Digital Signal Processing (DSP) — Exam Notes
1. Introduction to DSP Systems
Definition: DSP involves numerical manipulation of signals using digital systems (computers, DSP
chips). Key Features: Works on discrete-time signals, involves convolution, correlation, Fourier
transform, filtering. Applications: Audio/video compression, communication (5G/Wi-Fi), medical
imaging (ECG/MRI), control systems.
2. Sampling Theorem
Statement: A continuous-time signal x(t) can be represented completely by its samples x(nT) if
sampling frequency fs ≥ 2fmax. Nyquist Rate: fN = 2fmax. Nyquist Interval: Ts = 1 / (2fmax). If fs <
2fmax → Aliasing (overlap of spectra).
3. Discrete-Time Signal & Systems
Discrete-Time Signal: Sequence of numbers defined at discrete instants, written as x[n]. Examples:
Impulse δ[n], Step u[n]. Discrete-Time System: Maps input x[n] → output y[n]. Examples: Delay
y[n]=x[n-1], Amplifier y[n]=3x[n].
4. Representation of Discrete Systems in Z-Domain
Z-Transform: X(z) = Σ x[n] z^(-n). System Representation (LCCDE):
y[n]+a1y[n-1]+...+aNy[n-N]=b0x[n]+...+bMx[n-M]. Transfer Function: H(z)=Y(z)/X(z)= (b0+
b1z^-1+...+bMz^-M)/(1+a1z^-1+...+aNz^-N). Poles & zeros determine stability and response.
5. Basic DSP Algorithms
Convolution: y[n] = Σ x[k]h[n-k]. Correlation: r_xy[n] = Σ x[k] y[k+n]. DFT: X[k] = Σ x[n] e^(-j2πkn/N).
FFT: Efficient method for computing DFT.
6. Representation of DSP Algorithms
Block Diagram: Graphical representation using delay, multiplier, adder blocks. Data Flow Graph
(DFG): Nodes=operations, edges=data. Dependence Graph: Shows execution order. Signal Flow
Graph (SFG): Nodes=signals, edges=operations.
7. Loop Bound & Iteration Bound
Loop Bound: Minimum time for one iteration, based on longest path delay. Iteration Bound:
Minimum average time per iteration for entire algorithm. Important for pipelining and parallel
processing.
8. Recursive (IIR) vs Non-Recursive (FIR) Systems
FIR (Non-Recursive): y[n]=Σ b_k x[n-k], no feedback, always stable, can have linear phase. IIR
(Recursive): y[n]=Σ b_k x[n-k] - Σ a_k y[n-k], uses feedback, may be unstable, nonlinear phase.
Examples: FIR = Moving Average filter, IIR = Butterworth/Chebyshev filters.
Feature FIR (Non-Recursive) IIR (Recursive)
Equation y[n]=Σ b_k x[n-k] y[n]=Σ b_k x[n-k] - Σ a_k y[n-k]
Feedback No Yes
Stability Always stable May be unstable
Phase Can be linear Usually nonlinear
Impulse Response Finite Infinite
Example Moving Average Butterworth, Chebyshev
