COST 231 Hata
Extending Okumura-Hata to the PCS Band
When second-generation cellular networks moved from the 900 MHz GSM band up to the 1800 MHz (GSM 1800/DCS) and 1900 MHz (PCS) bands in the early 1990s, the existing Okumura-Hata model fell outside its validated frequency range. The original Hata equations were curve-fitted to Okumura's Tokyo measurements for 150 to 1500 MHz, so applying them at 1800 MHz introduced systematic error. The European cooperative research program COST 231 (Cooperation in Science and Technology, Action 231) addressed this by collecting additional propagation data and refitting the model constants, producing what is now universally called the COST 231 Hata or Extended Hata model.
The result preserves the convenient closed-form structure that planners liked: path loss is a simple sum of a frequency term, a base-height term, a distance-versus-height slope, and a mobile-antenna correction. Two changes distinguish it from its predecessor. First, the leading constant and the frequency coefficient are refit for the higher band, raising the baseline loss to match measured 1800 MHz behavior. Second, a discrete city-size correction Cm is appended, adding 3 dB in dense metropolitan clutter while leaving suburban and medium-city predictions unchanged. Both versions share the identical mobile-antenna correction function a(hm).
Because it is a macrocell model, COST 231 Hata assumes the base antenna clears the local rooftop line and that propagation is dominated by diffraction over building edges rather than street-level guiding. Engineers building a coverage plan typically pair the model with a clutter database, a terrain-elevation profile, and a link budget so the predicted median loss can be combined with fade and interference margins before cell sites are committed.
Governing Equations
L50 = 46.3 + 33.9·log10(f) − 13.82·log10(hb) − a(hm) + (44.9 − 6.55·log10(hb))·log10(d) + Cm dB
Mobile-antenna correction (small/medium city):
a(hm) = (1.1·log10(f) − 0.7)·hm − (1.56·log10(f) − 0.8) dB
City-size correction:
Cm = 0 dB (suburban / medium city), Cm = 3 dB (dense metropolitan)
Where f = frequency in MHz (1500 to 2000), hb = base-station effective height in m (30 to 200), hm = mobile antenna height in m (1 to 10), d = link distance in km (1 to 20). Example: f = 1800 MHz, hb = 50 m, hm = 1.5 m, d = 5 km, dense urban (Cm = 3) → a(hm) ≈ 0.04 dB and L50 ≈ 160 dB; the same link in suburban clutter (Cm = 0) gives ≈ 157 dB.
Model Comparison
| Model | Valid Frequency | Distance | Base Height | City Term | Best Use |
|---|---|---|---|---|---|
| COST 231 Hata | 1500 to 2000 MHz | 1 to 20 km | 30 to 200 m | Cm = 0 or 3 dB | GSM 1800 / PCS macrocells |
| Okumura-Hata | 150 to 1500 MHz | 1 to 20 km | 30 to 200 m | None | GSM 900 / VHF-UHF macrocells |
| COST 231 Walfisch-Ikegami | 800 to 2000 MHz | 0.02 to 5 km | 4 to 50 m | Street geometry | Urban micro and macrocells |
| Free-space (FSPL) | Any | Line of sight | N/A | None | Clear LOS reference |
Frequently Asked Questions
What frequency and distance range is the COST 231 Hata model valid for?
It is validated for 1500 to 2000 MHz, base-station effective heights of 30 to 200 m, mobile heights of 1 to 10 m, and link distances of 1 to 20 km. Below 1500 MHz use the original Okumura-Hata model; above 2000 MHz the fit degrades. As a macrocell model it assumes the base antenna sits above local rooftops, so it is not intended for microcell or indoor cases.
How does COST 231 Hata differ from the original Okumura-Hata model?
It keeps the Okumura-Hata functional form but refits the band-dependent constants: the leading term changes from 69.55 to 46.3 and the frequency coefficient from 26.16 to 33.9. It also adds a city-size term Cm (0 dB suburban/medium, 3 dB dense urban) that the original lacked. The mobile-antenna correction a(hm) is carried over unchanged.
What is the city-size correction factor Cm in COST 231 Hata?
Cm is an additive term for building density: 0 dB for suburban and medium-sized cities, 3 dB for dense metropolitan centers with continuous tall-building cover. The 3 dB increment reflects extra diffraction and shadowing from high-rise clutter. Planning tools usually map each clutter class (open, suburban, urban, dense urban) to a Cm value before the link budget is computed.