Embedding antennas with tilted beam patterns into parabolic wave equation‐based models for tunnel propagation

Abstract

Accurate radio wave prediction models are necessary for the innovation of mobile communication technology and the deployment of wireless sensor networks in railways. The parabolic wave equation (PWE) method is one commonly used method to predict the electromagnetic characteristic of long railway tunnels. However, the advent of novel antennas and technologies (e.g., beam steering, beamforming) has posed new challenges to PWE-based models such as effective means of taking tilted beam patterns into account. This study addresses such a challenge by using tilted Gaussian beams to embed antennas with oriented patterns or tilted beam directions into PWE-based models. The proposed approach does not reduce the simulation efficiency of PWE itself. The validity and practicability of the technique are demonstrated by comparing against ray-tracing as well as experimental measurements in various tunnel environments.

Efficient high‐fidelity deep convolutional generative adversarial network model for received signal strength reconstruction in indoor environments

“…These models necessitate detailed environmental data–like the geometry and material properties of specific sites—to accurately represent each unique setting, rendering it site‐specific and yielding more precise outcomes. Nevertheless, the computational efficiency of ray tracing (RT), a deterministic model frequently employed for indoor situations, is frequently impeded by the prolonged simulation time and memory demands [8], thereby creating a bottleneck for its implementation. Traditional methods have used spatial interpolation techniques like Kriging [9] to reconstruct complete RSS maps from a subset of measured RSS values, offering a time‐saving alternative.…”

Comparative analysis of finite‐difference and split‐step based parabolic equation methods for tunnel propagation modelling

“…Since the introduction of the PE method, a range of approaches, including the narrow-angle and wideangle PE methods, as well as the one-way and two-way PE (2W-PE) modelling, have been introduced to enhance the applicability of the PE technique in different scenarios [6,20]. More recently, various techniques have been introduced to PE to enable propagation modelling in tunnels with complex geometries/sources and rough surfaces [21][22][23][24][25]. PE methods are also combined with machine learning techniques, providing an alternative approach to accelerate wave propagation prediction [26][27][28][29][30].…”