the performance of this reconstruction method.
  • Keywords
    channel bank filters; matrix algebra; quantisation (signal); signal reconstruction; OFB structure; instantaneous erasure; oversampled filter bank; parity-check matrix; quantization noise; signal output perfect reconstruction; subband domain; subband vectors; synthesis filters; Filter banks; Finite impulse response filter; Noise; Quantization; Reconstruction algorithms; Vectors; Consistent Reconstruction; Instantaneous Erasure; Oversampled Filter Bank; Parity-Check Matrix;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Conference (EUSIPCO), 2012 Proceedings of the 20th European
  • Conference_Location
    Bucharest
  • ISSN
    2219-5491
  • Print_ISBN
    978-1-4673-1068-0
  • Type

    conf

  • Filename
    6333992