Page 53 - ITU Journal: Volume 2, No. 1 - Special issue - Propagation modelling for advanced future radio systems - Challenges for a congested radio spectrum

ITU Journal: ICT Discoveries, Vol. 2(1), December 2019

DIRECTIONAL ANTENNA CHANNEL MODELING IN URBAN AREA USING RAY TRACING

1
1
1
2
Zhuyin Li , Sithamparanathan Kandeepan , Andrea Giorgetti , Akram Al-Hourani , Kagiso Magowe 1
1 School of Engineering, RMIT University, Melbourne, Australia, DEI, University of Bologna, Italy
2
Abstract – Directional antennas are regarded as one of the key technologies to achieve higher signal quality and data rate.
Evidence has shown that the directional antenna channel features can be very different from the omnidirectional ones. There-
fore, it is essential to characterize the directional antenna channel model (DACM), since an accurate, easy-to-implement DACM
plays a vital role in the wireless network design, optimization and utilization. In this paper, we cluster the incident rays based
on the azimuth direction of arrival at the receiver and extract the features concerning the number of clusters, the distribution
of the cluster center and the cluster power ratio. The proposed DACM is derived by the ray tracing method and is feasible for a
typical urban scenario within a range of a few hundred meters. Moreover, our proposed model is veri ied by both ray tracing
and numerical simulations. In this paper, we present the model derivation methodology, recommend the model parameters
and offer an implementation guideline.
Keywords – Directional channel modeling, direction of arrival, multipath clustering, radio propagation, ray tracing

1. INTRODUCTION geometry-based technique, can provide an accurate esti-
mation of the radio propagation with high computational
5G networks are born with the tags of high capac- speed [6]. In a ray tracing simulation (RTS), the software
ity, high energy ef iciency and high data rate. In the estimates possible paths between the transmitter and the
last decades, multiple-input-multiple-output (MIMO) has receiver depending on the environment, and offers a list
been regarded as one of the most promising techniques to of outputs including path loss, DOA, DOD, delay etc. Fur-
enable some of the 5G vital features [1]. Directional an- thermore, even though the RTS can offer a great number
tennas can improve propagation quality by focusing the of multipath components (MPCs), it is not necessary to
beam within some desired direction. We can, therefore, deal with MPCs individually. Many measurements have
achieve longer transmission distance with higher perfor- revealed that MPCs can be grouped into clusters with sim-
mance and lower interference by enhancing the signal ilar features such as DOA, DOD or delay. Such cluster-
strength in the speci ic direction and suppressing the sig- based channel models are more compact and intuitive
nal in other directions [2]. For instance, higher frequency since the intra-cluster, and inter-cluster statistics can be
bands above 6 GHz e.g. millimeter-wave (mmWave), as treated separately [7].
part of the 5G spectrum offer huge bandwidth, but at the
same time, bring problems like high path loss, high pen- Considering the urge to have an accurate channel model
etration loss, limited delay spread and re lection domi- in a directional propagation environment, we propose a
nated propagation [3]. Therefore, directional antennas directional antenna channel model (DACM) and also offer
are deployed to overcome such challenges. On the other a generic methodology framework for channel modeling.
hand, theutility of directionalantennasatboth traditional In this paper, our main contributions can be summarized
and higher frequency bands can bring new challenges as follows: 1) We derive a clustering DACM using the ray
to the wireless network design, development and imple- tracing method. The model is feasible for a typical urban
mentation, with the increasing demand for the accuracy scenario in which the transmitter and the receiver are a
of the channel modeling [4]. Moreover, the angular infor- fewhundredmetersapart, suchasanurbanmicrocellnet-
mation such as direction of arrival (DOA) and direction of work. 2) We cluster the MPCs based on the azimuth DOA,
departure (DOD), cannot be neglected in directional chan- and further extract the probability mass function (PMF)
nels, because they are not only important features for the of the number of clusters, the probability density func-
propagation model, but also affect other parameters like tion (PDF) of the azimuth cluster center and the cluster
path loss and delay [5]. As a result, the traditional omni- power ratio. 3) The derived PDF expressions are simple
directional channel models cannot accurately estimate a functions of only a few distance-independent parameters.
directional antenna propagation.

Channel models are essential for wireless network design Theremainderofthepaperisorganizedasfollows. Insec-
and analysis, since they give insights into the character- tion 2, a literature review on the existing directional chan-
istics of the radio channel without conducting ield mea- nel model is presented. We also present the system model
surements, which are usually time-consuming and costly. in section 3 and the ray tracing simulation method in sec-
Among all the channel modeling methods, ray tracing, as a tion 4. The proposed DACM is presented in section 5, fol-

© International Telecommunication Union, 2019 37