Analytical Multi-User MIMO Channel Modeling: Subspace Alignment Matters

For a receiver that observes a mixture of desired and interfering multiple-input multiple-output (MIMO) transmissions, not all interference has the same severity: its effect is modulated by the degree of alignment between the eigenspaces of the desired and undesired channel matrices. An analytical channel model is proposed to account for this effect. Two metrics of eigenspace compatibility are studied, and a method to generate channels of a given degree of compatibility is derived. The resulting model is parameterized using radio channel measurement data, and an implementation recipe for immediate use of the model is provided.