Search

The objective of this Special Issue is to showcase a unified vision for the applications of AI in radio propagation for communications and other relevant aspects. More specifically, the initial announcement encouraged emphasis in, but not limited to, the following areas: novel AI or AI-enabled techniques for radio propagation characterization and analysis; AI-enabled data analysis and parameter extractions of wireless channels; clustering analysis for radio channel characterization and modeling; AI-enabled channel modeling and communication system simulation; AI algorithm design in radio propagation for the applications in communications.

In order to provide higher data rates, as well as better coverage, cost efficiency, security, adaptability, and scalability, the 5G and beyond 5G network are development with various artificial intelligence techniques. In this twopart paper, we investigates the application of artificial intelligence (AI) and in particular machine learning (ML) to the study of wireless propagation channels. It firstly provides a comprehensive overview of ML for channel characterization and ML-based antenna-channel optimization, and then gives state-of-the-art literature review of channel scenario identification and channel modeling in Part II. Fundamental results and key concepts of ML for communication networks are presented, and widely used ML methods for channel data processing, propagation channel estimation, and characterization are analyzed and compared. A discussion of challenges and future research directions for ML enabled next generation networks rounds off the paper.

This two-part paper investigates the application of artificial intelligence (AI) and, in particular, machine learning (ML) to the study of wireless propagation channels. In Part I of this article, we introduced AI and ML and provided a comprehensive survey on ML-enabled channel characterization and antenna-channel optimization, and in this part (Part II), we review the state-of-the-art literature on scenario identification and channel modeling here. In particular, the key ideas of ML for scenario identification and channel modeling/prediction are presented, and the widely used ML methods for propagation scenario identification and channel modeling and prediction are analyzed and compared. Based on the state of the art, the future challenges of AI-/ML-based channel data processing techniques are given as well.

Ray tracing is a powerful tool to obtain deterministic and dynamic descriptions of communication channels. However, performing ray tracing simulations at each discrete time instant is computationally expensive. Instead, a new approach is proposed to extrapolate results obtained from a single ray tracing simulation. It relies on the geometric tracking of interaction points (i.e. reflection or diffraction points), enabling analytical or numerical predictions of the evolution of any ray identified during an initial ray tracing simulation. The performance of this new approach is studied on several canonical vehicle-to-vehicle configurations, as part of a statistical study. Focus is also given to the time horizon during which dynamic ray tracing is possible, related to the lifetime of the main rays. This time horizon can be directly estimated based on the knowledge of the geometry and its evolution. It is found to be the main parameter influencing the accuracy and the computational gain of the presented approach.

Vehicle-to-vehicle (V2V) wireless communications have many envisioned applications for ensuring traffic safety and for addressing traffic congestion. However, developing suitable communication systems and standards for this purpose requires developers to have accurate models for the V2V propagation channel. Likewise, the dynamic evolution of multipath components (MPCs) in V2V channels has not been well modeled in existing models. In this paper, we propose a geometry-based stochastic channel model for a lightly built-up urban environment and then parametrize the model from measurements. The MPCs are extracted based on a high-resolution parameter estimation; they are tracked and clustered through a joint algorithm. The identified clusters are classified as line-of-sight, reflections from static scatterers, reflections from mobile scatterers, multiple-bounce reflections, and diffuse scattering. Specifically, the multiple-bounce reflections are modeled as twin clusters that follow the COST 273/COST2100 approach. The paper gives a full parameterization of the channel model and supplies a step-by-step implementation recipe. We verify the model by comparing two second-order statistics, i.e., the root-mean-square (RMS) delay spread and the angular spreads of arrival/departure derived from the channel model, to the results obtained directly from the measurements. Furthermore, we also identify several key factors that strongly impact the synthetic channel performance

Downlink exposure to electromagnetic fields due to cellular base stations in urban environments is studied using the stochastic geometry framework. A two-dimensional Poisson Point Process is assumed for the base station distribution. Mathematical expressions of statistics of exposure are derived from a simple propagation model taking into account the height of the base stations. The error on exposure made by taking a limited number of base stations, instead of the whole set, is quantified. The relative impact of the model parameters on the statistics of exposure is highlighted. The method is then applied and the model parameters are calibrated using experimental data obtained by drive-tests in two Brussels municipalities, in Belgium, for the 2100 MHz and 2600 MHz frequency bands. It is shown that the proposed model fits experimental values, paving the way to a new methodology to assess general public exposure to electromagnetic fields, for any frequency band. An insight is given on how to apply the methodology to a real case without access to experimental data.

In this work, the results of Ultra-Wideband air-to-ground measurements carried out in a real-world factory environment are presented and discussed. With intelligent industrial deployments in mind, we envision a scenario where the Unmanned Aerial Vehicle can be used as a supplementary tool for factory operation, optimization and control. Measurements address narrow band and wide band characterization of the wireless radio channel, and can be used for link budget calculation, interference studies and time dispersion assessment in real factories, without the usual limitation for both radio terminals to be close to ground. The measurements are performed at different locations and different heights over the 3.1-5.3 GHz band. Some fundamental propagation parameters values are determined vs. distance, height and propagation conditions. The measurements are complemented with, and compared to, conventional ground-to-ground measurements with the same setup. The conducted measurement campaign gives an insight for realizing wireless applications in smart connected factories, including UAV-assisted applications.

A measurement setup, including millimeter-wave and ultrawideband transceivers mounted on both a customized unmanned aerial vehicle and a ground station and a measurement procedure for full 3-D wireless propagation analysis, is described in this work. The custom-made developed system represents a flexible solution for the characterization of wireless channels and especially of urban propagation, as the drone might be easily located almost anywhere from ground level to the buildings’ rooftop and beyond. The double-directional properties of the channel can be achieved by rotating directive antennas at the link ends. Other possible applications in urban contexts include the study of outdoor-to-indoor penetration, line-of-sight to non-line-of-sight transition, 3-D scattering from buildings, and air-to-ground channel characterization for drone-assisted wireless communications.

A novel localization framework is presented based on ultra-wideband (UWB) channel sounding, employing a triangulation method using the geometrical properties of propagation paths, such as time delay of arrival, angle of departure, angle of arrival, and their estimated variances. In order to extract these parameters from the UWB sounding data, an extension to the high-resolution RiMAX algorithm was developed, facilitating the analysis of these frequency-dependent multipath parameters. This framework was then tested by performing indoor measurements with a vector network analyzer and virtual antenna arrays.The estimated means and variances of these geometrical parameters were utilized to generate multiple sample sets of input values for our localization framework. Next to that, we consider the existence of multiple possible target locations, which were subsequently clustered using a Kim-Parks algorithm, resulting in a more robust estimation of each target node. Measurements reveal that our newly proposed technique achieves an average accuracy of 0.26, 0.28, and 0.90 m in line-of-sight (LoS), obstructed-LoS, and non-LoS scenarios, respectively, and this with only one single beacon node. Moreover, utilizing the estimated variances of the multipath parameters proved to enhance the location estimation significantly compared to only utilizing their estimated mean values.

Diffuse scattering (DS), mostly caused by macroscopically rough surfaces, can be modeled by an effective roughness (ER) approach. In the ER approach, the object surface is divided into tiles, and the DS field amplitude associated with each tile is given. Assuming that the transmit antenna (or emitted field) is fixed and the phases associated with different tiles are independent, this paper proposes a phase evolution model based on the relative spatial locations of the receive antenna (Rx) to each tile. The proposed model contains two parts: the deterministic part that depends on the variation of the distance from Rx to tile center, and the correlated random part that depends on the variation of the angle between the Rx and the normal vector of tile. The correlation of the tile field phases resulting from the angular variation is modeled by the exponentially damped cosine correlation function. The proposed phase evolution model is evaluated by being applied to the ER approach to predict the DS component of radio channel transfer function (CTF). The predicted DS-CTF is compared with the reference (simulated data by physical optics and measured data in well-controlled environment) in terms of the spatial autocorrelation as well as the spatial Doppler spectrum.

As wireless connectivity becomes increasingly ubiquitous, a greater emphasis will be placed upon the seamless integration of dissimilar networking technologies. One such example of this will occur in urban environments, where wearable devices and vehicular networks will operate in close proximity to one another. Clearly, a natural extension to both types of network is their interconnectivity through vehicle-to-pedestrian (V2P) or equivalently pedestrian-to-vehicle (P2V) communications as part of a much greater vehicle-to-X (V2X) based Intelligent Transportation System (ITS). To this end, we empirically investigate the P2V communications channel at 5.8 GHz for the case of a moving vehicle when a person positioned by the edge of a road was either stationary or walking parallel to the side of the highway. The measurements considered a chest mounted transmitter and four receiver locations on the vehicle covering the front wing mirrors and two positions on the roof, which simultaneously recorded the received signal power. To characterize the propagation mechanisms which are responsible for shaping the received signal in the P2V channel we decomposed it into its path loss, large-scale and small-scale fading components. We first show that although there was evidence of interference caused by multiple rays interacting with one another, the popular Two- Ray ground-reflection path loss model was unable to adequately describe the compounded effects of the vehicle and pedestrian’s body on the signal attenuation in the majority of the considered scenarios. Instead, we found that the overall path loss was well characterized using a dual-slope log-distance model, with lognormal large-scale fading. Due to the often severe small-scale fading that was observed in the P2V channel, we have been able to utilize the kappa-mu Extreme distribution with considerable success to characterize the worse than Rayleigh fading conditions which were encountered.

This paper proposes a hybrid modeling approach for prediction of the room-to-room radio propagation channel. The model combines ray tracing with propagation graph. The propagation graph vertices are obtained at each room by ray tracing with the assumption that the receive antenna (or transmit antenna) virtually locates on the surface of the separating wall between two rooms. Rays transmitted from one room to the other through the separating wall are deterministically calculated by Snell’s law of refraction. Predictions by the proposed model are compared with measurement data from an office-to-office scenario. The results show that the proposed modeling works with the simplest parameter settings, i.e. assuming no propagation from the room containing receive antenna to the room containing transmit antenna, ray tracing applied separately in each room only involves mechanism of line-of-sight and first order specular reflection.