Introduction to the Wireless Channel
Propagation models describe average signal strength and path loss over a distance, providing a basis for network planning. They do not capture the full variability of a real wireless environment. A radio wave reflects, diffracts, and scatters, so its path can differ substantially from the direct line between transmitter and receiver.
The includes space, obstacles, reflectors, and signals competing with the transmission. These phenomena change the amplitude, phase, and arrival time of the signal. Three characteristics are especially useful in this context: multipath propagation, interference, and noise. Separating them helps engineers choose a remedy for the mechanism that degrades the link, while the measured result depends on bandwidth, antenna pattern, receiver implementation, and observation interval. These parameters support decisions about equalization, filtering, channel reuse, and link adaptation.
Multipath Propagation
In an ideal free space environment, the receiver would capture a single, clean copy of the transmitted signal. In reality, this almost never happens. The transmitted electromagnetic wave radiates outward and interacts with the surrounding environment. It reflects off buildings, the ground, and vehicles, diffracts around sharp corners, and scatters off smaller objects like trees and lampposts.
As a result, the receiver's antenna captures multiple copies that have traveled paths of different lengths. This phenomenon is called . Each path contributes its own delay, attenuation, and phase shift. The received signal is the vector sum of these copies, and the result also depends on their amplitudes, the signal bandwidth, and the observation time. For a path difference Δl, the phase change is approximately Δφ = 2πΔl/λ, so the same geometry has a different effect at another frequency.
Constructive and Destructive Interference (Fading)
When multiple copies arrive at the receiver, they interfere with one another. The result depends on their relative phases, amplitudes, and delays, which can change with position and frequency.
- Constructive Interference:If the peaks of two waves arrive at the same time (they are "in phase"), their amplitudes add together, resulting in a stronger signal than either of the individual copies.
- Destructive Interference:If the peak of one wave arrives at the same time as the trough of another wave (they are "out of phase"), their amplitudes cancel each other out, producing a much weaker signal or, in the limiting case, near cancellation and a "dead spot".
This rapid fluctuation in signal strength as a receiver moves through the field is known as small-scale fading. At 2.4 GHz, the wavelength is about 12.5 cm, so a displacement of a few centimeters changes path phases. The depth of a fade also depends on geometry, amplitudes, and signal bandwidth, so distance alone does not determine the result.
Intersymbol Interference (ISI)
Another consequence of multipath is that delayed copies spread the channel response in time. Digital communication works by sending discrete symbols, where each symbol represents one or more bits and lasts for a specific duration.
Imagine that the transmitter sends a sharp pulse representing a '1' followed by silence representing a '0'. The receiver gets the shortest-path copy first and weaker reflected copies later. If their energy overlaps the next symbol, it creates . The effect on bit error rate depends on the modulation, coding, and receiver algorithm.
The extent of this time dispersion is described by delay spread. It can be estimated as the time between the first and last significant component, or as the root-mean-square delay spread computed from the power-delay profile. RMS delay spread weights each path by its relative power, so it often describes channel impact better than an extreme component alone. A large value relative to the symbol period increases the risk of ISI, while the exact effect depends on the response shape and receiver.
Frequency-Selective Fading
The effects of constructive and destructive interference are highly dependent on the phase of the waves, which in turn depends on their frequency. This means that for a wideband signal (a signal that occupies a range of frequencies), the multipath interference will affect different frequencies within that signal in different ways.
This results in . The channel can act as a filter, attenuating one part of the spectrum more than another. A wide signal can therefore contain frequencies with different gains, while notch depth depends on the multipath profile and signal bandwidth. Coherence bandwidth is the frequency interval over which gain is considered similar under a chosen correlation criterion. It is inversely related to delay spread, with a constant that depends on the definition. The selected correlation threshold should be stated when reporting a value, because different thresholds produce different estimates. When signal bandwidth is below this interval, a flat-channel approximation may be used. When it is larger, the system needs techniques that handle different subband gains.
Interference in the Wireless Channel
Beyond the signal's multipath effects, it must also contend with external signals. This is called . While noise is random, interference is structured. It may be another user's or system's communication signal that is treated as an unwanted disturbance by your receiver.
Co-Channel Interference (CCI)
This is interference from other transmitters using the same frequency channel as the desired signal. Frequency reuse can increase cellular capacity, while the importance of CCI depends on distance, load, cell geometry, antenna directionality, receiver algorithms, and scheduling. A user at a cell edge may receive a weaker serving signal and a similarly strong signal from a neighboring cell using the same frequency. Network planning, antenna tilting, power control, coordination, and receiver processing limit this interference.
Adjacent-Channel Interference (ACI)
Neither a transmitter nor a receiver has an ideal rectangular response. Transmitter out-of-band emissions and limited receiver selectivity create ACI. In the 2.4 GHz Wi-Fi band, Channel 6 is centered near 2437 MHz and Channel 7 near 2442 MHz, while actual overlap depends on channel width, power, emission masks, filters, and receiver sensitivity. A strong adjacent signal can therefore degrade a weak signal. Transmit filters limit leakage and receiver filters reject energy outside the intended band.
Interference from Other Systems
This is particularly relevant in the unlicensed ISM band, where Wi-Fi shares spectrum with Bluetooth, cordless phones, and baby monitors. A microwave oven operates near , but its effect depends on shielding, distance, duty cycle, and radiation outside the enclosure. In a crowded environment these conditions can increase Wi-Fi retransmissions and reduce Bluetooth audio quality.
Noise Sources
Distinct from structured interference, is a random component from the receiver chain and the surrounding radio environment. It appears as hiss or static when a receiver is tuned away from a station. Its level depends on bandwidth, temperature, receiver noise figure, antenna, and the environment. Noise comes from many sources:
- Thermal Noise (Johnson-Nyquist Noise): It is generated by the random thermal motion of charge carriers in conductive materials such as wires, resistors, and transistors above absolute zero. For an ideal resistor, noise power over bandwidth B is N = kTB, where k is the Boltzmann constant and T is absolute temperature. A real receiver adds noise according to its noise figure, so bandwidth and the first receiver stage limit sensitivity. Filtering reduces integrated noise power only by reducing the effective bandwidth.
- Atmospheric Noise:It is caused by natural phenomena in the Earth's atmosphere, most notably lightning discharges. Lightning generates strong bursts of radio energy across a wide range of frequencies. This noise is most dominant in lower frequency bands such as AM radio and is usually a smaller factor for higher-frequency systems such as Wi-Fi and cellular networks, although its level depends on location, time, and antenna bandwidth.
- Extraterrestrial Noise: The Sun emits substantial radio noise, whose received level depends on observation conditions, and distant stars and galaxies also emit radio waves that contribute to the background noise floor. This cosmic noise can limit sensitive applications such as radio astronomy and deep-space communication, in which antenna direction and observation time affect the received level.
- Man-Made Noise: This refers to unintentional radiation from electrical and electronic devices that are not designed to be transmitters. Sources include sparks from car ignition systems, electric motors, switching power supplies in computers, and faulty power lines. This type of noise is typically strongest in dense urban and industrial areas, where many emitters share the same receiver bandwidth.
SNR and SINR at the Receiver
To successfully receive a wireless signal, the receiver must be able to distinguish the desired signal from all these competing disturbances. A common metric is the . It is a simple ratio comparing the power of the desired signal () to the power of the background noise ().
Expressed in decibels, the formula is:
A higher SNR increases the separation between signal and noise, while the resulting Bit Error Rate (BER) also depends on modulation, coding, synchronization, and receiver implementation. With other assumptions fixed, a higher SNR generally permits a lower BER. For an ideal Gaussian channel, the limiting relation is C = B log2(1 + SNR), while a practical system achieves less because of coding overhead, spectral constraints, and its target BER.
SINR in the Presence of Interference
When signals from other transmitters are present, link quality is also described by the . This metric compares desired-signal power with the sum of interfering signals () and background noise ().
A high SINR is useful, but achievable throughput also depends on bandwidth, spectral efficiency, coding, and link adaptation. MIMO can increase the number of independent streams, while beamforming can improve gain or reduce interference. Either technique may improve SINR, and its effect depends on the channel and system configuration. In a shared system, power control, scheduling, and retransmissions also matter, because a better SINR sample does not guarantee a sustained data rate. SINR is a power ratio in linear units; implementation margins and measurement windows are needed when it is reported in decibels.