Transmissive RIS for 6G Communications: Design, Prototyping, and Experimental Demonstrations
Extracted source text · arXiv 2206.15133
%\documentclass[10pt,twocolumn,twoside]{IEEEtran}
%\documentclass[11pt,journal,onecolumn]{IEEEtran}
\documentclass[12pt,draftclsnofoot,journal,onecolumn]{IEEEtran}
%\documentclass[12pt,draftclsnofoot,journal,onecolumn]{IEEEtran}
%\documentclass[journal]{IEEEtran}
%\linespread{1.6}
%\documentclass[conference]{IEEEtran}
%\documentclass[10pt,conference,twocolumn,]{IEEEtran}
%\setlength{\topmargin}{-0.7in}
%\documentclass[a4paper,12pt]{article}
% correct bad hyphenation here
%\hyphenation{op-tical net-works semi-conduc-tor IEEEtran}
\usepackage{graphicx}
\usepackage{booktabs}
%\DeclareGraphicsRule{.png}{eps}{.bb}{}
%\usepackage{mathrsfs} then \mathscr{FP}
%\usepackage{amsmath} then \mathcal{FP}
%\usepackage[Symbol]{upgreek}
\usepackage{subfigure}
\usepackage{color}
%\usepackage{setspace}
\usepackage{multicol}
\usepackage{amssymb}
\usepackage{multirow}
\usepackage{amsmath}
\usepackage{bm}
\usepackage{cite}
\usepackage{gensymb}
\usepackage{amsfonts}
\usepackage{mathrsfs}
\usepackage{amsmath}
\usepackage{algorithm}
\usepackage{algorithmic}
\usepackage{amsthm}
\renewcommand{\algorithmicrequire}{\textbf{Inputs:}}
\renewcommand{\algorithmicensure}{\textbf{Output:}}
\usepackage{tabularx}
\usepackage{balance}
\usepackage{mathrsfs}
%\usepackage{hyperref}
\usepackage{array}
\usepackage{multirow}
\newcommand{\PreserveBackslash}[1]{\let\temp=\\#1\let\\=\temp}
%\newcommand{\tabincell}[2]{\begin{tabular}{@{}#1@{}}#2\end{tabular}}
\newcolumntype{C}[1]{>{\PreserveBackslash\centering}p{#1}}
\newcolumntype{L}[1]{>{\PreserveBackslash\raggedright}p{#1}}
\newcolumntype{R}[1]{>{\PreserveBackslash\raggedleft}p{#1}}
\usepackage[flushleft]{threeparttable}
\newcommand{\mb}[1]{{ \mathbf #1}} %\mathbf \bm
\newtheorem{thm}{Lemma}
%\setlength\abovedisplayskip{3pt}
%\setlength\belowdisplayskip{3pt}
\begin{document}
\bibliographystyle{IEEEtran} % use IEEEtran.bst style
\title{Transmissive RIS for 6G Communications: Design, Prototyping, and
Experimental Demonstrations}
\author{\IEEEauthorblockN{ Junwen Tang, Mingyao Cui, Shenheng Xu,
\emph{Member, IEEE}, Linglong Dai, \emph{Fellow, IEEE}, Fan Yang,
\emph{Fellow, IEEE}, and Maokun Li, \emph{Senior Member, IEEE}}
\thanks{ Part of this work has been accepted by 2020 IEEE
International Symposium on Antennas and Propagation and North American
Radio Science Meeting \cite{1}.
All the authors are with the Department of Electronic
Engineering, Tsinghua University, Beijing 100084, China (e-mails:
tjw17@mails.tsinghua.edu.cn; cmy20@mails.tsinghua.edu.cn;
\{shxu, daill, fan\_yang, maokunli\}@tsinghua.edu.cn).
}
\vspace{-4mm}
}
\maketitle
\IEEEpeerreviewmaketitle
\begin{abstract}
Reconfigurable intelligent surface (RIS) has been widely considered as a
key technique to improve spectral efficiency for 6G communications.
Compared with most existing research that only focuses on the reflective RIS,
the design and prototyping of a novel transmissive RIS are presented in this
paper, and its enhancement to the RIS-aided communication system is
experimentally demonstrated. The 2-bit transmissive RIS element utilizes the
penetration structure, which combines a 1-bit current reversible dipole and a
90° digital phase shifter based on a quadrature hybrid coupler. A transmissive
RIS prototype with 16×16 elements is designed, fabricated, and measured to
verify the proposed design. The measured phase shift and insertion loss of the
RIS element validate the 2-bit phase modulation capability. Being illuminated
by a horn feed, the prototype achieves a maximum broadside gain of 22.0 dBi at
27 GHz, and the two-dimensional beamforming capability with scan angles up to
±60° is validated. The experimental results of the RIS-aided communication
system verify that by introducing the extra gain and beam steering capability,
the transmissive RIS is able to achieve a higher data rate, reduce the
transmit power, improve the transmission capability through obstacles, and
dynamically adapt to the signal propagation direction.
\end{abstract}
\begin{IEEEkeywords}
Transmissive RIS, 2 bit phase shifter, prototyping \vspace{-3mm}
\end{IEEEkeywords}
\section{Introduction} \label{sec:1}
With the explosive growth of the emerging new applications, such as holographic
video and meta-verse, 6G is expected to achieve a 10$\times$ increase in
spectral efficiency compared to 5G \cite{2}, \cite{3}. To
realize this goal, the reconfigurable intelligent surface (RIS) has become an
essential candidate \cite{4}. By adjusting the phase and amplitude
responses to an
incident electromagnetic wave, RIS can achieve dynamic beamforming to enhance
the spectral efficiency and overcome blockage \cite{5}.
Nevertheless, existing research contributions mainly consider the employment of
reflective RIS \cite{6, 7, 8}, which
may result in
coverage holes
in a cell.
Specifically, for reflective RIS, the base station and user need to be
located on the same side of the RIS, which brings extra geographical
constraints on the physical topology \cite{9}. For instance,
reflective RIS is
difficult to assist communication between a transmitter outside a vehicle and a
receiver inside it. To cope with this problem, transmissive RIS
has been recently proposed \cite{10}. Rather than being reflected,
signals can
transmit through a transmissive RIS to form directional beams. In this way,
transmissive RIS can potentially fill in the coverage holes of reflective RIS.
The phase reconfigurability is the most essential capability for a transmissive
RIS. Solid-state electronic devices are commonly integrated in each constituent
RIS element to dynamically control its phase response. Continuous phase shifts
can be realized using analog-type devices like varactor diodes
\cite{10, 11, 12, 13, 15, 14, 16, 17}. The
phase shift range usually exceeds 360° so that the phase errors are negligible.
However, the transmissive RIS has a very high transmission insertion loss,
which can be up to 5.7 dB \cite{13}. As a consequence, the
aperture
efficiency of the RIS is considerably reduced. At high frequencies, switch-type
devices are widely used to produce discrete phase shifts, including PIN
diodes, RF MEMS switches, and
etc
\cite{13,14,15,16,17}. Most existing designs focus on 1-bit phase
reconfigurability, wherethe
design and fabrication difficulties are manageable
\cite{18,20,21,23,24,22,25,26,19}. For
example, a
classic design of an O-slot patch loaded with two PIN diodes was developed in
\cite{18} to achieve a stable 180° phase shift in a large bandwidth using
the
current reversal mechanism. The minimum measured insertion loss is 1.7/1.9 dB
at 9.8 GHz. Several other attempts were made to obtain a 2-bit phase resolution
\cite{27,28}. Specifically, a monolithically fabricated 2-bit prototype using
MEMS
switches \cite{27} was able to produce 4 phase states, but the measured
insertion
loss is 4.2–9.2 dB at Ka band. A compact 2-bit element using a modified O-slot
patch \cite{28} was designed, fabricated and measured. The measured insertion
loss
is 1.5–2.3 dB at 29.0 GHz, and the measured 3-dB transmission bandwidth is
10.1–12.1\% for four phase states.
The phase quantization errors associated with discrete phase shifts inevitably
introduce performance degradation \cite{29,30,31}. It is observed that the
1-bit
element designs suffer from 3–4 dB insertion loss attributed to the coarse
phase resolution, which results in a low aperture efficiency. The sidelobe
levels are significantly higher, sometimes even causing unwanted grating lobes.
By contrast, 2-bit designs are considered a well-balanced choice between the
element performance and the design complexity. The phase quantization loss can
be greatly reduced to less than 1 dB, and the sidelobe envelop can be
significantly improved as well.
Nonetheless, there are only a few 2-bit designs in the literature. One of the
important reasons is that, compared to 1-bit designs, more transition
structures, electronic devices and DC/RF biasing circuits are needed to achieve
four stable phase states. The conventional design approach solely relies on the
two-dimensional unit-cell footprint because of the printed-circuit board (PCB)
fabrication process. It becomes extremely challenging to lay out a variety of
constituent element parts in a greatly limited sub-wavelength array grid.
Consequently, more numbers of stacked layers are used, which results in the
higher fabrication complexity and cost. To the best of our knowledge, the
design, fabrication, and measurement of the 2-bit transmissive RIS have not
been well studied.
In this paper, the 2.5-D penetration structure \cite{32} is exploited to make
the
best of the longitudinal space in the propagation direction to develop the
2-bit transmissive RIS. Specifically, our contributions are summarized
below.
\begin{itemize}
\item Firstly, a 90° digital phase shifter with twoPIN diodes based on a
quadrature hybrid coupler \cite{33} is proposed and realized. Combined
with a
1-bit current reversible dipole, 2-bit phase resolution is achieved at the
millimeter wave (mmWave) band. In addition to some preliminary results
reported in the authors’ conference paper \cite{1}, the 2-bit RIS element
is
further optimized and more detailed analysis is presented in this journal
paper.
\item Furthermore, a prototype with 16×16 RIS elements and a logic circuit
control board is fabricated and measured. The measured results show that
the prototype can achieve a maximum broadside gain of 22.0 dBi at 27 GHz,
with the corresponding aperture efficiency of 25.3\%. Moreover, the
measured radiation performances for scan angles are presented as well,
demonstrating the beamsteering capability in a large angle up to 60°.
\item Finally, a transmissive RIS-aided mmWave wireless communication
prototype is developed to demonstrate the performance of the fabricated
2-bit transmissive RIS. The experimental results verify that the 2-bit
transmissive RIS can achieve a higher data rate, reduce the transmit power,
improve the transmission capability through obstacles, and dynamically
adapt to the signal propagation direction.
\end{itemize}
\emph{Organization}: The reminder of this paper is organized as follows. The
communication model of transmissive RIS is introduced in Section II. Then, in
Section III, we describe the structure, biasing circuits, and simulated
performance of the element of the designed transmissive RIS. Section IV
provides the experimental
results for the 16×16 elements RIS prototype in the microwave anechoic chamber.
A transmissive RIS-aided wireless communication prototype is set up to further
measure the performance of the proposed RIS in Section V. Finally in Section
VI, conclusions are drawn.
\section{System Model} \label{sec:2}
In this section, we describe the system model of the transmissive
RIS-aided communication system.
\subsection{System Model} \label{sec:2-1}
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=3.5in]{figure/fig1.png}}
\caption{ System model of transmissive RIS-aided communication systems.
}
\label{fig1}
\vspace*{-1em}
\end{figure}
We consider that the transmitter (Tx) employs a horn antenna to communicate
with a single-antenna receiver (Rx), where an N-element transmissive RIS is
used for enhancing the signal power. As shown in Fig. \ref{fig1}, the RIS is a
uniformly
planar array (UPA) placed in the x-y plane, and its geometric center is located
at the origin of the coordinate system. Let $N_x$ denote the number of antennas
on the row and $N_y$ denote the number
of antennas on the column, with $N = N_xN_y$. The antenna spacing
along the $x$ axis and $y$ axis are $d_x$ and $d_y$ respectively, which are
usually half of the wavelength.
The coordinates of the Tx and Rx are $(x_t,y_t,z_t)$ and $(x_r, y_r, z_r)$.
The coordinate of the $(m, n)$-th antenna in the $m$-th row and $n$-th column
of the RIS is $(\delta_md_x, \delta_nd_y, 0)$, where $\delta_m = m - (N_x -
1)/2$ and $\delta_n = n - (N_y - 1)/2$ with $m \in\{0,1,\cdots,N_x - 1\}$ and
$n \in \{0,1,\cdots,N_y - 1\}$. We use the symbols $d_t$, $\theta_t$, and
$\phi_t$ to denote the distance, the azimuth angle, and the elevation angle
between the center of the RIS and the Tx. The relationship between $(x_t, y_t,
z_t)$ and $(d_t, \theta_t, \phi_t)$ is shown in Fig. \ref{fig1}. Similarly,
we use the
symbols $d_r$, $\theta_r$, and $\phi_r$ to denote the distance, the azimuth
angle, and the elevation angle between the center of the RIS and the Rx, where
the relationship between $(x_r, y_r, z_r)$ and $(d_r, \theta_r, \phi_r)$ is
shown in Fig. \ref{fig1} as well.
We denote $\mb{f} \in \mathbb{C}^{N\times 1}$ as the Tx-RIS channel,
and denote $\mb{g} \in \mathbb{C}^{N\times 1}$ as the RIS-Rx channel.
Since this paper aims to evaluate the function of the transmissive RIS, the
direct link between the Tx and Rx is ignored in this paper. Denote
$[\mb{g}]_{m,n}$ and $[\mb{f}]_{m,n}$ as
the channel from the $(m, n)$-th antenna of the RIS to the Rx and Tx,
respectively. Then, the free space channel model is adopted to represent
$[\mb{g}]_{m,n}$ and $[\mb{f}]_{m,n}$ as \cite{4}
\begin{align}
[\mb{g}]_{m,n} = \sqrt{\frac{\lambda G_g F_g(\theta_r, \phi_r)}{4\pi}}
\frac{e^{-j\frac{2\pi}{\lambda}d_{m,n}^r}}{d_{m,n}^r}, \\
[\mb{f}]_{m,n} = \sqrt{\frac{\lambda G_f F_f(\theta_t, \phi_t)}{4\pi}}
\frac{e^{-j\frac{2\pi}{\lambda}d_{m,n}^t}}{d_{m,n}^t},
\end{align}
where $\lambda$ represents the wavelength, $d_{m,n}^r$ and $d_{m,n}^t$
denote the distance from the $(m, n)$-th antenna of the RIS to the Rx and Tx,
respectively. Moreover, $G_g$ and $F_g(\theta_r, \phi_r)$
denote the antenna gain and the normalized power radiation pattern of the side
of RIS facing the receiver, while $G_f$ and $F_f(\theta_t, \phi_t)$
correspond to those of the side of RIS facing the transmitter \cite{4}.
Let $s$ denote the transmitted symbol with $\|s\|_2 = 1$.
We use symbols $G_t$, $F_t$, $P_t$, $G_r$, $F_r$ to represent the transmitter
antenna gain, the transmit power, the normalized radiation power of the
transmitter,
the receiver antenna gain, and the normalized radiation power of the receiver.
Then
the noiseless received signal by the RX can be modeled as \cite{4}
\begin{align}\label{eq:channel}
y &= \sqrt{P_tGF}\sum_{m=0}^{N_x - 1}\sum_{n=0}^{N_y -
1}[\mb{g}]_{m,n}
[\mb{f}]_{m,n}\Gamma_{m,n}e^{j\phi_{m,n}}s, \notag\\
&=\frac{\sqrt{P_tGF\lambda}}{4\pi}\sum_{m=0}^{N_x - 1}\sum_{n=0}^{N_y -
1}\frac{\Gamma_{m,n}}{d_{m,n}^td_{m,n}^r}e^{j\left(\phi_{m,n}
- \frac{
d_{m,n}^t+d_{m,n}^r}{\lambda}\right)}s,
\end{align}
where $\Gamma_{m,n}\in[0,1]$ is transmission loss of the $(m,n)$-th antenna of
the RIS, $\phi_{m,n}\in[0, 2\pi]$ denotes the phase shift of the $(m,n)$-th
antenna of the RIS, $G = G_tG_fG_gG_r$, and $F = F_tF_f(\theta_r,
\phi_r)F_g(\theta_r, \phi_r)F_r$.
\subsection{Beamforming for the Transmissive RIS} \label{sec:2-2}
For the desired receiver, the transmissive RIS is designed for maximizing the
received signal energy through beamforming. The noiseless
received signal energy can be presented as
\begin{align}
P_r = \frac{P_tGF\lambda^2}{16\pi^2}\left|\sum_{m=0}^{N_x - 1}\sum_{n=0}^{N_y -
1}
\frac{\Gamma_{m,n}}{d_{m,n}^td_{m,n}^r}e^{j\left(\phi_{m,n}
- \frac{
d_{m,n}^t+d_{m,n}^r}{\lambda}\right)}\right|^2.
\end{align}
To maximize $P_r$, it is obvious that the optimal $\phi_{m,n}$ is
\begin{align}
\bar{\phi}_{m,n} = \mod\left(C + \frac{
d_{m,n}^t+d_{m,n}^r}{\lambda}, 2\pi\right).
\end{align}
Here, $C$ is an arbitrary constant.
From (5), the phase shift $\phi_{m,n}$ can be obtained according to
$d_{m,n}^t$ and $d_{m,n}^r$. In our practical employment, the RIS is deployed
nearby the Rx and far away from the Tx. That is to say, the Rx is within the
near-field range of the RIS, while the Tx is within the far-field range of the
RIS. Based on this assumption, the distance $d_{m,n}^r$ can be derived
according to the
spherical wave model as
\begin{align}
d_{m,n}^r = \sqrt{(x_r - \delta_md_x)^2 + (y_r - \delta_nd_y)^2 + z_r^2},
\end{align}
while the distance $d_{m,n}^t$ can be simplified by the planar wave model as
\begin{align}
d_{m,n}^t &= \sqrt{(x_t - \delta_md_x)^2 + (y_t - \delta_nd_y)^2 + z_t^2}
\notag\\
&\approx r_t - \delta_md_x\sin\theta_t\cos\phi_t - \delta_nd_y\sin\theta_t
\sin\phi_t.
\end{align}
Therefore, the optimal solution to $\phi_{m,n}$ can be obtained accordingly.
Then, for practical implementation of RIS, the limited resolution of the phase
shift has to be considered. If there are $b$ bits for quantized phase shifter,
then the feasible set of phase shift is
\begin{align}
\mb{\Phi}_b = \left\{0, \frac{1}{2^{b-1}}\pi,\cdots,
\frac{2^{b}-1}{2^{b-1}}\pi\right\}.
\end{align}
Finally, the practical phased shift on a RIS element is given by
\begin{align}
\tilde{\phi}_{m,n} = \arg\min_{\phi \in \mb{\Phi_b}} |\phi-\bar{\phi}_{m,n}|.
\end{align}
In our system, we design a transmissive RIS with $b = 2$ bits. Notice that the
transmission model and the phase shift design mathematically have
no difference with the reflective RIS-aided communication system.
Nevertheless, the application scenarios and implementation and of them are
quite different. For instance, a transmissive RIS can be deployed on a vehicle
window to enhance communication between inside and outside the vehicle. In the
following sections, we will elaborate on the proposed transmissive RIS design
in our communication system.
\section{Proposed RIS Element Design}
We describe the proposed transmissive RIS element in this section,
including element design, 90° phase shifter exploited in the element, biasing
circuits, and element performance in simulation.
\subsection{Element Design}
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=3.5in]{figure/fig2.png}}
\caption{ Proposed 2.5-D structure of the transmissive RIS for mmWave
communications.}
\label{fig2}
\vspace*{-1em}
\end{figure}
{
\begin{table}[]
\caption{Simulated Performance of the 2-bit RIS Element}
\centering
\begin{tabular}{lccc}
\toprule
\multicolumn{1}{c}{State} & \begin{tabular}[c]{@{}c@{}}Element loss\\
@26.5 GHz\end{tabular} &
\multicolumn{1}{c}{\begin{tabular}[c]{@{}c@{}}Phase shift\\ @26.5
GHz\end{tabular}} & \multicolumn{1}{c}{\begin{tabular}[c]{@{}c@{}}3-dB
trans.\\ bandwidth\end{tabular}} \\
\midrule
1 (state-0$^{\circ}$) & 1.1
dB &
-141.2$^{\circ}$
&
23.6-28.9 GHz
(20.1\%) \\
\midrule
2 (state-90$^{\circ}$) & 1.3
dB &
-56.8$^{\circ}$
&
24.8-28.6 GHz
(14.2\%) \\
\midrule
3 (state-180$^{\circ}$) & 1.1
dB &
34.9$^{\circ}$
&
23.5-28.6 GHz
(19.6\%) \\
\midrule
4 (state-270$^{\circ}$) & 1.5
dB &
129.0$^{\circ}$
&
24.6-29.0 GHz
(16.4\%) \\
\bottomrule
\end{tabular}
\end{table}
}
Fig. \ref{fig2} illustrates the proposed 2.5-D structure of the 2-bit
transmissive RIS. Each RIS element consists of a number of one-dimensional
subarrays aligned in parallel. The constituent element utilizes the
penetration structure, which
is composed of a RIS-side Rx and a RIS-side Tx to receive and transmit the RF
signals impinging on the RIS, respectively, and a microstrip transmission line
to convey the signal through the RIS aperture. Instead of laying them out in
the aperture plane, two vertical dipoles \cite{35} are employed and arranged
along
the propagation direction to circumvent the space constraint imposed by the
unit cell footprint. The slightly extended longitudinal space easily
accommodates a 90° digital phase shifter between the RIS-side Rx and the
RIS-side Tx. Combined with the 1-bit current reversible dipole, the 2-bit
transmissive RIS element can be readily realized.
The 2.5-D structure slightly increases the aperture profile by 1~2 wavelengths,
but it is still much smaller compared with the aperture size. Especially at
mmWave frequencies, the aperture profile is merely a couple of centimeters and
will not considerably undermine the low profile and conformal features of the
RIS. Each subarray can be easily fabricated on a single layer of substrate,
thus avoiding the complicated multilayer PCB process and lowering the overall
cost. It is also a remarkable fact that the $//$-type 2.5-D structure can be
further developed into a $\#$-type or a $\Delta$-type formation where two or
three sets
of subarrays can be arranged in the same aperture, so that the 2.5-D unit cell
space can be more fully exploited for more constituent element parts. Hence,
more advanced features such as dual-polar or dual-band designs can be
accomplished.
The linearly incident wave propagating along the z axis is received by the
passive RIS-side Rx composed of a vertical dipole. Then, the received signal is
converted into a guided wave and conveyed along the microstrip transmission
line. The phase shifter structure based on the quadrature hybrid coupler is
employed. Instead of the varactor diodes used in the analog phase shifter
design [33], it is terminated with reflection-type loads composed of two p-i-n
diodes (model MADP-000907-14020) to provide a 0°/90° digital phase shift. The
two loaded p-i-n diodes are always tuned at the same ON- or OFF-states,
resulting in a 90° phase shift. The RIS-side Tx is similar to the RIS-side Rx,
but the vertical dipole structure \cite{35} is modified to integrate two p-i-n
diodes (MADP-000907-14020) that are symmetrically placed in the feeding
microstrip line. They are alternatively turned ON or OFF so that the excitation
current flows in opposite directions. Consequently, the reversible current flow
produces a stable 180° phase shift without significantly affecting magnitude
response. Finally, the RIS-side Tx re-radiates the guided wave into the free
space with four stable phase states of 0°/90°/180°/270°, and hence, the 2-bit
phase resolution is obtained.
The frequency of the designed transmissive RIS element is 26.5 GHz, and the
unit
cell size is $4.9\times4.9$ $\text{mm}^2$. The proposed 2-bit element is
printed
on a single layer of Taconic TLX-8 substrate ($\epsilon_r$ = 2.55,
$\tan\delta$= 0.0019, $h$ =
0.254 mm). The microstrip line and the 90° digital phase shifter are on the top
of the substrate, while the vertical dipoles and the ground plane are arranged
on the bottom of the substrate. The total profile of the element is 21.8 mm,
corresponding to 1.9$\lambda$ at the design frequency. The following
geometrical
parameters are optimized after a comprehensive parametric study: $l_d$ = 2.1
mm,
$wd$ = 0.6 mm, $l_s$ = 1.6 mm, $w_s$ = 0.3 mm.
\subsection{90° Digital Phase Shifter}
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=5in]{figure/fig3.png}}
\caption{ Proposed 90° digital phase shifter based on a quadrature hybrid
coupler: (a) the illustrative structure; (b) simulated reflection
magnitude, (c) transmission magnitude, and (d) transmission phase.}
\label{fig3}
\vspace*{-1em}
\end{figure}
The proposed 90° digital phase shifter integrated in the 2-bit transmissive RIS
element is designed to provide 0°/90° phase switching, as shown in Fig.
\ref{fig3} (a).
It is a reflection-type phase shifter based on the quadrature coupler
terminated with two reflective loads, which are composed of p-i-n diodes and
short-circuited stubs. The realization of 90° phase shift is by tuning the two
p-i-n diodes to work at ON-ON or OFF-OFF state, respectively. Two tuning stubs
are introduced to maintain a stable desired phase shift by tuning their
lengths. The digital phase shifter is modeled and simulated with a periodic
boundary condition and a wave-port excitation. As can be observed from the
simulation results plotted in Fig. \ref{fig3} (b) , a 10-dB return loss
bandwidth of
22\% (22.6-28.2 GHz) is achieved.
Good impedance matching performance minimizes
the insertion loss to less than 1 dB within the frequency band from 24.2 to
28.0 GHz. Meanwhile, from the curves plotted in Fig. \ref{fig3} (d) , we can
find that
the phase difference is close to 90° in a wide frequency band. We can conclude
that the phase shifter is capable of generating desired 90° phase shift by
electrically adjusting the two p-i-n diodes, thus achieving the 0°/90° digital
phase switching capability with the low insertion loss.
\subsection{Performance of the designed Transmissive RIS Element}
\begin{figure}
\centering
% \vspace*{-0.5em}
\subfigure[]
{\includegraphics[width=3in]{figure/fig4a.png}}
\subfigure[]
{\includegraphics[width=3in]{figure/fig4b.png}}
\caption{ Simulated transmission (a) magnitude and (b) phase of the
proposed 2-bit RIS element.}
\label{fig4}
\vspace*{-1em}
\end{figure}
We simulate the proposed 2-bit RIS element with a periodic boundary condition
to mimic an infinite array environment and the mutual
coupling effect is also considered. The element is simulated. The simulated
element performances
under normal incidence for the four working states are plotted in Fig.
\ref{fig4}. The
transmission magnitude responses at 26.5 GHz are summarized in Table I. It can
be seen that the average transmission insertion loss of the four states is
around 1.3 dB. Moreover, the insertion losses are maintained below 3 dB ranging
from 24.8 to 28.6 GHz, corresponding to the 3-dB transmission bandwidth of
14.2\%. The phase curves of the four states are roughly parallel with around
90° phase difference in the operating frequency band. Hence, by switching the
p-i-n diodes alternatively, the RIS element successfully achieves 2-bit phase
tuning capability.
\section{Ris Design, Fabrication, and Measurement}
In this section, a transmissive RIS prototype composed of $16\times16$ proposed
elements is designed, fabricated, and measured to experimentally validate the
performance of the proposed RIS element.
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=5in]{figure/fig5.png}}
\caption{ Photography of (a) the fabricated transmissive RIS prototype, (b)
the measurement setup in the microwave anechoic chamber.}
\label{fig5}
\vspace*{-1em}
\end{figure}
The photographs of the fabricated RIS prototype and the measurement setup in
the microwave anechoic chamber are presented in Fig. \ref{fig5}. Taking
advantage of
the PCB technology, the RIS is easy to fabricate. The RIS is space-fed by a
linearly polarized horn. The effective aperture size of the RIS prototype is
$78.4\times78.4$ $\text{mm}^2$. The feeding horn illuminates the array at an
optimized distance of 50 mm. To achieve the goal of phase reconfigurability,
each element of the array needs to be controlled independently. There are 512
biasing lines since each element has two biasing lines to apply DC voltages to
four p-i-n diodes. Two logic circuit controlling boards are attached to supply
independent DC voltages of 512 channels, and they are connected to the biasing
lines through several connectors.
\begin{figure}
\centering
% \vspace*{-0.5em}
\subfigure[]
{\includegraphics[width=3in]{figure/fig6a.png}}
\subfigure[]
{\includegraphics[width=3in]{figure/fig6b.png}}
\caption{Performance of the broadside beam: (a) measured gain bandwidth,
(b) calculated and measured normalized radiation patterns at 27.0 GHz.}
\label{fig6}
% \vspace*{-1em}
\end{figure}
We measure the fabricated prototype in a planar near-field anechoic chamber.
From the measured performance of the broadside beam depicted in Fig.
\ref{fig6}, a peak
gain of 22.0 dBi is obtained at 27.0 GHz, which is 9.3 dB higher than that of
the feeding horn. The corresponding aperture efficiency is 25.3\%. The slight
frequency shift to higher frequencies is mainly attributed to the tolerance in
fabrication and the supporting structure. The measured radiation pattern is in
good agreement with the theoretical prediction based on the array theory. The
measured sidelobe level is below -22 dB, remarkably lower than that of the
1-bit designs, which is mainly attributed to the finer phase resolution of the
2-bit RIS. The measured cross-polarization level is -21.3 dB.
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=5in]{figure/fig7.png}}
\caption{Measured radiation patterns of scanned beams at 27.0 GHz in (a)
E-plane, (b) H-plane.}
\label{fig7}
% \vspace*{-1em}
\end{figure}
The radiation behaviors of the scanned beams are measured as well to
experimentally verify the dynamic beam scanning capability of the transmissive
RIS in two-dimensional space. By controlling the phase of each RIS element, the
transmitted focused beam can be formed by the RIS in the given direction. As
shown in Fig. \ref{fig7}, the measured scanned beams in the angular region from
0° to
-60° show accurate beam pointing and well-defined patterns in both E-plane and
H-plane. Note that the beams in the opposite angular region are omitted due to
the symmetry. The measured 60° scan gain losses are 4.0 dB and 5.0 dB in two
principal planes, respectively. Table II compares the performance of the
proposed 2-bit RIS with other RIS designs published in literature. It is
evident that this work achieves the highest aperture efficiency.
\begin{table}[]
\centering
\caption{Performance Comparison with Other transmissive RIS Designs}
{
\begin{tabular}{cccccc}
\toprule
\textbf{Ref.} & \begin{tabular}[c]{@{}c@{}}\textbf{Freq.}\\
\textbf{{[}GHz{]}}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Tunable}\\
\textbf{device}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Phase}\\
\textbf{resolution}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Gain}\\
\textbf{{[}dBi{]}}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Aper. eff.}\\
\textbf{{[}\%{]}}\end{tabular} \\
\midrule
{[}18{]} &
9.8 &
p-i-n &
1-bit &
22.7 &
15.4
\\
\midrule
{[}26{]} &
34.8 &
MEMS &
2-bit &
9.2 &
6.2
\\
\midrule
{[}27{]} &
29 &
p-i-n &
2-bit &
19.8 &
15.9
\\
\midrule
Our work &
27 &
p-i-n &
2-bit &
22.0 &
25.3
\\\bottomrule
\end{tabular}}
\end{table}
\section{Transmissive RIS-Aided Wireless Communication Prototype}
In this section, a transmissive RIS-aided wireless communication prototype is
set up to further measure the performance of the designed transmissive RIS in
practical communication scenarios.
\subsection{Measurement Setup}
\begin{figure}
\centering
% \vspace*{-0.5em}
{\includegraphics[width=3.5in]{figure/fig8.png}}
\caption{Transmissive RIS-aided wireless communication prototype.}
\label{fig8}
% \vspace*{-1em}
\end{figure}
As illustrated in Fig. \ref{fig8}, we carry out the experiment in an office
environment. The prototype is composed of the transmitter side, the RIS, and
the
receiver side. Both the receiver and the transmitter utilize the
software-defined
radio
(SDR) platform, composed of the PXI hardware architecture and the LabVIEW
system design software, to realize transmission at the mmWave band.
% Please add the following required packages to your document preamble:
% \usepackage{multirow}
\begin{table}[]
\caption{System Parameters}
\renewcommand\arraystretch{1.25}
\centering
\begin{tabular}{|c|c|cc|c|}\hline
\specialrule{0em}{1.2pt}{0pt}
\hline
\textbf{Parameters}
& \textbf{Value} &
\multicolumn{2}{c|}{\textbf{Parameters}}
&
\textbf{Value} \\ \hline
\begin{tabular}[c]{@{}c@{}}Carrier\\
frequency\end{tabular} & 27.0 GHz &
\multicolumn{1}{c|}{\multirow{2}{*}{\begin{tabular}[c]{@{}c@{}}Antenna
\\ gain\end{tabular}}} &
Tx &
22.7 \\ \cline{1-2} \cline{4-5}
System
bandwidth &
800 MHz &
\multicolumn{1}{c|}{}
&
Rx &
9.2 \\ \hline
Waveform
& CP-OFDM &
\multicolumn{1}{c|}{\multirow{4}{*}{\begin{tabular}[c]{@{}c@{}}Trans.\\
\\ RIS\end{tabular}}} & Array
gain & 19.8
\\ \cline{1-2} \cline{4-5}
FFT
size
&
2048 &
\multicolumn{1}{c|}{}
&
Frequency &
22.0 \\ \cline{1-2} \cline{4-5}
\begin{tabular}[c]{@{}c@{}}Carrier\\
spacing\end{tabular} & 75 KHz &
\multicolumn{1}{c|}{}
& \begin{tabular}[c]{@{}c@{}}Phase\\
resolution\end{tabular} & 2 bit \\ \cline{1-2} \cline{4-5}
\multirow{2}{*}{\begin{tabular}[c]{@{}c@{}}ADC/DAC\\
resolution\end{tabular}} & \multirow{2}{*}{14 bit} &
\multicolumn{1}{c|}{}
& Array
size & 16 $\times$ 16
\\ \cline{3-5}
& & \multicolumn{2}{c|}{Tx/Rx
height}
&
1.5 m \\ \hline
Modulation
& 16-QAM &
\multicolumn{2}{c|}{$d_1$}
&
2.4 m \\ \hline
Coding
& Turbo &
\multicolumn{2}{c|}{$d_2$}
&
0.2 m \\ \hline
Code
rate
&
1/2 &
\multicolumn{2}{c|}{$d_3$}
&
0.05 m \\ \hline \specialrule{0em}{1.2pt}{0pt}
\hline
\end{tabular}
\end{table}
The transmitter consists of the transmitter host, the FPGA, the ADC module,
the mmWave upconverter, and a horn antenna. The transmitter host controls the
system
parameters, including modulations, the transmit power, the central frequency,
and so on. A high-speed bitstream is first delivered by the host to
the FPGA module. The FPGA module realizes complex signal processing, including
channel coding, constellation mapping, OFDM modulation, and etc. After that,
the signals are sequentially processed by the ADC and the mmWave upconverter to
form mmWave analog signals suitable for transmission in wireless channels.
The transmissive RIS has 256 2-bit elements. It receives the signals from the
transmitter, tunes the phase of received signals, and then directionally
transmits them towards the receiver. Moreover, for
validating that a transmissive RIS is able to overcome obstacles, we can also
place an obstacle, such as a piece of marble, between the RIS and the base
station.
Similar to the transmitter, the receiver is composed of the receiver host, the
FPGA, a DAC module, a mmWave down converter, and a horn antenna.
At the receiver side, the Rx horn antenna first receives the signal and
transforms
them the digital domain by the mmWave down converter and the DAC module. Then,
the
FPGA module is responsible for recovering the bitstream by channel estimation,
OFDM demodulation, decoding, CRC check and etc. Moreover, according to the
results of the CRC check, the host obtains the block error rate and calculates
the data rate. Finally, the received constellation and the data rate are
illustrated on the screen of the host.
In Table III, the detailed settings of our mmWave platform are provided. We
consider a downlink communication scenario. The bandwidth is 800 MHz
and the carrier frequency is 27.0 GHz. The horn antenna of the transmitter, the
transmissive RIS, and the
horn antenna of the receiver are on the same height of 1.5 m and are all
horizontally
polarized. A piece of marble with thickness of 0.03 m is employed to mimic the
obstacle. Let $d_1$, $d_2$, and $d_3$ denote the the
distance between the Tx antenna and the obstacle, the distance between the
obstacle and the RIS, the distance between the RIS and the Rx antenna. They are
set as $d_1$ = 2.4 m, $d_2$ = 0.2 m, and $d_3$ = 0.05 m, respectively.
The fabricated
transmissive RIS works at 27.0 GHz with an additional 9.3 dB array gain when
being illuminated by the Rx horn antenna. The azimuth angle and the elevation
angle from the RIS center to the Tx/Rx are all zero, i.e., $\theta_t
= \theta_r = \phi_t = \phi_r$ = 0° as shown in Fig. \ref{fig1}. We verify the
performance of this RIS in terms of array gain and data rate.
\subsection{Array Gain Performance Assessment}
\begin{table}[]\centering
\caption{Array Gain Performance}
\begin{tabular}{ccc}
\toprule
\textbf{With or without RIS} & \multicolumn{1}{l}{\textbf{Transmit
power}} &
\multicolumn{1}{l}{\textbf{Data rate}} \\
\midrule
$\times$ & 13.6 dBm & 1024
Mbps \\
\midrule
$\checkmark$ & 5.4 dBm & 1121
Mbps \\
\bottomrule
\end{tabular}
\end{table}
To evaluate the array gain, we remove the obstacle and assess the system
performance with or without the transmissive RIS. The experiment results are
presented in Table IV. Specifically, without the transmissive RIS, when the
transmit power reaches 13.6 dBm, the data rate can reach 1024 Mbps. By
contrast, to achieve a similar transmission rate of 1121 Mbps, only 5.4 dBm
transmit power is required with the help of transmissive RIS. It is obvious
that the
transmit power is reduced by 8.2 dB. As we measured in the anechoic chamber,
the RIS can provide a 9.3 dB array gain, which is consistent with the transmit
power reduction measured by our prototype. Therefore, it is obtained that the
transmissive RIS-aided communication prototype can reduce the transmit
power by the magnitude close to the array gain provided by the RIS, while
maintaining the similar communication performance.
\subsection{Data Rate Assessment}
\begin{table}[]\centering
\caption{Data Rate Performance}
\begin{tabular}{ccccc}
\toprule
\begin{tabular}[c]{@{}c@{}}\textbf{With or} \\ \textbf{without
RIS}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{With or} \\ \textbf{without
obstacle}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Azimuth}\\
\textbf{angle}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Data} \\ \textbf{rate}\end{tabular}
\\
\midrule
$\times$ &
$\times$ &
$0^{\circ}$
& 1024
Mbps \\
\midrule
$\checkmark$ &
$\times$ &
$0^{\circ}$
& 1683
Mbps \\
\midrule
$\times$ &
$\checkmark$ &
$0^{\circ}$
& 0
Mbps \\
\midrule
$\checkmark$ &
$\checkmark$ &
$0^{\circ}$
& 1683
Mbps \\
\bottomrule
\end{tabular}
\end{table}
To further validate that the fabricated transmissive RIS is able to overcome
obstacles, we also introduce a piece of marble into the system to evaluate the
data rate performance. Specifically, depending on whether the obstacle and/or
the RIS exist, there are in total four communication scenarios. The transmit
power is fixed to 13.6 dBm. The experiment results are tabulated in Table V.
Two conclusions can be derived accordingly. One is that the fabricated RIS is
able to improve the transmission data rate. When the obstacle is removed, the
introduction of the transmissive RIS increases the data rate from 1024 Mbps to
1683 Mbps. It is worth mentioning that 1683 Mbps is the highest data rate that
the communication prototype can support with 16-QAM modulation. The other is
that the transmissive RIS is able to overcome obstacles because of the
additional array gain it provides. Specifically, when the direct Tx-Rx link is
blocked by the marble, data transmission is completely interrupted
without the help of transmissive RIS. By contrast, the introduction of the
transmissive RIS
re-establishes the communication link and achieves the highest 1683 Mbps. This
is because that the array gain provided by the RIS is higher than the
attenuation of the obstacle.
\subsection{Beam Steering Assessment}
\begin{table}[]\centering
\caption{Beam Steering Evaluation}
\begin{tabular}{ccccc}
\toprule
\begin{tabular}[c]{@{}c@{}}\textbf{With or} \\ \textbf{without
RIS}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{With or} \\ \textbf{without
obstacle}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Azimuth}\\
\textbf{angle}\end{tabular} &
\begin{tabular}[c]{@{}c@{}}\textbf{Data} \\ \textbf{rate}\end{tabular}
\\
\midrule
$\times$ &
$\times$ &
$30^{\circ}$
& 450
Mbps \\
\midrule
$\checkmark$ &
$\times$ &
$30^{\circ}$
& 1683
Mbps \\
\midrule
$\times$ &
$\checkmark$ &
$30^{\circ}$
& 0
Mbps \\
\midrule
$\checkmark$ &
$\checkmark$ &
$30^{\circ}$
& 1683
Mbps \\
\bottomrule
\end{tabular}
\end{table}
Finally, the beam steering ability of the fabricated transmissive
array is validated in this subsection. Compared to the communication settings
in previous experiments, the transmit power is fixed to 13.6 dBm and the
azimuth angle from the transmissive RIS to the transmitter is
$\theta_t=$ 30°. The other settings are the same.
The experiment
results are shown in Table VI. Without the assistance of transmissive RIS, the
data rate is
reduced from 1024 Mbps to 450 Mbps due to the misalignment between the Tx and
Rx horns. By introducing the transmissive RIS and scanning the beam it
forms to
the 30° direction, the maximum 1683 Mbps data rate is achieved. As a result, we
can conclude that the transmissive RIS is able to dynamically steer its beam to
adapt to the variation of the incident angle of the signal. Similar to the
experiment in the previous subsection, the obstacle is again introduced into
the communication prototype. The experiment results validate that the
transmissive RIS is able to improve the data rate and overcome obstacles at an
oblique incident angle $\theta_t = $ 30°.
\section{Conclusion}
The transmissive RIS has a promising prospect in 6G communications to provide
enhanced signal coverage in various communication scenarios. Compared to the
reflective RIS, the research and experimental demonstration of the transmissive
RIS and the transmissive RIS-aided communication system are considerably fewer.
This work proposes a novel 2-bit transmissive RIS design with a 90° digital
phase shifter and a 1-bit vertical current reversible dipole. The 2-bit phase
controlling capability is achieved and measured. Experimental results show the
transmissive RIS prototype with $16\times16$ elements can provide a broadside
gain of
22.0 dBi at 27.0 GHz. Moreover, a transmissive RIS-aided communication
prototype is set up, and the system performances for three representative
scenarios are measured. It is experimentally demonstrated that the transmissive
RIS is able to achieve higher data rate, reduce the transmit power, improve the
transmission capability through obstacles, and dynamically adapt to the signal
propagation direction.
\bibliographystyle{IEEEtran}
\bibliography{IEEEabrv,refs}
\end{document}
% Generated by IEEEtran.bst, version: 1.14 (2015/08/26)
\begin{thebibliography}{10}
\providecommand{\url}[1]{#1}
\csname url@samestyle\endcsname
\providecommand{\newblock}{\relax}
\providecommand{\bibinfo}[2]{#2}
\providecommand{\BIBentrySTDinterwordspacing}{\spaceskip=0pt\relax}
\providecommand{\BIBentryALTinterwordstretchfactor}{4}
\providecommand{\BIBentryALTinterwordspacing}{\spaceskip=\fontdimen2\font plus
\BIBentryALTinterwordstretchfactor\fontdimen3\font minus
\fontdimen4\font\relax}
\providecommand{\BIBforeignlanguage}[2]{{%
\expandafter\ifx\csname l@#1\endcsname\relax
\typeout{** WARNING: IEEEtran.bst: No hyphenation pattern has been}%
\typeout{** loaded for the language `#1'. Using the pattern for}%
\typeout{** the default language instead.}%
\else
\language=\csname l@#1\endcsname
\fi
#2}}
\providecommand{\BIBdecl}{\relax}
\BIBdecl
\bibitem{1}
J.~Tang, S.~Xu, and F.~Yang, ``Design of a {2.5-D} 2-bit reconfigurable
transmitarray element for {5G mmWave} applications,'' in \emph{Proc. 2020
IEEE Int. Symp. Antennas and Propag.}, 2020, pp. 631--632.
\bibitem{2}
F.~E. Idachaba, ``{5G} networks: Open network architecture and densification
strategies for beyond 1000x network capacity increase,'' in \emph{proc.
Future Technol. Conf.}, 2016, pp. 1265--1269.
\bibitem{3}
T.~S. {Rappaport}, Y.~{Xing}, O.~{Kanhere}, S.~{Ju}, A.~{Madanayake},
S.~{Mandal}, A.~{Alkhateeb}, and G.~C. {Trichopoulos}, ``Wireless
communications and applications above 100 {GHz}: Opportunities and challenges
for 6{G} and beyond,'' \emph{IEEE Access}, vol.~7, pp. 78\,729--78\,757, Jun.
2019.
\bibitem{4}
W.~{Tang}, M.~Z. {Chen}, X.~{Chen}, J.~Y. {Dai}, Y.~{Han}, M.~{Di Renzo},
Y.~{Zeng}, S.~{Jin}, Q.~{Cheng}, and T.~J. {Cui}, ``Wireless communications
with reconfigurable intelligent surface: Path loss modeling and experimental
measurement,'' \emph{IEEE Trans. Wireless Commun.}, vol.~20, no.~1, pp.
421--439, Jan. 2021.
\bibitem{5}
P.~Wang, J.~Fang, X.~Yuan, Z.~Chen, and H.~Li, ``Intelligent reflecting
surface-assisted millimeter wave communications: Joint active and passive
precoding design,'' \emph{IEEE Trans. Veh. Tech.}, vol.~69, no.~12, pp.
14\,960--14\,973, Dec. 2020.
\bibitem{6}
C.~Huang, A.~Zappone, G.~C. Alexandropoulos, M.~Debbah, and C.~Yuen,
``Reconfigurable intelligent surfaces for energy efficiency in wireless
communication,'' \emph{IEEE Trans. Wireless Commun.}, vol.~18, no.~8, pp.
4157--4170, Aug. 2019.
\bibitem{7}
L.~Dai, B.~Wang, M.~Wang, X.~Yang, J.~Tan, S.~Bi, S.~Xu, F.~Yang, Z.~Chen,
M.~D. Renzo, C.-B. Chae, and L.~Hanzo, ``Reconfigurable intelligent
surface-based wireless communications: Antenna design, prototyping, and
experimental results,'' \emph{IEEE Access}, vol.~8, pp. 45\,913--45\,923,
2020.
\bibitem{8}
Z.~Zhang and L.~Dai, ``A joint precoding framework for wideband reconfigurable
intelligent surface-aided cell-free network,'' \emph{IEEE Trans. Signal
Process.}, vol.~69, pp. 4085--4101, Jun. 2021.
\bibitem{9}
J.~Xu, Y.~Liu, X.~Mu, and O.~A. Dobre, ``{STAR-RISs}: Simultaneous transmitting
and reflecting reconfigurable intelligent surfaces,'' \emph{IEEE Commun.
Lett.}, vol.~25, no.~9, pp. 3134--3138, Sep. 2021.
\bibitem{10}
\BIBentryALTinterwordspacing
``{DOCOMO} conducts world’s first successful trial of transparent dynamic
metasurface,'' 2020. [Online]. Available:
\url{www.nttdocomo.co.jp/english/info/media center/pr/2020/0117 00.html}
\BIBentrySTDinterwordspacing
\bibitem{11}
M.~Sazegar, Y.~Zheng, C.~Kohler, H.~Maune, M.~Nikfalazar, J.~R. Binder, and
R.~Jakoby, ``Beam steering transmitarray using tunable frequency selective
surface with integrated ferroelectric varactors,'' \emph{IEEE Trans. Antennas
Propag.}, vol.~60, no.~12, Dec. 2012.
\bibitem{12}
L.~Boccia, I.~Russo, G.~Amendola, and G.~Di~Massa, ``Multilayer antenna-filter
antenna for beam-steering transmit-array applications,'' \emph{IEEE Trans.
Microw. Theory Techn.}, vol.~60, no.~7, pp. 2287--2300, Jul. 2012.
\bibitem{13}
J.~Y. Lau and S.~V. Hum, ``Analysis and characterization of a multipole
reconfigurable transmitarray element,'' \emph{IEEE Trans. Antennas Propag.},
vol.~59, no.~1, pp. 70--79, 2011.
\bibitem{15}
W.~Pan, C.~Huang, X.~Ma, and X.~Luo, ``An amplifying tunable transmitarray
element,'' \emph{IEEE Antennas Wireless Propag. Lett.}, vol.~13, pp.
702--705, 2014.
\bibitem{14}
J.~Y. Lau and S.~V. Hum, ``Reconfigurable transmitarray design approaches for
beamforming applications,'' \emph{IEEE Trans. Antennas Propag.}, vol.~60,
no.~12, pp. 5679--5689, Dec. 2012.
\bibitem{16}
C.~Huang, W.~Pan, X.~Ma, B.~Zhao, J.~Cui, and X.~Luo, ``Using reconfigurable
transmitarray to achieve beam-steering and polarization manipulation
applications,'' \emph{IEEE Trans. Antennas Propag.}, vol.~63, no.~11, pp.
4801--4810, Nov. 2015.
\bibitem{17}
M.~Frank, F.~Lurz, R.~Weigel, and A.~Koelpin, ``Electronically reconfigurable 6
× 6 element transmitarray at {K-Band} based on unit cells with continuous
phase range,'' \emph{IEEE Antennas Wireless Propag. Lett.}, vol.~18, pp.
796--800, 2019.
\bibitem{18}
A.~Clemente, L.~Dussopt, R.~Sauleau, P.~Potier, and P.~Pouliguen, ``1-bit
reconfigurable unit cell based on {PIN} diodes for transmit-array
applications in {X}-band,'' \emph{IEEE Trans. Antennas Propag.}, vol.~60,
no.~5, pp. 2260--2269, May 2012.
\bibitem{20}
W.~Pan, C.~Huang, X.~Ma, B.~Jiang, and X.~Luo, ``A dual linearly polarized
transmitarray element with {1-Bit} phase resolution in {X}-band,'' \emph{IEEE
Antennas Wireless Propag. Lett.}, vol.~14, pp. 167--170, 2015.
\bibitem{21}
L.~Di~Palma, A.~Clemente, L.~Dussopt, R.~Sauleau, P.~Potier, and P.~Pouliguen,
``Circularly-polarized reconfigurable transmitarray in {Ka}-band with beam
scanning and polarization switching capabilities,'' \emph{IEEE Trans.
Antennas Propag.}, vol.~65, no.~2, pp. 529--540, Feb. 2017.
\bibitem{23}
B.~D. Nguyen and C.~Pichot, ``Unit-cell loaded with {PIN} diodes for {1-Bit}
linearly polarized reconfigurable transmitarrays,'' \emph{IEEE Antennas
Wireless Propag. Lett.}, vol.~18, pp. 98--102, 2019.
\bibitem{24}
M.~Wang, S.~Xu, F.~Yang, and M.~Li, ``A {1-Bit} bidirectional reconfigurable
transmit-reflect-array using a single-layer slot element with {PIN} diodes,''
\emph{IEEE Trans. Antennas Propag.}, vol.~67, no.~9, pp. 6205--6210, Sep.
2019.
\bibitem{22}
L.~Di~Palma, A.~Clemente, L.~Dussopt, R.~Sauleau, P.~Potier, and P.~Pouliguen,
``Experimental characterization of a circularly polarized 1 {Bit} unit cell
for beam steerable transmitarrays at {Ka}-band,'' \emph{IEEE Trans. Antennas
Propag.}, vol.~67, no.~2, pp. 1300--1305, Feb. 2019.
\bibitem{25}
Y.~Wang, S.~Xu, F.~Yang, and D.~H. Werner, ``1 {Bit} dual-linear polarized
reconfigurable transmitarray antenna using asymmetric dipole elements with
parasitic bypass dipoles,'' \emph{IEEE Trans. Antennas Propag.}, vol.~69,
no.~2, pp. 1188--1192, Feb. 2021.
\bibitem{26}
C.-W. Luo, G.~Zhao, Y.-C. Jiao, G.-T. Chen, and Y.-D. Yan, ``Wideband 1 bit
reconfigurable transmitarray antenna based on polarization rotation
element,'' \emph{IEEE Antennas Wireless Propag. Lett.}, vol.~20, pp.
798--802, 2021.
\bibitem{19}
A.~Clemente, L.~Dussopt, R.~Sauleau, P.~Potier, and P.~Pouliguen, ``Wideband
400-element electronically reconfigurable transmitarray in {X} band,''
\emph{IEEE Trans. Antennas Propag.}, vol.~61, no.~10, pp. 5017--5027, Oct.
2013.
\bibitem{27}
C.-C. Cheng, B.~Lakshminarayanan, and A.~Abbaspour-Tamijani, ``A programmable
lens-array antenna with monolithically integrated mems switches,'' \emph{IEEE
Trans. Microw. Theory Techn.}, vol.~57, no.~8, Aug. 2009.
\bibitem{28}
F.~Diaby, A.~Clemente, R.~Sauleau, K.~T. Pham, and L.~Dussopt, ``2 bit
reconfigurable unit-cell and electronically steerable transmitarray at {Ka}
-band,'' \emph{IEEE Trans. Antennas Propag.}, vol.~68, no.~6, pp. 5003--5008,
2020.
\bibitem{29}
B.~Wu, A.~Sutinjo, M.~E. Potter, and M.~Okoniewski, ``On the selection of the
number of bits to control a dynamic digital {MEMS} reflectarray,'' \emph{IEEE
Antennas Wireless Propag. Lett.}, vol.~7, pp. 183--186, 2008.
\bibitem{30}
H.~Yang, F.~Yang, S.~Xu, M.~Li, X.~Cao, J.~Gao, and Y.~Zheng, ``A study of
phase quantization effects for reconfigurable reflectarray antennas,''
\emph{IEEE Antennas Wireless Propag. Lett.}, vol.~16, pp. 302--305, 2017.
\bibitem{31}
Q.~Wu and R.~Zhang, ``Beamforming optimization for wireless network aided by
intelligent reflecting surface with discrete phase shifts,'' \emph{IEEE
Trans. Commun.}, vol.~68, no.~3, pp. 1838--1851, Mar. 2020.
\bibitem{32}
Y.~Xiao, F.~Yang, S.~Xu, M.~Li, K.~Zhu, and H.~Sun, ``Design and implementation
of a wideband 1-bit transmitarray based on a {Yagi–Vivaldi} unit cell,''
\emph{IEEE Trans. Antennas Propag.}, vol.~69, no.~7, pp. 4229--4234, Jul.
2021.
\bibitem{33}
T.~Lambard, O.~Lafond, M.~Himdi, H.~Jeuland, and S.~Bolioli, ``A novel analog
360° phase shifter design in {Ku and Ka} bands,'' in \emph{Proc. 4th Euro.
Conf. Antennas Propag. (EuCAP)}, 2010, pp. 1--4.
\bibitem{35}
S.~X. Ta, H.~Choo, and I.~Park, ``Broadband printed-dipole antenna and its
arrays for {5G} applications,'' \emph{IEEE Antennas Wireless Propag. Lett.},
vol.~16, pp. 2183--2186, 2017.
\end{thebibliography}