ABSTRACT
The discrete time systems that process data at more than one sampling rate are known as multirate systems. The two
basic operations in multirate signal processing are decimation and interpolation.One of the important applications of
multirate signal processing is sub-band coding of speech signal. In the proposed system, speech signal is taken as input
signal. Additive White Gaussian Noise is added with the input speech signal. The input speech signal spectrum is divided
into frequency sub-bands using a bank of finite response filters. Hamming, Hanning, Blackman, Rectangular and Kaiser
windowing methods are used to implement the low pass and high pass filters. Finally performance of the proposed
system is evaluated on the TIMIT data base using the parameters like leakage factor, main lobe width, side lobe
attenuation, peak amplitude of side lobe and signal to noise ratio. The performance evaluation shows which window is
suitable for designing the finite impulse response filters and sub-band coding system.
Keywords: - Windowing, Signal to Noise Ratio, FIR Filter, Multirate.