Discrete-time signals: Conversion of analogue signals to discrete time representation, digital representation of signals, sampling and sampling theorem, aliasing introduced by sampling, reconstruction of signals, sample rate conversion.
Z-transform: Z-transform, properties of z-transforms, transfer functions, bilinear transformation, causality and stability, determining frequency response from poles and zeroes.
Fourier Transform: Discrete-time Fourier transform, properties of Fourier transforms, Fourier transforms for periodic signals and applications of the Fourier transforms
Basic DSP Concepts: Meaning of frequency in discrete-time signals, discrete-time frequency analysis, DTFT and DFT, how to interpret DFT (cycles/sample vs. cycles/sec, zero-padding, spectral leakage and windowing), fast algorithms (FFT) for spectral analysis, discrete-time filters (moving average, autoregressive, and their combination).
Analysis and design of digital filters: Analysis of filters, impulse response, frequency response (magnitude vs. phase, ripple and group delay), FIR vs. IIR, stability of IIR filters, motivation for Z-transform, pole-zero analysis, design of FIR filters to specification (windowing and optimization techniques), design of IIR filters, classical designs (Butterworth, Chebyshev, etc)
Architectures and implementation: Algorithms for implementing filters, Convolution, correlation and their implementation, quantization, , quantization noise, factors affecting quantization of filter coefficients and practical issues such as limit-cycles and dead bands, up sampling and down sampling (how a 1-bit ADC and DAC work)
Examples and applications: Audio processing examples using DSP (equalization and reverb), speech processing examples (The spectrogram implementation methods in software and hardware (DSP architectures).
Matlab exercises will accompany each of the above
Lab sessions will involve implementing DSP applications on a suitable DSP Chip.
Programme: CE
Discrete-time signals: Conversion of analogue signals to discrete time representation, digital representation of signals, sampling and sampling theorem, aliasing introduced by sampling, reconstruction of signals, sample rate conversion. Z-transform: Z-transform, properties of z-transforms, transfer functions, bilinear transformation, causality and stability, determining frequency response from poles and zeroes. Fourier Transform: Discrete-time Fourier transform, properties of Fourier transforms, Fourier transforms for periodic signals and applications of the Fourier transforms Basic DSP Concepts: Meaning of frequency in discrete-time signals, discrete-time frequency analysis, DTFT and DFT, how to interpret DFT (cycles/sample vs. cycles/sec, zero-padding, spectral leakage and windowing), fast algorithms (FFT) for spectral analysis, discrete-time filters (moving average, autoregressive, and their combination). Analysis and design of digital filters: Analysis of filters, impulse response, frequency response (magnitude vs. phase, ripple and group delay), FIR vs. IIR, stability of IIR filters, motivation for Z-transform, pole-zero analysis, design of FIR filters to specification (windowing and optimization techniques), design of IIR filters, classical designs (Butterworth, Chebyshev, etc) Architectures and implementation: Algorithms for implementing filters, Convolution, correlation and their implementation, quantization, , quantization noise, factors affecting quantization of filter coefficients and practical issues such as limit-cycles and dead bands, up sampling and down sampling (how a 1-bit ADC and DAC work) Examples and applications: Audio processing examples using DSP (equalization and reverb), speech processing examples (The spectrogram implementation methods in software and hardware (DSP architectures). Matlab exercises will accompany each of the above Lab sessions will involve implementing DSP applications on a suitable DSP Chip.