Analysis of using TV whitespaces for broadband connectivity in selected areas of Uganda.
Abstract
The project is about the analysis of the feasibility of using TV Whitespace for broadband connectivity in rural areas of Uganda whereby the case study is Luwero district.
We first determined the TV Whitespace spectrum available in Luwero. This was done by using the energy detection method which compares the received signal energy to a certain predefined threshold, if the signal is above the threshold, the band is considered to be occupied, otherwise, the band is free and is considered a TV Whitespace. We noted that the number of whitespaces varied from point to point due to the fact that the TV transmitters were transmitting to different distances away from the transmitters.
We also determined the most suitable propagation model for path loss prediction. We selected three propagation models that is Hata, Cost231, and extended Hata models basing on their range of frequency, area of application that is rural area, and the complexity of the data needed to provide accurate results. In this case, we first did a prediction using the propagation model equations in MATLAB using the parameters such as the antenna heights, gains and others which we obtained from SIGNET Uganda whereby we obtained the prediction values for Path loss.
We then carried out field measurements where we obtained the received power, and given that we already know the transmitter power at Kololo, we calculated the path loss for the measured values at different distances from the transmitters. To note, these measurements were done at six different points in Luwero.
We then compared the measured values of path loss with the prediction values. We obtained the difference between the measured and prediction values whereby Extended Hata model had a small deviation between these values, so we decided to modify the extended Hata model. In this modification, we obtained the Root Mean Square Error value which we added to the extended Hata equation. This RMSE is to cater for the environment of Luwero.
We also looked at the derivation of the key parameters for TV Whitespaces.
The key results include; we identified a significant amount of spectrum that can be used to provide extra mobile broadband connectivity in rural areas like Luwero, we assessed the accuracy of the three selected propagation models in predicting the TV signal, we modified the extended Hata model to suite the environment and which gives the best prediction of the TV signal in Luwero.
Conclusion is that the project has achieved the propagation model for accurate prediction of DVB-T spectrum occupancy in Luwero.
Recommendations are; the results from this study could serve as input for future regulatory decisions on the UHF band and cost-benefit analysis for deploying new systems to access this spectrum band. We also recommend the project to be taken up by government for database implementation.