18 GHz microwave link: link budget versus rain margin and Fresnel clearance
A utility is designing a 12 km point-to-point microwave link between two hilltop sites to carry 155 Mbit/s. The chosen 18 GHz radio delivers +23 dBm at the antenna port; each end uses a 0.6 m dish with 38 dBi gain, 2.5 dB of feeder and connector loss, and a receiver threshold of −78 dBm for the planned 16-QAM modulation. The design rain rate exceeded for 0.01% of the year is 45 mm/h, for which the manufacturer's propagation data give a specific attenuation of 3.0 dB/km at 18 GHz; the effective path length may be taken as d/(1 + d/35) with d in km. The Site A antenna is at 70 m AMSL and Site B at 60 m AMSL. A ridge at 4.8 km from Site A has its top at 62 m AMSL, and the earth curvature bulge at that point is 2.0 m. Licensing permits either 18 GHz or 7 GHz; at 7 GHz the same dish gives 30 dBi, the receiver threshold is unchanged, and rain attenuation falls to about 0.6 dB/km.
1. Carry out the link budget and determine the fade margin available at 18 GHz.
Read the worked answer
Free-space loss. FSL = 92.45 + 20 log₁₀(f in GHz) + 20 log₁₀(d in km) = 92.45 + 20 log₁₀(18) + 20 log₁₀(12) = 92.45 + 25.11 + 21.58 = 139.14 dB
Received power. Pr = Pt + Gt + Gr − Lfeeder,A − Lfeeder,B − FSL = 23 + 38 + 38 − 2.5 − 2.5 − 139.14 = −45.14 dBm
Fade margin. FM = Pr − threshold = −45.14 − (−78) = 32.9 dB
So 32.9 dB of fade margin is available before the receiver reaches threshold at 16-QAM. Note that the dish gain is quoted at 18 GHz; the same physical dish at 7 GHz delivers 30 dBi, which is consistent with a 20 log(f) scaling plus a small aperture-efficiency difference, and the feeder loss is unchanged. Nothing in this step depends on the terrain, which is why the clearance must be checked separately - the budget can look healthy while the path is optically blocked.
2. Check the rain availability and the Fresnel-zone clearance at 18 GHz, and state whether the path as surveyed is usable.
Read the worked answer
Rain. Effective path length = 12/(1 + 12/35) = 12/1.343 = 8.94 km. Attenuation at 0.01% of the year = 3.0 × 8.94 = 26.8 dB, against a fade margin of 32.9 dB. The rain criterion is met with about 6 dB to spare, so 99.99% availability is achievable on rain alone.
Clearance. At 4.8 km from Site A, d₁ = 4.8, d₂ = 7.2. First Fresnel radius: F₁ = 17.3 √(d₁d₂/(f·D)) = 17.3 √(4.8 × 7.2/(18 × 12)) = 17.3 × 0.40 = 6.92 m The usual 60% criterion therefore requires 4.15 m of clearance.
Line-of-sight height at that point = 70 − (70 − 60)(4.8/12) = 70 − 4.0 = 66.0 m. Subtract the 2.0 m earth bulge: 64.0 m effective. The ridge tops out at 62 m, leaving 2.0 m clearance against 4.15 m required.
Verdict. The budget passes but the path does not. At 2.0 m the obstruction intrudes well inside the first Fresnel zone, so diffraction loss of several dB will appear immediately, and the link will fail whenever k falls below 4/3 and the bulge grows. The path as surveyed is not usable at 18 GHz.
3. Compare the 7 GHz option and recommend a course of action, with the numbers that justify it.
Read the worked answer
7 GHz budget. FSL = 92.45 + 16.90 + 21.58 = 130.93 dB. Received power = 23 + 30 + 30 − 5 − 130.93 = −52.93 dBm; fade margin = 25.1 dB. Rain = 0.6 × 8.94 = 5.4 dB, so rain is trivial, but at 25 dB margin multipath becomes the design case and must be checked by a multipath outage method for the climate zone.
Clearance at 7 GHz gets worse, not better. F₁ = 17.3 √(34.56/(7 × 12)) = 11.1 m; 60% = 6.66 m required. The lower frequency widens the zone, so the obstruction problem is aggravated.
Height needed. Required LOS height at 4.8 km = 62 + 2.0 + clearance.
- For 18 GHz: 68.15 m, so hA = (68.15 − 24)/0.6 = 73.6 m, i.e. raise the Site A antenna about 4 m.
- For 7 GHz: 70.66 m, so hA = 77.8 m, i.e. raise it about 8 m.
Recommendation. Raise the Site A mast by about 5 m (to 75 m AMSL, giving margin for survey error and sub-refraction) and retain 18 GHz, which already has 6 dB of rain margin and smaller dish loading. Raising by 8 m to keep 7 GHz buys nothing and costs more steel and licensing rework. Before committing: re-survey the obstruction, run the path profile at k = 4/3 and k = 2/3, check mast and tower loading for the extra wind area, and confirm with the regulator that the higher antenna stays inside the licensed height envelope. If the mast cannot be raised, a passive repeater or a different Site A is the fallback - not a lower frequency.