Empirical approach for 5G NR 3.5GHZ dense urban propagation model tuning

Abstract

With the increasing demand for wireless networks and rapid growth in wireless access technologies, there are numerous technical and economic problems during the planning, deployment and optimization phases of radio networks from the operator’s point of view. Wireless signal prediction is one of the major approaches to minimize those problems by pre-analyzing coverage, capacity and interference through specific propagation modelling. With the introduction of fifth generation (5G) wireless systems in the last decade, several new technical aspects such as higher operating frequency, larger bandwidth, massive multiple-input and multiple-output (MIMO) and beamforming features have been incorporated into 5G wireless networks. However, these new technical aspects are not properly addressed in legacy propagation modelling. Hence, telecommunication operators find it difficult to have accurate propagation models which minimize the gap between simulation results and actual field measurements. This thesis presents a novel approach of propagation modelling for 3.5 GHz 5G network in urban environment. The proposed methodology first provides a method for purifying field-collected signal strength samples by new filtering and averaging approaches. The purified samples are then used to accommodate 5G features in different scenarios in the Planet tool, which is the most popular engineering tool among wireless network operators for network coverage simulation. Furthermore, this thesis compares the accuracy of the proposed universal model with the basic universal model and other legacy models, which do not capture the main characteristics of 5G wireless access technology..