Comparison between path-loss prediction models for wireless telecommunication system design

For the cell design of mobile communication systems, the path loss prediction models are very important. These models can be classified into theoretical and experimental models. The main experimental models are the Okumura-Hata, Cost231-Hata, and ITU-R model. The problem of these models is that these prediction expressions are based on the qualitative propagation environments such as urban, suburban, and open areas. The Cost231-Walfisch-Ikegami model (Cost231-WI) is a result of the effort to use a quantitative description of the propagation environments. In addition to the height of T/sub x/ and R/sub x/ antennas, the quasi-uniform building height and width of street are considered in this model. A mistake in formulating the Cost23 I-WI from the original Ikegami (see IEEE Trans Antennas Propagat., vol.32, p.822-29) as found by Har, Watson and Chadney (see IEEE Trans. Veh. Technol., p.1451-52, Sep. 1999). The mistake happened due to the misunderstanding between reflection coefficient and reflection loss. We comment on this. For a more quantitative understanding of the propagation environment when Oukumura measured path losses near Tokyo, we compared the path losses of Cost231-Hata with those of Cost231-WI based on the same frequency (f=2000 MHz, and 1500 MHz), T/sub x/ and R/sub x/ antenna heights (h/sub b/=40 m and h/sub m/=1.5m), but changing building height (10-25 m) and street widths (20-25 m). We also compared Okumura-Hata model with the ITU-R model using a similar approach. Through such a parametric study, a quantitative propagation environment based on which Okumura curves were extracted is estimated.