Wireless Friendliness Evaluation of Building Materials as Reflectors

The enormous popularity of wireless devices has prompted a rapid growth of indoor
wireless traffic. To meet the high data demand and avoid inconvenient usages of a room,
indoor base stations (BSs) and Wi-Fi access points (APs) with large-scale multiple-input
multiple-output (MIMO) antenna arrays are likely to be deployed in the vicinity of a wall,
which therefore results in non-negligible interactions between indoor electromagnetic (EM)
wave propagations and building materials. The reflection characteristics of building materials,
which depend on their intrinsic EM and physical properties, play a crucial role in indoor
wireless communications. However, the relationship between the material properties and the
indoor wireless performance has not been sufficiently studied.

In this thesis, wireless friendliness is proposed as a new metric to measure the impact
of a building material on indoor wireless performance as a function of its EM and physical
properties. The main objectives are to develop wireless friendliness evaluation schemes for
building materials as reflectors on indoor line-of-sight (LOS) MIMO communications, and to
provide insights into the appropriate design and/or selection of building materials according
to their wireless friendliness.

To achieve these objectives, the thesis presents four major contributions. The first
contribution is to propose a new two-ray channel model and a new multipath channel model
that incorporate both the LOS path and the wall reflection (WR) path for indoor LOS
MIMO downlink transmissions. For the first time, the relative permittivity (EM property)
and thickness (physical property) of a building material are encapsulated into the channel
models through the reflection coefficient of the building material, which provides theoretical
prerequisites for the subsequent tractable analysis.

The second contribution is to reveal the analytical relationship between the relative permittivity
and thickness of building materials and the MIMO channel capacity. By exploiting
the expressions of indoor wireless capacity and their asymptotic forms, four effective metrics
for evaluating the wireless friendliness of building materials are proposed, i.e., the spatially
averaged capacity, the spatially averaged logarithmic eigenvalue sum (LES), the spatially
averaged logarithmic eigenvalue product (LEP), and the upper-bound outage probability,
which are all over the room of interest.

The third contribution is to develop the evaluation schemes for the wireless friendliness
of building materials. The optimal values of the relative permittivity and thickness of a
building material that maximise the indoor wireless capacity are obtained, shedding light on
the selection and/or design of a building material accordingly, and thus paving the way for
wireless friendly architectural design.

The fourth contribution is to analyse the effects of the WR from building materials
on the per-antenna power distribution across a precoded antenna array at a BS or an AP
deployed near a wall. An uneven power distribution across antenna elements may reduce
the efficiencies of their corresponding radio frequency (RF) power amplifiers. How the
per-antenna power distribution changes with the building material’s relative permittivity and
thickness is investigated, providing guidelines on the selection and/or design of a building
material that alleviates the unevenness of per-antenna power distribution.

Simulation results validate the correctness of analytical results as well as the effectiveness
of the four proposed evaluation metrics, and demonstrate that the EM and physical
properties of building materials have to be delicately selected or designed to avoid the risk of
reducing indoor wireless capacity and RF power amplifier efficiency. More specifically, the
inappropriate choices of relative permittivity and thickness of a building material may reduce
the indoor wireless capacity by up to 13.5% or cause severe unevenness as large as 8 dB in
the per-antenna power distribution across a precoded antenna array. The outcomes of this
thesis would enable appropriate design and/or selection of building materials for building
designers, e.g., civil engineers and architects, and provide wireless-friendliness information
for communications engineers.