Q: Explain Sampling Theorem with example
Sampling Theorem states that a continuous-time signal can be completely reconstructed from its
samples if the sampling frequency is at least twice the highest frequency present in the signal.
Mathematically, fs ≥ 2fmax. This is called the Nyquist rate. Example: For a 3 kHz audio signal, fs ≥
6 kHz. In practice, engineers use a slightly higher rate (8 kHz or more) to allow a transition band for
the anti-aliasing filter.
Q: What is aliasing? How to prevent it?
Aliasing is distortion caused when higher frequency components fold back into the baseband due to
undersampling. It makes two different frequencies appear as the same frequency. Solution: Use an
anti-aliasing low-pass filter to remove components above fs/2 before sampling and always choose
fs ≥ 2fmax.
Q: Difference between convolution and correlation
Convolution is used to find the output of an LTI system given input and impulse response (y[n] =
x[n] * h[n]). It involves flipping one signal, shifting, and multiplying. Correlation measures similarity
between two signals (no flipping). Used for signal detection and synchronization.
Q: FIR vs IIR Filters
FIR filters always have finite impulse response and can provide exact linear phase, making them
ideal for applications needing phase preservation. They are inherently stable. IIR filters use
feedback and can achieve a given response with fewer coefficients but may have nonlinear phase
and can be unstable if not carefully designed.
Q: Explain FFT and its advantage over DFT
DFT directly computes N frequency bins using N² multiplications. FFT uses divide-and-conquer to
reduce computation to N log■N operations. This makes spectrum analysis fast enough for real-time
applications like OFDM communication, radar, and audio analysis.
Q: Frequency translation in DSP
Multiplying a signal x(t) with cos(2πfct) or e^(j2πfct) shifts its frequency spectrum by ±fc. This is
used in modulation/demodulation, frequency upconversion/downconversion, and superheterodyne
receiver design.
Q: Radar Range Equation
The received power in a radar system is given by: Pr = (Pt * Gt * Gr * λ² * σ) / ((4π)³ * R■ * L) It
shows power falls with R■ (two-way propagation loss). This equation helps design transmit power
and antenna gain needed for required detection range.
Q: Radar Cross Section (RCS)
RCS (σ) represents how detectable an object is by radar. It depends on target size, shape, material,
and aspect angle. Larger RCS → stronger echo. Example: fighter jet has low RCS (stealth).
Q: Doppler Effect in Radar
Moving target causes frequency shift fd = 2v/λ (for monostatic radar). This shift helps measure
radial velocity. Positive fd indicates target moving toward radar; negative means moving away.
Q: Matched Filter
A filter designed to maximize output SNR when input contains a known signal in noise. It is
equivalent to cross-correlation with the transmitted waveform. Used for pulse compression in radar
to improve detection of weak targets.
Q: AM, FM, PM Differences
AM changes carrier amplitude with message signal, FM changes carrier frequency, PM changes
carrier phase. FM/PM are more robust to noise. AM is bandwidth efficient but noise-sensitive.
Q: BPSK and QAM
BPSK uses two phase states (0° and 180°) to transmit 1 bit per symbol. QAM uses combination of
amplitude and phase changes (e.g., 16-QAM transmits 4 bits per symbol). Higher-order QAM
increases data rate but requires higher SNR.
Q: Where QAM is used
QAM is widely used in WiFi, LTE, cable modems, and DSL because it provides high spectral
efficiency and supports high data rates in limited bandwidth.
Q: MATLAB FFT Code
Example: Fs = 1000; t = 0:1/Fs:1-1/Fs; x = cos(2*pi*50*t); X = fft(x); f = (0:length(X)-1)*Fs/length(X);
plot(f,abs(X)); xlabel('Frequency'); ylabel('|X(f)|');
Q: Moving Average Filter in C
Uses circular buffer to compute running average: #define N 5 float moving_avg(float x) { static float
buffer[N]={0}; static int index=0; static float sum=0; sum -= buffer[index]; buffer[index]=x; sum += x;
index=(index+1)%N; return sum/N; }
Q: Challenges in Real-Time DSP
Limited processing time per sample, memory constraints, fixed-point precision issues, interrupt
latency, jitter control, and efficient use of processor pipelines/DMAs.
Q: Testing and Validation
Verify with test vectors, compare with MATLAB reference output, check frequency response, use
hardware-in-loop testing, and analyze logs to catch edge cases.
