Home > Seminars > Seminar By Hamza Soury, PhD Candidate (KAUST)

Reduced Rank Adaptive Filtering in Impulsive Noise Environments By Hamza Soury, PhD Candidate (KAUST)

  • Class schedule:  Sunday April 19th, 2015 from 12:30 pm to 01:00 pm
  • Location: Building 9, Room 2322
  • Refreshments: Available at 11:45 am


Adaptive filtering is extensively used in signal processing applications and attracts attention in different fields of research such as communications, equalization and multi user detection. The most used criteria in the existent adaptive filters is the Minimum Mean Square Error (MMSE), which gives efficient results in Gaussian noise scenario. However, in real life applications, severe impairments and perturbations may have an impulsive nature. These perturbations can deteriorate the performance of many adaptive filters. An impulsive noise environment, modeled by an alpha-stable distribution, is considered in this talk. Two approaches are used to deal with impulsive noise. The first one considers the received signal without preprocessing. In this case, from a received signal, we estimate directly the corresponding filter. This technique conserves a lot of information about the signal. However the impulses coming from the noise may deteriorate it. The other approach truncates the received signal in order to limit it in a given interval. This technique may loose some information about the signal, but the high impulses will no more affect the signal. Both techniques are tested in this work. In parallel of the use of truncated signal, the reduced-rank filter is combined with this approach. In fact, a full rank direct solution is derived followed by an iterative solution. The reduced rank adaptive filter is presented in this environment by using two methods for rank reduction, while the minimized objective function is defined using the Lp norm. The results are presented and the efficiency of each method is discussed.


Hamza Soury received Diplome d’Ingenieur from Ecole Polytechnique de Tunisie (EPT),Tunis, Tunisia in June 2010, and M.S. from KAUST in 2012. Currently, he is a PhD Candidate  in Electrical Engineering department at KAUST. His research interests include the analysis performance of communication systems perturbed by non-Gaussian noise/interference, with current research focusing on adaptive filtering in impulsive noise environments and probability of error of communication systems perturbed by Laplacian noise.