Evaluating Spatial Stream Selection and Guard Intervals in Heavy RF Reflection
Heavy RF reflection forces extended guard intervals and capped spatial streams matched to channel rank, preventing throughput collapse from inter-symbol interference.
Condition Number
Meaning
Matrix sensitivity under inversion or linear equation solving is measured by a dimensionless scaling ratio. In wireless communications, the condition number of the channel matrix indicates how easily a receiver can separate multiplexed data streams. High values represent an ill-conditioned channel where noise is amplified during the detection process.
Mathematical Computation
Calculating the metric requires dividing the largest singular value of the channel matrix by its smallest non-zero singular value. When these two values are close, the ratio approaches unity, representing an ideal orthogonal channel. Conversely, a large spread between the singular values produces a high ratio, signaling that some spatial paths are heavily attenuated compared to others.
This disparity complicates the matrix inversion needed to decouple the transmitted streams.
Physical Interpretation
Propagation environments with strong line-of-sight components and little scattering often produce high values. In these scenarios, the antenna paths are highly correlated, which makes the spatial signatures of the antennas almost identical. Multipath scattering helps to lower the ratio by decorrelating the paths.
Receiver Impact
Zero-forcing and minimum mean square error receivers experience severe noise enhancement when the channel matrix is poorly conditioned. In such situations, the demultiplexing algorithm amplifies both the signal and the thermal noise, degrading the post-processing signal-to-noise ratio. Lowering the modulation order is often required to maintain an acceptable bit error rate.