The monitoring platform we had helped bring to market on CAT-M1 was later extended to a new deployment: equipment installed in underground utility vaults in dense urban areas, accessed through street-level manhole covers. This changed the cellular problem fundamentally. Above ground, the question had been which standard and which carriers to certify. Below ground, the first question was whether a signal could reach the device at all.

We ran Verizon CAT-M1 and Verizon NB-IoT units in parallel at underground vault sites in the San Francisco Bay Area, across multiple vault configurations. LTE connectivity was achievable at every site. NB-IoT was significantly more reliable than CAT-M1 across the test locations, and the lab-measured sensitivity advantage translated directly into real-world performance where CAT-M1 was marginal or inconsistent.

The testing also produced a result we hadn’t expected. Higher-frequency bands at 1900 MHz outperformed 700 MHz inside the vaults, which runs against the usual rule that lower frequencies penetrate solid materials more effectively. The explanation is that underground vault access isn’t primarily through solid walls. RF enters through the apertures of grate covers and ventilation openings, and for openings of that geometry, shorter wavelengths (higher frequency bands) pass through more effectively than longer ones. Carrier and band selection for underground IoT devices, it turns out, should be driven by vault geometry rather than by the assumptions that apply above ground.

‟customer comment”

— author