Abstract-Substantial modifications have been made to the expressions for calculating distance factor and extrapolation techniques in the latest ITU-R P.530-14. However, its performance has not been rigorously evaluated in the tropical and equatorial climates. In this article, the new ITU-R method and three prediction models are validated using measurement data from tropical Malaysian climate. The data were collected on six geographically spread terrestrial microwave DIGI MINI-LINKs operating at 15 GHz. When tested against measurements, the Da Silva Mello model yields a significant improvement for the prediction of rain attenuation distributions. The prediction errors observed in the ITU-R model suggest the need for more data campaign in the afore-mentioned climates.
I. INTRODUCTIONTraditionally frequencies above 10 GHz are commonly employed in the current microwave point-topoint networks, due to large bandwidth and freedom from traffic congestion. However, in these frequency bands, the performance of the fixed service (FS) is predominantly controlled by rain in 2012; however, its performance has not been largely evaluated using data collected from tropical climates. Therefore, the main aim of this article is to evaluate the performance of the newly released ITU-R model and the three afore-mentioned models, using measurement data from six locations in a tropical climate. The data used are the rain rate and rain attenuation data banks available from six geographically spread terrestrial microwave DIGI MINI-LINKs operating at 15 GHz as described in[13]. It should be mentioned here that the ITU-R P.530-14 was not yet in force when our previous works were reported.
II. THEORETICAL BACKGROUND OF RAIN ATTENUATION PREDICTION MODELSRain attenuation can be conveniently defined as the product of specific attenuation (dB/km) corresponding to the point rain rate (typically measured at one end of the link) and the effective propagation path length (km) [1]. The path correction factor is defined as the ratio of effective path length to the physical path length of a microwave link, and its value is usually less than unity, except in rear cases [14].The concept of effective path length is thus a way of 'averaging out' the spatial inhomegeneity of rain rate and thus specific attenuation [15]. Note that the degree of spatial inhomogeneity in rain rate generally varies with rainfall intensity. Therefore the variation of path length reduction factor can be expressed as a function of rain rate or the corresponding time exceedance [16].Attenuation p A exceeded at p % of time can either be obtained from direct measurements, or predicted from the knowledge of long-term rainfall rate. Generally, the required inputs in most attenuation prediction models for terrestrial point-to-point microwave links are the rainfall rate exceeded at p % of time, the effective propagation path length and the link's operating frequency [7], [16]. Thus p