Side-by-side comparison
| Parameter | Continuous | Discrete Signals |
|---|---|---|
| Definition | Defined for every value of continuous time t | Defined only at discrete time instants n = 0, ±1, ±2, … |
| Time variable | t ∈ (−∞, ∞), real-valued | n ∈ integers only |
| Mathematical notation | x(t) | x[n] |
| Signal energy example | ∫|x(t)|² dt over all t | Σ|x[n]|² over all n |
| Processing hardware | Op-amps, RC circuits, analog ICs like LM741 | DSPs, microcontrollers, FPGAs like STM32 or Xilinx Artix |
| Noise susceptibility | Accumulates noise continuously along the channel | Noise immunity high; bit decisions made at threshold |
| Storage | Requires analog media (magnetic tape, vinyl) | Stored as binary in RAM, flash, or SD cards |
| Transforms used | Fourier Transform, Laplace Transform | DTFT, DFT, Z-Transform |
| Typical bandwidth | Audio: 20 Hz – 20 kHz; RF: MHz–GHz range | Determined by sampling rate: fs/2 (Nyquist limit) |
| Real-world example | ECG waveform from electrodes before sampling | MP3 audio at 44,100 samples/second |
Key differences
A continuous signal x(t) carries information at every time instant, so its Fourier Transform spans the entire frequency axis. A discrete signal x[n] only exists at integer indices; its DTFT is periodic with period 2π, which is why aliasing occurs when fs < 2fmax. The Laplace Transform suits continuous systems (RLC circuits), while the Z-Transform is the discrete counterpart. Real ADCs like the ADS1115 operate at fixed fs, making the Nyquist criterion — fs ≥ 2fmax — the first design constraint you must satisfy.
When to use Continuous
Use a continuous signal representation when designing or analysing analog front-end circuits — for example, when computing the frequency response of a Sallen-Key active filter built around the TL072 op-amp.
When to use Discrete Signals
Use discrete signals when the system processes data digitally — for example, when implementing an FIR notch filter at 50 Hz on an STM32 microcontroller to suppress power-line interference from an ECG signal.
Recommendation
For most placement tests and university practicals, choose the discrete signal framework as soon as sampling or digital processing is involved. Nearly every modern system — from audio codecs to 5G baseband — is discrete after the ADC stage, so fluency with x[n] notation and Z-transforms matters more in interviews.
Exam tip: Examiners frequently ask you to sketch x(t) vs x[n] for a given expression and to state whether energy or power is finite — know both the integral and summation forms of energy.
Interview tip: Interviewers at core companies like Qualcomm or Texas Instruments expect you to state the Nyquist theorem with a number — "sampling at 8 kHz captures audio up to 4 kHz" — not just the formula.