Path Loss Prediction Model for Mobile Radio Wave Propagation into a Multi-Floored Building

This thesis presents the development of a new path loss prediction model for mobile communication field due to wave propagation into multifloored building. Field strength measurements from four different base transceivers (BTS) located at Universiti Putra Malaysia (UPM) campus and Taman Desa Serdang were conducted at two buildings using an Advantest U3641 spectrum analyzer and AHS519-4 log-periodic antenna. A computer program has been developed to retrieve the measured field strength data from the spectrum analyzer and convert the values gained to path loss using Agilent VEE software. Line-of-sight propagation (for open area) and non-line-of-sight propagation (for building wall obstruction) have been investigated. The measured path loss data have been compared with the results obtained using various path loss prediction models such as COST231 line-of-sight (CLOS), COST231 non-line-of-sight (CNLOS), Gahleitner-Stochastic (GS), Paulsen-Microcell (PMI) and Paulsen-Macrocell (PMA). The results demonstrate poor agreement between the predicted and the true measured path loss. For line-of-sight case; CLOS, GS, PMI and PMA models have overestimated the path loss as high as 16.1%, 78%, 8.61% and 35.8% respectively. For non-line-of-sight case; CNLOS, GS, PMI and PMA models have overestimated the path loss as high as 14%, 91%, 5.56% and 56.17% respectively in all measurement frequencies. An improved version of the PMI model has been developed and tested where the mean error values are found to be approximately 2.5% for all the measurement frequencies. In addition, integrated software UPMIPL for path loss prediction of wave propagation in both line-of-sight and non-line-of-sight cases has been developed and implemented using Agilent VEE. The UPMIPL program provides the utility for calculating the signal characteristics of radio propagation paths and is realized in the run time version.

Preview PDF FS_2007_59.pdf Download (140kB)

Actions (login required)

View Item