How does below-rooftop propagation differ from macro-cell?

In macro-cell deployments (antenna above rooftop, 25-50 m height), the dominant propagation path goes over building rooftops via multiple diffraction. The Walfisch-Ikegami and COST 231 models describe this regime. Path loss depends on building height distribution, street width, and number of diffracting edges. In below-rooftop deployments (antenna at 3-10 m, below average 15-25 m roofline), the signal cannot propagate over buildings. Instead: 1) LOS street canyon: signal travels along the street like a waveguide. Path loss exponent n ≈ 2.0-2.5 (near free-space). Range: 100-300 m. The buildings create a multipath-rich environment with strong reflections from facades. 2) NLOS around corners: signal must diffract around building corners. Path loss exponent n ≈ 3.0-4.5. Additional 15-25 dB loss at each corner. Range limited to 100-200 m. 3) Through-building: for indoor coverage, signal must penetrate exterior walls. Building entry loss: 10-30 dB depending on material (glass: 3-8 dB, concrete: 15-25 dB, low-E glass: 25-40 dB). 4) Scattering: vehicles, trees, and street furniture create time-varying multipath. Delay spread: 30-300 ns (BRT) versus 500-3000 ns (macro-cell).

What propagation models apply to below-rooftop scenarios?

The primary models used for BRT planning: 3GPP TR 38.901 UMi (Urban Micro) Street Canyon: LOS: PL = 32.4 + 21·log₁₀(d_3D) + 20·log₁₀(f_c). Valid: 10 m < d < 5 km, 0.5-100 GHz. Shadow fading σ = 4.0 dB. NLOS: PL = 35.3·log₁₀(d_3D) + 22.4 + 21.3·log₁₀(f_c) - 0.3·(h_UT - 1.5). Shadow fading σ = 7.82 dB. LOS probability: P_LOS = min(18/d, 1) × (1 - exp(-d/36)) + exp(-d/36). WINNER II B1 (Urban Micro-cell): Similar structure with different coefficients. Adds O-to-I (outdoor-to-indoor) sub-scenario. ITU-R P.1411: Site-specific model for short-range outdoor. Includes street canyon waveguide model and over-rooftop diffraction for mixed scenarios. For 5G NR at 28 GHz (mmWave), BRT propagation is particularly challenging: oxygen absorption adds 0.01 dB/m, rain attenuation adds 1-10 dB/km (heavy rain), and foliage attenuation adds 10-20 dB per tree canopy. Beamforming (64-256 element arrays) is essential to overcome the higher path loss.

What are the antenna requirements for below-rooftop small cells?

Below-rooftop antenna design differs significantly from macro-cell: Height: 3-10 m (lamppost, utility pole, building facade). Must be inconspicuous for municipal permitting. Typical form factor: integrated small cell unit 30×30×10 cm or less. Beamwidth: Horizontal: 90-120° (sector) or 360° (omnidirectional) for street-level coverage. Vertical: 20-40° (downtilted to cover street level, not upper floors of adjacent buildings). For sub-6 GHz (3.5 GHz n78): 4T4R or 8T8R MIMO, 8-16 dBi gain, 10-20 W per port. Beamwidth provides natural coverage limitation to ~200-300 m. For mmWave (28 GHz n257): 64-256 element phased array, 25-30 dBi gain with beam steering ±60°. EIRP up to 55 dBm. Coverage limited to 100-200 m LOS due to high path loss and building blockage. Array scanning range must cover ±60° horizontal to serve users across the street and around corners. Backhaul: fiber preferred (1-10 Gbps), wireless backhaul via mmWave point-to-point (E-band 71-86 GHz) as alternative. Power: 200-500 W total (including baseband processing), requiring -48 VDC or PoE++ (90 W max for low-power nodes).

RF Propagation

Below Rooftop

Q: What are the antenna requirements for below-rooftop small cells?

Below-rooftop antenna design differs significantly from macro-cell: Height: 3-10 m (lamppost, utility pole, building facade). Must be inconspicuous for municipal permitting. Typical form factor: integrated small cell unit 30×30×10 cm or less. Beamwidth: Horizontal: 90-120° (sector) or 360° (omnidirectional) for street-level coverage. Vertical: 20-40° (downtilted to cover street level, not upper floors of adjacent buildings). For sub-6 GHz (3.5 GHz n78): 4T4R or 8T8R MIMO, 8-16 dBi gain, 10-20 W per port. Beamwidth provides natural coverage limitation to ~200-300 m. For mmWave (28 GHz n257): 64-256 element phased array, 25-30 dBi gain with beam steering ±60°. EIRP up to 55 dBm. Coverage limited to 100-200 m LOS due to high path loss and building blockage. Array scanning range must cover ±60° horizontal to serve users across the street and around corners. Backhaul: fiber preferred (1-10 Gbps), wireless backhaul via mmWave point-to-point (E-band 71-86 GHz) as alternative. Power: 200-500 W total (including baseband processing), requiring -48 VDC or PoE++ (90 W max for low-power nodes).

/bih-LOH ROOF-top/

Propagation regime where the TX antenna is below the surrounding building roofline (3 to 10 m height vs. 15 to 25 m rooftops). Dominant mechanisms: street canyon waveguiding (n ≈ 2.0 to 2.5), corner diffraction (15 to 25 dB/corner), facade reflection, vehicle scattering. Delay spread: 30 to 300 ns. Models: 3GPP TR 38.901 UMi, WINNER II B1, ITU-R P.1411. Standard for 5G NR small cells: 100 to 500 m ISD at 3.5/28 GHz.

Height: 3–10 m

ISD: 100–500 m

PL exp: 2.0–4.5

Understanding Below-Rooftop Propagation

The shift from macro-cell (above-rooftop) to small-cell (below-rooftop) deployment fundamentally changes the RF planning problem. Above-rooftop antennas see relatively predictable over-building diffraction paths described by well-established models (Walfisch-Ikegami, COST 231). Below-rooftop antennas live in the street-level environment where every building corner, vehicle, and tree creates significant propagation effects.

This complexity is both the challenge and the opportunity of 5G densification. Below-rooftop cells provide short-range, high-capacity coverage precisely where users are concentrated (street level, ground-floor retail, transit corridors). The limited range (100 to 300 m) means aggressive frequency reuse, and the rich multipath environment enables spatial multiplexing (MIMO) that further increases capacity.

3GPP UMi Path Loss Model

LOS (Street Canyon):
PL = 32.4 + 21·log₁₀(d_3D) + 20·log₁₀(f_c)
σ_SF = 4.0 dB
Valid: 10 m < d < 5 km, 0.5–100 GHz

NLOS:
PL = 35.3·log₁₀(d_3D) + 22.4
+ 21.3·log₁₀(f_c) − 0.3·(h_UT − 1.5)
σ_SF = 7.82 dB

LOS Probability:
P_LOS = min(18/d, 1)·(1 − e^−d/36) + e^−d/36

Building Entry Loss:
Glass: 3–8 dB | Concrete: 15–25 dB
Low-E glass: 25–40 dB

Below-Rooftop vs. Above-Rooftop

Parameter	Below Rooftop (UMi)	Above Rooftop (UMa)
Antenna height	3–10 m	25–50 m
Cell radius	100–300 m	500–2,000 m
PL exponent (LOS)	2.0–2.5	2.0–2.2
PL exponent (NLOS)	3.0–4.5	2.8–3.5
Delay spread	30–300 ns	500–3,000 ns
Shadow fading (NLOS)	7.8 dB	6.0 dB

Common Questions

Frequently Asked Questions

BRT vs. macro-cell?

BRT: street-level waveguiding, corner diffraction (15 to 25 dB/corner), facade reflection. Macro: over-rooftop diffraction. BRT delay spread 30 to 300 ns vs. 500 to 3,000 ns macro. BRT range 100 to 300 m vs. 500 to 2,000 m.

Propagation models?

3GPP UMi: LOS PL = 32.4 + 21·log(d) + 20·log(f). NLOS: 35.3·log(d) + 22.4 + 21.3·log(f). WINNER II B1, ITU-R P.1411. Valid 0.5 to 100 GHz.

Antenna design?

Sub-6 GHz: 4T4R/8T8R, 8 to 16 dBi, 90 to 120° sector. mmWave: 64 to 256 element phased array, 25 to 30 dBi, ±60° steering. EIRP to 55 dBm. Form factor <30×30 cm.

Related Terms

← Bellcore TR-332 Bench Cable →

Small Cell RF

Precision RF Components

RF Essentials provides precision terminations and custom waveguide assemblies for small cell antenna testing, urban propagation measurement campaigns, and 5G NR channel sounding equipment.

Request a Quote