Update on MW Radio Rain Fading Estimation George Kizer

Major Topics MW Path Design Point Rain Attenuation Point to Path Conversion Factor Rain Fading Variability

Rain Fading

Microwave Path Design Parameters North America Europe (ITU-R) Quality (one way) Not Calculated Multipath (worst month) (tested) Short Term Interference Availability (two way) Multipath Rain (annual) Rain Long Term Interference Interference Hardware Calculations are typical (median) estimates, not worst case.

Microwave Path Rain Design North America Europe (ITU-R) Rice and Holmberg (ITS, 1973)* Lin (Bell Labs, 1977)** Crane (1980)* ITU-R P.530-1 (1978)* Crane (1996)* Crane (2003)** ITU-R P.530-10 (2001)** * Rain Zones ** Specific Locations Crane is de facto standard in North America ITU-R is standard in Europe

How Are Crane 2003 and ITU-R P.530 Different Crane 2003 ITU-R P.530 Based upon 5 minute integration (commonly available) Based upon 1 minute integration (less available) Works for all probabilities Only works for 0.001% to 1% (two 9 s to five 9 s, not possible to extend range) Curves based upon actual rain data (based upon 5 parameters derived from actual site data) Rain curves are predetermined (one for > ±30 0 latitude, one for < ±30 0 latitude, uses 0.01% unavail. value based upon 3 interpolated parameters) Point to Path conversion factors are quite different

Crane Model Based Upon Two Components San Francisco Miami Low Density Rain ( debris ) Contribution = exponential distribution (two parameters) Intense Rain Cell ( cell ) Contribution = lognormal distribution (three parameters)

ITU-R P.530 Based Upon Shapes and Offsets Zone D3 1 2 Rain Distribution Shape 1 for latitude ±30 0 Rain Distribution Shape 2 for latitude < ±30 0 Shape shifted up or down from 0.01% rain rate Rain rate for 0.01% determined by three geographically based parameters Only applicable for probability between 0.001% and 1%

Major Topics MW Path Design Point Rain Attenuation Point to Path Conversion Factor Rain Fading Variability

Microwave Path Rain Design M = γ R d eff M = path flat fade margin (db) γ R = single location ( point ) rain attenuation (db/km) d eff = effective path length (km) = d F pp d = actual path distance F pp = point to path conversion factor 10

Microwave Path Rain Design γ R = point rain attenuation = KR α (db/km) K and α frequency dependant R = site or zone rain rate (mm/hr) for specific outage probability (e.g., 0.01%) 11

Point Rain Attenuation

Need for Current Point Rain Attenuation More than 100 years of NOAA rain data

Typical Rain City Data (99.9999%) (99.999%) (99.99%) typical measurement limit

Why create curves for parameter model (rather than actual measurements)? North American 11 GHz calculations for high availability require rain availabilities as high as 99.99995% (0.00005% unavailability) For 5 minute integrated data, that is one measurement in 20 years Actual data does not have reliable values for high availabilities (those rain events are too rare) Only the model can give reliable high availability estimates (it is based upon the underlying physical mechanisms)

Rain City Data Crane 2003 data has been augmented with Segal (CRC) and Kizer (NOAA) data Data represents 282 cities in Canada and the United States Curves derived from curve fits of actual data to Crane 5 parameter rain model

Why Migrate to Site Specific Rain Data? Crane 1996 All zones are averages of a wide range of specific site data Some zones misrepresent some locations

Crane 1996 Rain Regions

Site Data Different than Zone Average Zone C Zone F

Zone Average Misrepresents Site Rain Rates Zone D2 Zone C

Differences in Crane and ITU-R Site Data Crane Zone B1 Crane 2003 ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone C ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone D1 ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone D2 ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone D3 ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone E ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Differences in Crane and ITU-R Site Point Data Crane 2003 Crane Zone F ITU-R P.530 ITU rain curves are simplified to eliminate frequency dependency This introduces about 10% error for frequencies between 11 and 23 GHz

Major Topics MW Path Design Point Rain Attenuation Point to Path Conversion Factor Rain Fading Variability

Differences in Point to Path Factor

Differences Due to Point to Path Factor Crane Zone C Average Rain Rates (low rain rate region) ITU-R methods not valid for 99.9999% path availabilities

Differences Due to Point to Path Factor Crane Zone D2 Average Rain Rates (typical rain rate region) ITU-R methods not valid for 99.9999% path availabilities

Differences Due to Point to Path Factor Crane Zone E Average Rain Rates (high rain rate region) ITU-R methods not valid for 99.9999% path availabilities

Major Topics MW Path Design Point Rain Attenuation Point to Path Conversion Factor Rain Fading Variability

Rain Fading Variability

Philidelphia Example (2008-2009)

Rain Fading Variability

Major Topics MW Path Design Point Rain Attenuation Point to Path Conversion Factor Rain Fading Variability

george.kizer@alcatel-lucent.com