Semi Deterministic Hybrid Model for Path Loss Prediction Improvement

Mobile radio path loss modeling is an important aspect of network planning which ensures an improved quality of service to the subscribers. In this paper, the predictions made by COST 231 Walfisch-Ikegami propagation model to estimate the path loss of the mobile signals are improved by proposing a hybrid model. The modified model combines the deterministic aspects of ray tracing along with statistical processing of the empirical measurements. It includes a loss term computed for multiple reflections by using the method of images. The ten ray urban street canyon model is extended to compute reflections from multiple paths. The deterministic approach is merged with the statistical variations in the original Walfisch-Ikegami model. It consists of modeling the multi screen diffraction loss, roof to street loss and street orientation function by considering the height of the buildings, the separation and the road orientation angle as random variables with Gaussian distribution. The proposed hybrid model is validated by measurements of GSM 900MHz signals collected in Hyderabad city of Southern India. The performance of the proposed model is evaluated in terms of Prediction Error, Root Mean Square Error and other metrics. The results justify the improvement in the path loss prediction of the proposed semi deterministic hybrid model. This could be useful in link budget analysis and deployment of future cellular networks for the specified urban region and similar scenarios.