Introduction to Fading
Propagation models predict average path loss and the expected weakening of a signal over a distance, providing a basis for network design. They do not capture the full variability of a channel: a signal can be clear at one moment and weak or temporarily undetectable at the next, even after a small change in position. This variation in received signal strength is known as fading.
is variation in attenuation caused by the interaction of radio waves with the environment. A wave reflects, diffracts, and scatters, reaching the receiver over multiple paths whose superposition can reinforce or weaken the signal. Understanding the types of fading helps engineers select network planning, signal processing, and coding methods for the link conditions.
Interactive example: two-path interference
Use the controls to examine how path-length difference and reflection strength change the constructive and destructive sum at the receiver.
The chart traces instantaneous field versus path difference, while the diagram shows how direct and reflected components add in a two-path model. Changing path difference shifts the reflected component phase, so the same model produces alternating reinforcement and cancellation.
Large-Scale and Small-Scale Fading
Fading phenomena can be classified by the spatial scale over which they are observed. The distinction indicates whether a change comes from obstacles and average power over a longer path or from local path superposition. The mitigation method follows that cause.
Large-Scale Fading (Shadowing)
Large-scale fading describes variation in average received power over distances larger than local phase fluctuations. The scale depends on frequency, environment, and geometry. Large obstacles in the propagation path can block the signal and create an RF "shadow."
- Cause: Large obstructions such as buildings, hills, or dense forests create the shadowing component. As a user moves, for example, from an open street into an area blocked by a tall building, the average signal strength can decrease substantially. Path-loss models often represent this component as a random term in dB added to the mean loss, with variance depending on building type and averaging scale.
- Characteristics: These fluctuations are slow and occur as the user moves over distances larger than the local multipath scale. The statistical variation of this shadowing effect is often modeled by a log-normal distribution, with variance depending on environment and averaging scale. The mean follows the path-loss model, while obstacles can produce deeper and prolonged drops in a shadow.
- Mitigation: Large-scale fading is limited through network planning, base-station placement, and coverage control. Macro-diversity uses observations from multiple cells or transmission points to select or combine a more favorable path. A handoff (or handover) is a mobility procedure that transfers a connection between cells; it can use macro-diversity measurements, but it is not another name for macro-diversity. When a serving cell becomes weak in a shadow, the phone may be transferred to another cell.
Small-Scale Fading
This type of fading is observed over distances comparable with a wavelength and also with changes in time or frequency. A signal at has a wavelength of about 12.5 centimeters, so a displacement of about 6 centimeters can change path phases. Fade depth also depends on geometry, amplitudes, and bandwidth.
- Cause: The dominant mechanism of small-scale fading is constructive and destructive interference among copies arriving through multipath propagation. Motion, Doppler spread, and signal bandwidth also affect the observed variation.
- Characteristics: Amplitude and phase can fluctuate strongly after a small position change. The relevant spatial scale is set by wavelength and by the geometry of the scatterers, while temporal variation follows motion. A Rayleigh model is used when no dominant direct path is present, while a Rician model represents such a component; the K-factor describes its share, so the transition between models is continuous.
- Mitigation: Network planning determines transmission-point locations, while local fades also require signal processing at the transmitter and receiver. Systems use diversity, channel equalization, coding, and adaptive modulation. The selected method must follow the channel estimate and the time available for feedback and retransmission.
Small-Scale Fading Mechanisms
Small-scale fading can be described on two axes: time dispersion from multipath, called delay spread, and frequency variation from motion, called Doppler spread. Comparing these parameters with signal bandwidth and symbol time leads to flat or frequency-selective fading and to fast or slow fading.
Fading Based on Multipath Time Delay Spread
Delay spread can mean the time between the first and last significant component, or the root-mean-square delay spread computed from the power-delay profile. The fading type depends on this parameter relative to symbol time and on the signal bandwidth relative to coherence bandwidth.
1. Flat Fading
Condition: Occurs when the signal bandwidth is smaller than the channel's . Equivalently, the symbol period is longer than the characteristic delay spread.
Effect: All frequency components within the signal experience similar fading, including attenuation and phase shift. The amplitude fluctuates while the spectral shape is approximately preserved. A single channel coefficient can therefore describe the channel across that signal bandwidth.
Applies to: Narrowband communication systems, like pagers or some slow data links. The condition must be checked against the actual signal bandwidth and coherence bandwidth rather than inferred from a standard name.
2. Frequency-Selective Fading
Condition:Occurs when the signal bandwidth exceeds the channel's coherence bandwidth. Equivalently, the characteristic delay spread is not small relative to the symbol period.
Effect: Different frequency components experience different, partly correlated fading. The channel acts as a filter, attenuating parts of the spectrum. This distorts the received signal and can cause Intersymbol Interference (ISI), because delayed components of one symbol interfere with the next. Equalization, a cyclic prefix, or narrow subcarriers can limit the effect when system assumptions are satisfied.
Applies to: Many wideband communication systems, including Wi-Fi, 4G, and 5G configurations. The classification depends on signal bandwidth and the channel profile, so one radio may observe different categories at different positions or bandwidths.
Fading Based on Doppler Spread
describes the rate at which the channel changes and the frequency shifts of signal components. For speed v, the maximum shift is approximately f_D = v/λ, with a factor determined by arrival angle. It results from relative motion of the transmitter, receiver, or objects in the environment. The relation between coherence time and symbol duration determines whether fading is fast or slow. Coherence time is defined using a selected temporal-correlation threshold, so its value depends on that convention and on the motion model.
1. Fast Fading
Condition: The channel response changes on a time scale comparable to or shorter than a symbol. Approximately, the coherence time is shorter than the symbol period and Doppler spread is high.
Cause: High-speed user mobility, for example in a car on a highway, or a rapidly changing environment. At the same speed, a higher carrier frequency produces a shorter wavelength and a larger Doppler shift.
Effect: The channel can change during a symbol, causing distortion and channel-tracking errors. The severity depends on the modulation, observation time, and channel-estimation algorithm. At the same speed, a higher carrier frequency gives a shorter wavelength and a larger Doppler shift.
2. Slow Fading
Condition: The channel response changes much more slowly than a symbol. Approximately, the coherence time is longer than the symbol period and Doppler spread is low.
Cause: This condition can occur when the user is stationary, moving at pedestrian speed, or the environment changes slowly, depending on carrier frequency and symbol time.
Effect: The channel can be treated as nearly constant over many symbol periods. The main concerns are its amplitude and phase and the accuracy of channel estimation, rather than rapid variation within a symbol. The validity of this approximation depends on the scenario, carrier frequency, and observation interval.
Diversity Against Fading
If a signal can be weak at a particular position, frequency, or time, a system can improve reliability through .
The fundamental principle of diversity is to provide the receiver with multiple copies of the same transmitted signal that have experienced different, or ideally independent, fading conditions. It is based on the statistical observation that the probability of all copies being simultaneously in a deep fade is far lower than the probability of a single copy being in a deep fade. If one copy is weak, there is a good chance another one is strong. The benefit decreases when copies are strongly correlated, so antenna spacing, frequency separation, or time separation is chosen from channel statistics. There are several ways to obtain these independent signal copies.
1. Space Diversity
This is a common form of diversity. It uses two or more antennas separated by a distance chosen for the frequency and environment. Half a wavelength is a rule of thumb, while correlation also depends on arrival angle, polarization, scatterers, and the enclosure. The receiver can select or combine antenna signals to form a more stable result. Diversity gain is determined by branch independence, not by spacing alone. Techniques include:
- Selection Combining: A low-complexity method. The receiver measures the signal strength from all antennas and selects the one with the best signal for decoding.
- Maximal-Ratio Combining (MRC): A more optimal but complex method. The receiver coherently combines the signals using complex weights derived from channel estimates and noise levels. With correct weights, MRC maximizes SNR for this model; stale estimates or phase errors reduce the gain. MIMO can provide diversity, but it also includes spatial multiplexing.
2. Frequency Diversity
This involves transmitting the same information on at least two carrier frequencies. When their separation exceeds the coherence bandwidth under the selected criterion, the fades may be less correlated, so a deep fade on one component need not occur on the other. The technique consumes additional spectrum for one information stream and requires frequency synchronization.
3. Time Diversity
This involves transmitting the same information at different times. If the separation is longer than the coherence time, successive copies may experience less correlated fading. It is often combined with error-correction coding and . Interleaving separates adjacent bits in time, so a fade that creates a burst of errors can produce dispersed errors after de-interleaving, which a code can correct more effectively. The delay must remain acceptable for the application, and the code rate determines how many errors can be recovered.
4. Polarization Diversity
This technique uses two antennas with orthogonal polarizations, for example vertical and horizontal. Reflections can alter signal polarization, and coupling between branches depends on geometry and depolarization. Providing two polarized versions increases the chance that at least one will be received strongly as the polarization state changes through the channel.
Fading Mitigation Techniques
Beyond diversity, wireless systems use signal processing matched to the distortions caused by fading.
- Channel Equalization: With frequency-selective fading, an equalizer limits Intersymbol Interference (ISI). It is a digital filter that uses a channel estimate to compensate for the channel before the receiver makes a decision. Its coefficients must track channel changes and account for noise. Training symbols or pilot subcarriers provide the estimates used by the receiver.
- Orthogonal Frequency Division Multiplexing (OFDM): OFDM is widely used in Wi-Fi, 4G, and 5G variants. It divides a wideband channel into many narrow, orthogonal subcarriers, each of which can approximate flat fading. A cyclic prefix and receiver equalization limit multipath effects when their parameters match the channel assumptions. The number of subcarriers, spacing, and prefix length depend on the standard and configuration.
- MIMO Systems: MIMO (Multiple-Input Multiple-Output) uses multiple antennas at the transmitter and receiver. It can improve reliability through diversity or send independent streams in the same band through spatial multiplexing. The selected mode depends on channel rank and correlation, SNR, interference, and channel estimation. Spatial multiplexing requires several sufficiently independent paths; with high correlation, a system usually favors diversity. The channel matrix and its singular values indicate how many streams can be supported at a useful SNR.