Elektrik- Elektronik Mühendisliği
Permanent URI for this community
Browse
Browsing Elektrik- Elektronik Mühendisliği by Subject "Adaptive signal processing."
Now showing 1 - 2 of 2
Results Per Page
Sort Options
Item Audio source separation with convolutively mixed signals(Thesis (M.S.)-Bogazici University. Institute for Graduate Studies in Science and Engineering, 2005., 2005.) Başaran, Doğaç.; Harmancı, Kerem.In this thesis, we addressed the problem of audio source separation of convolutivelymixed signals using microphone arrays. Independent Component Analysisis a major statistical tool for solving this problem. In real room environments, the recordings of audio signals usually involve the signal itself as well as some delayed andamplitude modulated versions of this signal. This is due to reverberation or echo ofthe room which occurs as a result of reflection of walls, ceiling, ground as well as thefurniture inside. Separation of signals that are mixed in these kinds of environments is a challenging problem. There exist both time-domain and frequency-domain solutionsto this problem. We mostly focus on frequency domain methods where ICA isperformed separately in each frequency bin. Permutation ambiguity which is the basicproblem in frequency domain ICA, is also handled with two basic approaches which are, direction of arrival method which is motivated by conventional beamforming theoryand interfrequency correlations which is motivated by nonstationarity of speechsignals. Conventional and Adaptive Beamforming methods are also implemented here.These methods separate sources by exploiting the physics of the propagation.Item Transform domain adaptive filtering using Walsh transforms(Thesis (M.S.) - Bogazici University. Institute for Graduate Studies in Science and Engineering, 1986., 1986.) Üsküdar, Derviş.; Morgül, Avni,The purpose of this study is to investigate the behaviour of the transform domain adaptive filter in the Walsh domain. As it is well known, because of the binary nature of the basis functions, the computationally intensive frequency domain adaptive filter algorithm lends itself to easier implementation if Walsh transform is employed. This thesis starts with a brief review of Walsh functions, Walsh transforms and adaptive filtering. The Walsh domain adaptive filter has been simulated on the digital computer. Simulation results for various input signals embedded in noise are given. It is shown that better convergence can be achieved when the adaptive filter is operated in the Walsh domain, rather than time domain. The effect of initial weight on convergence has been investigated. It has been observed that small values of the initial weight are necessary for better convergence performance. Although the theoretical basis of the effect of small initial weight on convergence characteristics has not been established yet, simulation results indicate improved performance with a small initial weight. Since the Walsh transform requires only additions and subtractions. the digital implementation of Walsh domain adaptive filter is easily achievable, and can be thought as the scope of a further study.