Rayleigh Fading

The defining mathematical property of Rayleigh fading is that the complex channel impulse response is a circularly symmetric complex Gaussian random process with zero mean. This arises from the Central Limit Theorem: the received signal is the superposition of many independent scattered multipath components, each with random phase and amplitude. The in-phase (I) and quadrature (Q) components are independent and identically distributed (i.i.d.) Gaussian random variables with zero mean and equal variance. This statistical foundation makes Rayleigh fading the worst-case scenario for wireless links where no stable, direct path exists.

While the I and Q components are Gaussian, the signal envelope—the magnitude of the complex baseband signal—follows a Rayleigh probability density function (PDF). The phase is uniformly distributed over [0, 2π]. Key statistical metrics include:

• Mean envelope: proportional to the root-mean-square (RMS) voltage
• Median envelope: approximately 0.939 times the RMS value
• Mode: the most probable value, which is less than the mean This distribution accurately models the rapid amplitude fluctuations observed in dense urban environments or heavily forested areas where a direct path is obstructed.

Rayleigh fading channels are characterized by deep fades—periods where the instantaneous signal power drops significantly below the average. The received signal-to-noise ratio (SNR) follows an exponential distribution. The probability that the SNR falls below a required threshold (the outage probability) is a critical design parameter:

• A 20 dB fade margin provides roughly 99% reliability
• A 30 dB fade margin provides roughly 99.9% reliability
• Without diversity techniques, deep fades cause burst errors that dominate link performance This behavior drives the need for diversity combining, interleaving, and forward error correction in modern wireless standards.

When the receiver, transmitter, or scatterers are in motion, Rayleigh fading becomes time-selective. The classic Jakes' Doppler spectrum—also called the U-shaped spectrum—describes the power spectral density of the fading process for isotropic scattering. The maximum Doppler shift is given by f_d = v/λ, where v is velocity and λ is wavelength. The coherence time (T_c), approximately 0.423/f_d, defines the duration over which the channel impulse response remains correlated. If a symbol period exceeds T_c, the channel is fast fading, causing irreducible error floors unless robust pilot-based estimation is employed.

Multipath components arrive at the receiver with different time delays, characterized by the power delay profile (PDP). The RMS delay spread (σ_τ) quantifies this temporal dispersion. The coherence bandwidth (B_c) is inversely proportional to σ_τ (B_c ≈ 1/(5σ_τ)). When the signal bandwidth exceeds B_c, the channel becomes frequency-selective: different frequency components experience uncorrelated fading. This causes intersymbol interference (ISI) and transforms the channel into a tapped-delay-line filter. Equalization techniques like OFDM with cyclic prefixes are standard countermeasures.

Two second-order statistics quantify the temporal dynamics of Rayleigh fading:

• Level Crossing Rate (LCR): The expected rate at which the signal envelope crosses a specified amplitude threshold in the positive-going direction. It is directly proportional to the maximum Doppler shift and the threshold level normalized to the RMS value.
• Average Fade Duration (AFD): The average time the envelope remains below a given threshold. It is the ratio of the outage probability to the LCR. These metrics are essential for designing packet lengths, timeout intervals, and automatic repeat request (ARQ) protocols to ensure data is not transmitted during prolonged deep fades.