Radio Site, Coverage, and RF Propagation Analysis
What this domain tests
Domain 4 moves from link budgets to coverage acceptance: how sites are designed, how coverage is predicted and measured, and how in-building and simulcast problems are diagnosed.
Core rules
- Public safety coverage is normally specified as 95% of the service area at Delivered Audio Quality 3.4, measured separately for talk-out (site to portable) and talk-in (portable to site).
- Design inputs: HAAT, antenna gain and downtilt, ERP, receiver sensitivity, noise figure, combiner and duplexer loss, and site noise (desense).
- Propagation models, from free space through Okumura-Hata, COST 231 and Longley-Rice, differ in how they handle clutter and terrain. Predictions must be verified by drive or grid testing.
- Fresnel zone clearance matters: obstructing the first zone degrades a path even when optical line of sight exists.
- Simulcast cells rely on GPS-based timing alignment; overlap regions are engineered with LSM/CQPSK so delay spread is tolerated rather than avoided.
- In-building coverage comes from bi-directional amplifiers or DAS. NFPA 1225 and the International Fire Code set DAQ-based acceptance, and BDAs must not oscillate or raise site noise.
- Site infrastructure counts toward coverage availability: battery and generator backup, grounding, lightning protection, tower lighting and alarm monitoring.
Worked example
A 450 MHz site runs 100 W ERP (50 dBm) into a portable with 0 dBi gain. Free-space loss at 5 km is 32.44 + 20log(450) + 20log(5), about 99.5 dB, so the portable sees -49.5 dBm. With -116 dBm sensitivity the margin is roughly 66 dB, enough for a 25 dB clutter allowance. At 20 km the loss rises about 12 dB to 111.5 dB, still -61.5 dBm; terrain shadowing, not distance, is what actually breaks coverage.
Traps
Presenting a prediction as an acceptance result, confusing talk-in failures with talk-out failures (talk-in usually points to site desense or receiver problems), assuming more ERP always fixes coverage, and ignoring delay spread in simulcast overlap.