8.2 Technical Noise

Technical noise refers to the physical and electrical disturbances that corrupt signals as they travel through communication channels—wires, optical fibers, wireless links, circuit components, and amplifiers—causing the received signal to differ from the transmitted one. Unlike semantic or psychological forms of communication noise, technical noise operates at the physical layer of communication systems and is characterized by precise mathematical models that allow engineers to quantify its effects, predict its impact on system performance, and design countermeasures. Technical noise sets the fundamental physical limits on the accuracy and speed of information transmission in any electromagnetic, photonic, or acoustic communication system.

The most ubiquitous form of technical noise is thermal noise, also called Johnson-Nyquist noise, which arises from the random thermal agitation of charge carriers in any resistive element at a temperature above absolute zero. Electrons in a resistor undergo random thermal motion that produces spontaneous voltage fluctuations across the resistor's terminals. These fluctuations represent a noise source that adds to any signal the resistor carries. The power spectral density of thermal noise is:

where k_B is Boltzmann's constant (1.38 × 10⁻²³ J/K), T is the absolute temperature in Kelvin, and R is the resistance in ohms. This spectrum is flat (white) over the frequency range of interest in most practical systems, meaning that thermal noise power is distributed equally across all frequencies up to very high values. The thermal noise power available from a resistor in a bandwidth B is N = k_B T B, and the noise figure of a receiving system quantifies how much the system degrades the signal-to-noise ratio above the thermal noise floor of the input.

Shot noise is a second fundamental form of technical noise arising from the discrete quantum nature of electrical charge. In any device where current flows by the movement of individual electrons or photons—diodes, transistors, photodetectors, laser sources—the arrivals of charge carriers are random events governed by Poisson statistics. Even when the average current is constant, the instantaneous current fluctuates around the mean because charge carriers do not arrive in a smooth continuous stream but in a random sequence of individual events. The shot noise power spectral density for a current I is:

where q is the elementary charge (1.6 × 10⁻¹⁹ C) and I is the mean current. Shot noise is especially significant in optical communication systems, where the detection of individual photons at a photodiode produces current pulses whose random timing generates shot noise that limits the minimum detectable optical power.

Flicker noise, also known as 1/f noise or pink noise, is a form of technical noise whose power spectral density is inversely proportional to frequency, so that it is most intense at low frequencies and falls off as frequency increases. Unlike thermal and shot noise, which have flat (white) spectra, flicker noise has a distinctly colored spectrum that makes it the dominant noise source in electronic devices at audio frequencies and below. It arises from slow fluctuations in the electrical properties of semiconductor devices—trapping and release of charge at defects and interfaces—and is characterized by its corner frequency, the frequency at which flicker noise power equals the white noise floor:

where K is a device-dependent constant. Flicker noise is critical in oscillators, phase-locked loops, and precision measurement instruments where low-frequency stability is required.

Intermodulation noise is produced when two or more signals pass through a nonlinear device or medium and mix to produce new frequency components at the sum and difference of the original frequencies and their harmonics. In a communication system carrying multiple channels, the nonlinear mixing of signals in amplifiers or optical fibers generates intermodulation products that fall within the frequency bands of other channels, acting as noise sources that degrade those channels. Third-order intermodulation products—generated at frequencies 2f₁ − f₂ and 2f₂ − f₁ for two input signals at f₁ and f₂—fall close to the original signal frequencies and are particularly troublesome because they cannot easily be filtered out. The third-order intercept point (IP3) characterizes how severely a device generates intermodulation products and is a key figure of merit for amplifiers and mixers in multi-carrier communication systems.

Phase noise is the random fluctuation of the instantaneous phase of an oscillator or carrier signal, causing the frequency of the signal to deviate randomly from its nominal value over time. In communication systems that use phase modulation (PSK, QAM), phase noise degrades the receiver's ability to correctly detect the transmitted symbol because it causes the received symbol to rotate randomly in the complex signal constellation. Phase noise is characterized by its power spectral density as a function of offset frequency from the carrier, expressed in dBc/Hz (decibels relative to the carrier power per hertz of bandwidth). High-quality oscillators with low phase noise are essential for maintaining the phase coherence required for reliable detection in modern high-order modulation schemes.

Quantization noise is introduced by analog-to-digital conversion, in which a continuous-amplitude analog signal is approximated by a discrete sequence of digital values. The difference between the true analog value and its nearest digital representation is the quantization error, which adds as noise to the digitized signal. For a uniform quantizer with n bits and full-scale range V_FS, the quantization noise power is:

Each additional bit of resolution reduces quantization noise power by a factor of 4 (6 dB improvement), which is why high-resolution audio and instrument digitizers use 16, 24, or even 32-bit converters to achieve very low quantization noise floors.

Mitigation of technical noise relies on a combination of strategies at different stages of the communication system. At the signal level, filtering removes noise outside the signal bandwidth, shielding prevents electromagnetic interference, and cryogenic cooling reduces thermal noise in sensitive receivers such as radio telescope front ends. At the system level, error-correcting coding introduces redundancy that allows the receiver to detect and correct errors caused by noise without requiring retransmission. At the link level, link budgets are calculated to ensure that the received signal power exceeds the noise floor by a sufficient margin (the link margin) to maintain the required bit error rate with acceptable probability under the expected channel conditions. Understanding the sources and statistical properties of technical noise is essential for designing communication systems that achieve reliable transmission at the rates and distances required by modern applications.