How does multipath propagation behave differently at millimeter wave versus sub-6 GHz?
mmWave Multipath Behavior
The multipath behavior at mmWave has implications for channel modeling, system design, antenna architecture, and deployment strategy that are fundamentally different from sub-6 GHz.
Frequently Asked Questions
Do ground reflections help at mmWave?
Yes, but less than at lower frequencies: the ground reflection at mmWave depends on the surface type: smooth asphalt: reflection coefficient ≈ 0.5-0.7 at glancing angles (-3 to -1.5 dB). This provides a viable reflected path that can fill coverage gaps. Rough terrain or grass: reflection coefficient ≈ 0.1-0.3 (the surface roughness is significant compared to the mmWave wavelength; the reflection is scattered). At sub-6 GHz: even rough ground provides a reasonable reflection (the wavelength is large compared to the surface roughness). In the two-ray ground-reflection model: the path loss at mmWave shows the classic 40 dB/decade roll-off at distances beyond the breakpoint distance: d_bp = 4 × h_tx × h_rx / lambda. For h_tx = 10 m, h_rx = 1.5 m, f = 28 GHz: d_bp = 4 × 10 × 1.5 / 0.0107 = 5607 m. Since mmWave cells are much shorter (< 400 m): operation is always within the breakpoint distance, and the two-ray model is less relevant.
How does foliage affect mmWave?
Vegetation causes significant attenuation at mmWave: single tree (in-leaf): 10-20 dB loss through the canopy at 28 GHz. Dense foliage (row of trees): 20-40 dB loss. The attenuation is caused by scattering from the leaves and branches (the leaf size is comparable to the wavelength). Seasonal variation: deciduous trees in winter (no leaves): 2-5 dB loss (only trunk and branch scattering). Same trees in summer (full canopy): 10-20 dB. For mmWave deployment: avoid placing small cells behind or within dense tree cover. Street-level small cells on the same side of the street as trees may have acceptable LOS performance (the antennas are typically above the tree line).
Is rain a problem for mmWave?
Rain attenuation at 28 GHz: approximately 1 dB/km in light rain (5 mm/hr), 5 dB/km in heavy rain (25 mm/hr), and 10 dB/km in very heavy rain (50 mm/hr). For a typical mmWave small cell with 200 m range: maximum rain loss = 10 × 0.2 = 2 dB (very heavy rain). This is small compared to the link margin (typically 10-15 dB). Rain is NOT a significant problem for mmWave small cells. However: at 39 GHz, rain attenuation is approximately 50% higher than at 28 GHz. And for longer links (microwave backhaul at 70-80 GHz): rain can be significant (5-10 dB at 1 km in heavy rain).