Spatial Multiplexing

The primary benefit of spatial multiplexing is a linear increase in data rate without requiring additional bandwidth or transmit power. The multiplexing gain is defined as the rate at which capacity scales with the signal-to-noise ratio (SNR). In an ideal rich-scattering environment, the maximum number of independent streams is limited by min(N_t, N_r), where N_t is the number of transmit antennas and N_r is the number of receive antennas. This allows a system to achieve a multiplexing gain equal to the rank of the channel matrix.

The effectiveness of spatial multiplexing is governed by the rank and condition number of the MIMO channel matrix H. The rank determines the maximum number of independent spatial streams that can be supported. The condition number, the ratio of the largest to smallest singular value of H, indicates how easily the streams can be separated. A high condition number signifies a poorly conditioned channel where noise amplification during linear detection (like Zero-Forcing) severely degrades performance, making non-linear techniques like Successive Interference Cancellation (SIC) necessary.

Unlike diversity techniques that transmit redundant information, spatial multiplexing creates inter-stream interference (ISI) at the receiver. Each receive antenna captures a superposition of all transmitted streams, weighted by the channel coefficients. The core challenge of a spatial multiplexing receiver is to separate these overlapping signals. Linear detectors like Minimum Mean Square Error (MMSE) balance interference suppression with noise enhancement, while optimal Maximum Likelihood Detection (MLD) exhaustively searches for the most probable transmitted vector but incurs exponential computational complexity.

Spatial multiplexing requires a rich multipath environment to decorrelate the spatial channels. If antenna elements are too closely spaced or the propagation environment lacks sufficient scatterers, the channel coefficients become highly correlated. This spatial correlation reduces the effective rank of the channel matrix, collapsing the available spatial degrees of freedom and preventing the formation of independent streams. A Rayleigh fading model, which assumes no dominant line-of-sight path, is often ideal for achieving full multiplexing gain.

Spatial multiplexing can operate in two modes:

• Open-Loop: The transmitter has no knowledge of the channel state information (CSI). It transmits independent streams equally across all antennas. This is robust but suboptimal in poor channel conditions.
• Closed-Loop: The receiver feeds back CSI to the transmitter, which uses precoding to weight the streams optimally. By performing Singular Value Decomposition (SVD) on the channel matrix, the MIMO channel is decomposed into parallel, non-interfering eigenmodes, maximizing capacity and simplifying the receiver.

In wideband channels, frequency-selective fading causes inter-symbol interference. Spatial multiplexing is combined with Orthogonal Frequency-Division Multiplexing (OFDM) to form MIMO-OFDM. This transforms a frequency-selective wideband channel into multiple parallel flat-fading narrowband subcarriers. Spatial multiplexing is then applied independently on each subcarrier, allowing per-subcarrier precoding and detection. This architecture is the foundation of modern standards like 4G LTE, 5G NR, and Wi-Fi 6/7.

The linear increase in data rate capacity achieved by transmitting independent data streams over multiple spatial paths. This gain scales directly with the minimum number of transmit and receive antennas (min(Nt, Nr)). For example, a 4x4 MIMO system can theoretically quadruple throughput compared to a SISO system under ideal, rich-scattering channel conditions. The gain is fundamentally limited by the rank of the channel matrix and the degree of spatial correlation between antenna elements.

A matrix factorization method that decomposes the MIMO channel into parallel, non-interfering eigenmodes. By applying SVD to the channel matrix H = UΣV^H, the transmitter can use V as a precoding matrix and the receiver can use U^H to create independent spatial pipes. Each eigenmode's gain is given by the corresponding singular value in Σ. This enables capacity-achieving eigen-beamforming, where power is optimally allocated across spatial streams using water-filling algorithms.

A beamforming technique applied at the transmitter that weights the signal across multiple antennas to maximize signal power at the intended receiver while minimizing interference to others. In spatial multiplexing, precoding maps data streams to antenna elements using a precoding matrix derived from Channel State Information (CSI). Common approaches include:

• Codebook-based precoding: Selecting from predefined matrices (PMI feedback)
• Non-codebook precoding: Computing optimal weights from SVD or reciprocity

A metric describing the sensitivity of a MIMO channel matrix to inversion, defined as the ratio of the largest to smallest singular value (σ_max/σ_min). A high condition number indicates a poorly conditioned channel where spatial streams experience vastly different gains, limiting multiplexing performance. This often occurs in keyhole channels or environments with strong line-of-sight dominance. A condition number near 1 (0 dB) represents an ideal, well-conditioned channel where all spatial streams have equal gain.

A non-linear detection technique that iteratively decodes the strongest signal stream, subtracts its reconstructed contribution from the received signal, and repeats the process for remaining streams. This approach achieves higher throughput than linear receivers like ZF or MMSE by exploiting the capacity-achieving properties of decision feedback. In V-BLAST architectures, optimal ordering of stream decoding further maximizes the post-detection SNR for each layer.

The process of characterizing the propagation channel's impulse response using known reference signals or pilot symbols. Accurate channel estimation is critical for spatial multiplexing because both precoding and detection rely on precise knowledge of the channel matrix H. Techniques include:

• Least Squares (LS): Simple but noise-sensitive
• Minimum Mean Square Error (MMSE): Balances estimation accuracy and noise suppression
• Compressed sensing: Exploits channel sparsity in massive MIMO systems