A transmit‑side signal processing method that multiplies a vector of symbols by a linear matrix (the precoder) across transmit antennas or channels so as to shape the transmitted signals’ spatial or multi‑channel structure to improve reception, reduce inter‑user interference, or enable spatial multiplexing under given power constraints.

Principle

A linear precoder implements a linear mapping from symbol space to antenna signals: the choice of precoding matrix governs inter‑user interference and per‑stream effective channel gains, and optimal linear precoders under common metrics (e.g., zero‑forcing, MMSE, maximum‑ratio transmission) trade off interference suppression, noise amplification, and transmit power distribution.

Demonstration

Situation: A base station with four transmit antennas serves two single‑antenna users simultaneously. Recognition: The transmitter has current CSI for both users. Action: It computes a zero‑forcing precoder that projects each user’s symbol into the subspace orthogonal to the other’s channel and scales to meet the power budget. Consequence: Inter‑user interference is (ideally) suppressed at the receivers and each user observes an effective single‑stream channel with improved SINR; with imperfect CSI, residual interference appears.

Misapplication

Assuming linear precoding removes all interference independent of CSI quality: the mistake is to ignore that linear precoding performance depends on the accuracy and timeliness of channel knowledge and on the number of transmit degrees of freedom relative to served streams.

Consequence

Proper linear precoding increases spectral efficiency by enabling simultaneous transmission to multiple users and improves per‑user SINR; improper design or outdated CSI leads to residual interference, noise amplification, and reduced throughput or fairness problems.

Reversal

When channels are highly correlated or the number of served streams approaches or exceeds the transmitter’s spatial degrees of freedom, linear precoding cannot fully cancel interference and nonlinear or user‑scheduling strategies (or reducing the number of streams) may be required; at high SNR and strong interference, nonlinear precoding can yield strictly better rates.

Boundary

Clearly within: a transmitter applying a matrix multiply to symbol vectors across antennas (e.g., ZF, MRT, MMSE precoders). Boundary case: hybrid precoding where analog phase networks impose constraints on the linear transform. Clearly outside: per‑stream power control without spatial mixing or receiver‑side equalization alone.

Semantic Tension

Linear Precoding ↔ Nonlinear Precoding — linear methods trade lower complexity and tractable optimization against potentially suboptimal rates compared with nonlinear schemes that can manage interference more effectively at higher complexity.

Synthesis

Linear precoding is the practical family of transmit transforms that use linear algebra to shape multi‑antenna signals; its utility depends on CSI quality, available spatial degrees of freedom, and the chosen performance metric (e.g., sum‑rate, fairness, error rate).