Correlation properties of doppler fading in cellular networks in Ghana

Thumbnail Image
Journal Title
Journal ISSN
Volume Title
Fading is the most observed and frustrating problem in transmitting and receiving radio signals. Studying the correlation properties of fading helps us to design proper communication systems that are robust to it. In this project, received power signal is obtained by drive-test technique. The data collection is done in an operating frequency of 900MHz and 1800MHz and the environments used for the data collection are urban and suburban. The analysis is done using curve fitting tool and distribution fitting tool in Matlab software 7.5.0 (R2007b).The following fading distributions are used, Rayleigh, Rician, Nakagami-m, Lognormal and Weibull. Weibull fading distribution is found to give the best description of fading experienced in Ghana. On the other hand, Rician distribution, Nakagami-m distribution and lognormal distribution have less agreement with the measured data while Rayleigh fading distribution has much less agreement with measured data. The obtained mean scale parameter and the mean shape parameter in the Weibull distribution for the environment considered are 133.76 and 15.93, respectively. Furthermore, the path loss exponents in the near and far fields are also calculated; and the standard deviation of shadowing falls in the range of 4.368dB to 6.180dB. Last but not least, a clear dependence between the root-mean square (rms) delay spread values of the received power signals and the transmitter-receiver separation distance is observed. Ultimately, the analysis on the large-scale fading tends to describe the terrain profile of the environment and it is dependent on the distance.
A Thesis submitted to the Department of Telecommunication Engineering, Kwame Nkrumah University of Science and Technology in partial fulfillment of the requirements for the degree of Master of Science