Two-Dimensional Deterministic Propagation Models Approach and Comparison With Calibrated Empirical Models

Since the development of wireless cellular networks in the early 1970s, their unpredictable success has led to particular interest in developing accurate propagation models, particularly for indoor environments where small cells are deployed. Many propagation models have been proposed in literature, and they are classified into three major categories, i.e., empirical, semiempirical, and deterministic models. Deterministic models such as ray tracing and radiosity are very attractive due to their accuracy, but their 3-D formulation makes them tedious to use, and their migration to a 2-D representation becomes a necessity. In this paper, we propose to develop and implement a 2-D approach for ray tracing and radiosity that allows saving memory and processing time while maintaining prediction accuracy. The performance of the two methods is evaluated using measurements taken in the CITI laboratory building. It is shown that 2-D radiosity meets the ITU accuracy statistics recommendations for a 2.4-GHz indoor test environment with a mean error near to zero and a standard deviation of 5.98 dB. The comparison to 2-D ray tracing and an empirical calibrated model such as log-normal shadowing and the Cheung model shows that 2-D radiosity is the most accurate and saves more than 70% of the processing time compared with 2-D ray tracing while using almost the same memory load.