1. The Physics of Sub-Terahertz (100 GHz – 3 THz) Wave Propagation
1.1 Transitioning Beyond 5G mmWave to 6G Sub-THz
While 5G networks exploit frequency ranges up to 52.6 GHz within Frequency Range 2 (FR2-1) and extended FR2-2 (52.6 GHz to 71 GHz), next-generation 6G mobile architectures are targeting sub-terahertz frequency bands ranging from 100 GHz to 1 THz (specifically D-band: 110–170 GHz, G/H-band: 220–325 GHz, and sub-millimeter bands from 300 GHz to 3 THz). These super-high-frequency bands offer tens of gigahertz of contiguous bandwidth, enabling terabit-per-second (Tbps) data rates, sub-millisecond latencies, and ultra-dense wireless fronthaul/backhaul fabrics defined under the ITU-R IMT-2030 vision.
1.2 The Free-Space Friis Attenuation Challenge
Under free-space conditions, electromagnetic power radiated by an isotropic antenna spreads spherically over distance d. The isotropic Free-Space Path Loss (FSPL) is derived directly from Friis' transmission equation:
In logarithmic decibel units:
FSPLdB = 20 × log10(dkm) + 20 × log10(fGHz) + 92.45 dB
Because free-space path loss scales quadratically with frequency (proportional to f²), moving from 28 GHz in 5G mmWave up to 140 GHz in 6G D-band incurs an automatic isotropic attenuation penalty of 20 × log10(140 / 28) = 13.98 dB over identical transmission distances. At 300 GHz, this isotropic loss reaches 20.60 dB over 28 GHz.
2. ITU-R P.676 Molecular and Atmospheric Gas Absorption
Unlike microwave frequencies where dry air and water vapor are almost completely transparent, sub-THz electromagnetic radiation interacts quantum-mechanically with the rotational and vibrational absorption resonances of ambient atmospheric molecules. Recommendation ITU-R P.676 ("Attenuation by atmospheric gases and related effects") establishes the definitive global standards for quantifying these losses.
2.1 Oxygen (O2) and Water Vapor (H2O) Absorption Lines
Total specific atmospheric gaseous attenuation γgas (dB/km) is the sum of dry air attenuation (γo, dominated by molecular oxygen) and water vapor attenuation (γw):
Total Atmospheric Gas Loss (Agas) = γgas × dkm [dB]
Key resonance peaks across the 100 GHz to 1000 GHz spectrum include:
- 118.75 GHz (O2 Peak): A strong isolated oxygen magnetic dipole resonance that generates approximately 1.5 to 2.0 dB/km of attenuation at sea level.
- 183.31 GHz (H2O Peak): A severe electric dipole resonance of water vapor. Under moderate sea-level humidity (ρ = 7.5 g/m³), attenuation exceeds 30 dB/km, creating an impenetrable transmission barrier for terrestrial backhaul.
- 325.15 GHz & 380.20 GHz (H2O Peaks): Higher-order water vapor resonance lines producing attenuation rates of 25 dB/km and 45 dB/km respectively.
2.2 Strategic Low-Attenuation Transmission Windows
To maintain multi-gigabit carrier viability, 6G wireless network planners position transmission channels within distinct "atmospheric transmission windows" where molecular absorption hits local minima:
- D-Band Transmission Window (130–150 GHz): Center frequency 140 GHz. Total gaseous absorption remains near 1.0 dB/km under standard conditions, offering up to 20 GHz of contiguous bandwidth for urban microcell backhaul links up to 1 km.
- H-Band Transmission Window (210–240 GHz): Center frequency 220 GHz. Total attenuation averages 2.0 dB/km, providing massive capacity for high-density small-cell mesh clusters up to 500 meters.
- 300 GHz Sub-THz Window (280–310 GHz): Gaseous attenuation remains between 3.2 and 4.5 dB/km, making it ideal for campus-scale kiosk downloads, intra-data-center wireless fabrics, and chip-to-chip interconnects up to 200 meters.
3. Antenna Effective Aperture & Ultra-Massive MIMO Scaling
While isotropic path loss increases with f², antenna effective aperture area is given by Ae = (λ² / 4π) × G. Because operating wavelength λ contracts from 21.4 mm at 14 GHz down to 2.14 mm at 140 GHz (and 1.0 mm at 300 GHz), hundreds to thousands of sub-wavelength antenna elements can be integrated into tiny physical footprints.
By maintaining a constant physical aperture size (e.g., a compact 10 cm × 10 cm planar array), antenna directivity gain G scales quadratically with frequency (G ∝ 4π Aphys / λ² ∝ f²). When directional beamforming arrays are implemented at both the transmitter and receiver, the combined gain scales as Gtx × Grx ∝ f4, which completely overwhelms the f² free-space spreading loss, delivering a net received power advantage over lower-frequency systems with identical physical antenna dimensions.
4. Extended Close-In (CI) and Alpha-Beta-Gamma (ABG) Path Loss Formulations
In complex real-world microcell environments, scattering and reflection must be accounted for beyond standard free space. 3GPP and academic measurement campaigns extend standard propagation using two primary multi-frequency models:
PLCI(f, d) = FSPL(f, 1m) + 10 × n × log10(d) + Xσ + γgas(f) × d
where n is the path loss exponent (n ≈ 2.0 for Line-of-Sight, n ≈ 3.2 to 3.8 for Non-Line-of-Sight), and Xσ represents log-normal shadow fading.
PLABG(f, d) = 10 × α × log10(d) + β + 10 × γ × log10(f / 1 GHz) + XσABG
5. Techno-Commercial Impact on 6G Backhaul & Fronthaul Networks
Trenching dark fiber optics in dense metropolitan downtown areas routinely costs between $100,000 and $300,000 per kilometer, with civil permitting delays extending past 12 to 18 months. Sub-THz wireless bridges operating at 140 GHz or 220 GHz offer rapid, same-day deployment with full 100 Gbps to 1 Tbps throughput capabilities, bypassing right-of-way bottlenecks. Furthermore, because sub-THz signals propagate through air at the speed of light (c ≈ 300,000 km/s) compared to single-mode silica fiber (v ≈ 200,000 km/s), sub-THz wireless saves approximately 1.67 microseconds of transport latency per kilometer, directly easing the strict latency and jitter budgets of Open-RAN eCPRI fronthaul interfaces.