Noise in Communication Systems and SNR With easy examples.

Noise is the enemy of every communication system: random, unwanted electrical disturbances that hitch a ride with your signal and corrupt it. It comes from the thermal jiggling of electrons in every resistor and amplifier (unavoidable), from lightning and motors (man-made), and from the sky itself (cosmic and atmospheric). No wire or wireless link is ever perfectly clean.

The single number that measures how badly noise is winning is the signal-to-noise ratio (SNR): the signal power divided by the noise power. Every design decision in communications — transmit power, antenna size, modulation choice, coding — is ultimately a fight to raise the SNR enough for the message to get through.

Key Concepts

• SNR: Psignal/Pnoise, almost always quoted in decibels: SNR(dB) = 10 log10(Psignal/Pnoise).
• Thermal noise: white noise from electron motion, with power N = kTB watts — it exists wherever there is temperature.
• Noise figure (NF): how much an amplifier degrades the SNR, in dB; a perfect amplifier has 0 dB.
• Noise temperature: another way to express an amplifier's noisiness, in kelvin; satellite receivers boast cryogenic low-noise amplifiers near 20 K.
• Shannon's law: channel capacity C = B log2(1 + SNR) bits/s — noise fundamentally caps how fast you can communicate.
• Eb/N0: energy-per-bit over noise density, the fair way to compare digital modulations at equal data rates.

Formulas

• SNR in dB: SNR(dB) = 10 log10(Ps/Pn); a ratio of 1000× is 30 dB, 100× is 20 dB, 10× is 10 dB.
• Thermal noise power: N = kTB, with Boltzmann's constant k = 1.38×10⁻²³ J/K, T in kelvin, B in Hz.
• Noise figure cascade (Friis): F = F1 + (F2 − 1)/G1 + … — the first stage's noise figure dominates, which is why low-noise amplifiers sit right at the antenna.
• Capacity: C = B·log2(1 + SNR) — units: B in Hz, SNR as a plain ratio, C in bits per second.
• Assumptions: thermal noise is additive white Gaussian (AWGN) in the textbook models; real channels add fading and interference on top.

Worked Example

A receiver measures signal power Ps = 1 mW and noise power Pn = 1 µW in its 1 MHz bandwidth. Step 1: ratio = 10⁻³/10⁻⁶ = 1000. Step 2: SNR = 10 log10(1000) = 30 dB — a healthy link. Step 3: check against thermal noise: N = kTB = 1.38×10⁻²³ × 290 × 10⁶ ≈ 4×10⁻¹⁵ W (−114 dBm). Step 4: Shannon capacity C = 10⁶ × log2(1001) ≈ 10⁶ × 9.97 ≈ 10 Mbps — the theoretical ceiling for this 1 MHz channel at 30 dB SNR.

Common Misconceptions

• "Amplifying the signal fixes a noisy link." Wrong: an amplifier boosts noise exactly as much as signal — SNR barely changes. Only a lower-noise first stage or a bigger antenna helps.
• "Zero noise is achievable with good enough gear." Wrong: thermal noise is set by temperature and bandwidth — physics, not engineering. You can cool the amplifier, not silence the universe.
• "A high SNR guarantees a good call." Not quite: distortion, interference, and dropouts can wreck quality even when random noise is tiny.

Real-World Applications

• Satellite TV dishes use ultra-low-noise block downconverters (LNBs) because the received signal is barely above thermal noise.
• Mobile networks plan cell sizes and transmit powers from link-budget SNR calculations.
• Wi-Fi rate adaptation: your router drops to slower, tougher modulations as SNR falls with distance.
• Deep-space probes like Voyager communicate at SNRs near 0 dB using massive dishes and powerful error-correction codes.

Noise is the tax physics levies on every message, and SNR is the receipt. Thermal noise sets the floor, the first amplifier stage sets the damage, and Shannon's law sets the speed limit. Design anything that talks — radio, fibre, or copper — and you are really designing an SNR budget.