Multiple-Input Multiple-Output (MIMO)

MIMO and spatial channels

In radio communication, occurs when reflections and other paths deliver copies of a signal to the receiver. The copies can have different delays and phases, so a single-antenna link may experience interference, fading, and distortion.

MIMO (Multiple-Input Multiple-Output) uses multiple antennas at the transmitter and receiver to process those paths jointly. Depending on the channel and the selected transmission strategy, the system can send independent spatial layers, add diversity, or form beams within the same frequency band. MIMO is a multi-antenna implementation of Space-Division Multiplexing (SDM) when independent spatial streams are transmitted.

How MIMO works: the signal model

A MIMO system is often denoted as an $M \times N$ system, with $M$ transmit antennas and $N$ receive antennas. For a narrowband linear model, the received vector can be written as $y = Hx + n$ , where $H$ is an $N \times M$ channel matrix, $x$ is the transmitted vector, and $n$ represents noise. If $L$ independent spatial layers are transmitted, then $L \leq \operatorname{rank}(H) \leq \min(M,N)$ .

Data and layer mapping: The payload is divided into one or more coded streams, called layers in many radio standards. The selected number of layers depends on the estimated channel and its usable rank.
Precoding and transmission: Each layer is modulated and encoded. A precoder can map the layers to the transmit antennas, which then transmit the resulting signals over the same time and frequency resources.
Radio channel: The transmitted signals propagate through the same radio environment. Reflections and path differences produce a channel matrix that mixes the layers at the receive antennas.
Reception: The receive array samples the mixture. Each antenna observes a different linear combination because its position and radio path differ.
Channel estimation and separation: Reference signals allow the receiver to estimate . In systems with CSI feedback, the transmitter uses reported channel information, such as a rank or precoding indication, to select the layer count and precoder. Digital Signal Processing (DSP) then uses the channel estimate in a detector or equalizer to separate and decode the layers.
Recombination: The decoded layers are ordered and combined to recreate the original data stream.

MIMO operating modes

A MIMO system selects a transmission strategy according to the channel and the target. The main objectives are higher throughput, greater reliability, and improved received signal quality. Choosing a strategy requires balancing these objectives.

1. Spatial multiplexing (increased throughput)

This mode sends independent layers over the same time and frequency resources. The number of useful layers is limited by the channel rank and cannot exceed the smaller of the transmit and receive antenna counts. With a well-conditioned channel and sufficient SNR, additional layers can raise throughput without additional bandwidth; the gain is not a fixed capacity multiplier. Spatial multiplexing is used in high-speed Wi-Fi and in 4G and 5G radio systems.

2. Spatial diversity (increased reliability)

This mode adds redundant observations of a data stream across antennas and, in some schemes, across time. Space-time coding can create these observations without requiring independent payload layers. The receiver combines the observations, and the reliability gain depends on the independence of the channel paths; strongly correlated paths provide less diversity.

3. Beamforming (increased range and quality)

In this mode, precoding weights adjust the phase and amplitude at the transmit antennas to shape the received field. With channel information, the precoder can strengthen the desired signal or place a spatial null toward an interferer. The resulting SINR and coverage benefit depend on channel knowledge, array geometry, and propagation conditions. Beamforming is used in 5G NR, including deployments referred to as Massive MIMO.