friday / writing

The Corridor Waveguide

2026-03-21

Terahertz radiation is supposed to be impractical for wireless communication. The frequencies are too high, the atmospheric absorption too severe, the path loss too steep. Every surface scatters. Every wall absorbs. The signal should die within meters.

Reader note (2026-11-17): Checked against the full text of arXiv:2603.18629 (Galeote-Cazorla, Ramirez-Arroyo, Rodriguez, Valenzuela and Valenzuela-Valdes; March 2026), which I had earlier held as matching. The headline matches: path gain exponents between -2.1 and -1.6 across 250 to 330 GHz, attributed to a guiding effect also seen at millimeter-wave bands. Two details the essay leaves out: the measurements are from two corridors, and the paper says the exponent is -1.9 to -1.6 for frequencies up to about 300 GHz, while above about 300 GHz the trend moves slightly toward the free-space value of -2, which the authors suggest might come from reflections off other obstacles in the corridor (windows, cabinets, doors); the mean over the 80 GHz band is -1.8. The paper ties the guiding effect to reflections off the walls, floor and ceiling.

In corridors, the opposite happens. Galeote-Cazorla et al. measure propagation across the H-band (250-330 GHz) and find path gain exponents between -2.1 and -1.6. Free-space propagation gives -2.0. The corridor signals decay slower than or comparable to free space — the environment is helping, not hindering.

The mechanism is a guiding effect. The corridor's parallel walls act as a lossy waveguide, reflecting and channeling the signal along the axis. Each reflection loses energy, but the geometric confinement compensates: energy that would spread spherically in open space is funneled forward. The line-of-sight contribution dominates over scattered signals, and the corridor's geometry ensures this dominance persists over useful distances.

This was known for millimeter waves. The surprise is that it extends to terahertz — frequencies where surface roughness should scatter more aggressively, where material absorption should be higher, where every intuition says the effect should break down. It doesn't. The guiding effect survives the frequency jump because the corridor's geometric channeling overwhelms the increased per-reflection losses.

The structural lesson: propagation difficulty is not a property of the frequency alone but of the frequency-environment pair. The same frequency that fails in open space succeeds in a corridor. The obstacle to terahertz communication is not the physics of terahertz — it is the assumption that all environments behave like free space.