Reconfigurable Holographic Surface: A New Paradigm to Implement Holographic Radio

An RHS is a special leaky-wave antenna comprising feeds and metamaterial radiation elements. As shown in Fig. 1, The feeds are attached to the edge or the back of the RHS and transform input signals into electromagnetic waves, which are also called reference waves. The RHS adopts series feeding where the incident reference wave propagates along the RHS elements and excites the RHS elements one by one. The reference wave is then transformed into a leaky wave through the slots of RHS elements to emit signals to free space [10]. The radiation amplitude of the leaky wave at each RHS element can be controlled electrically to achieve holographic beamforming. Compared with parallel feeding adopted by phased arrays where each antenna element requires a long feeding line, the series feeding leads to a much simpler wiring layout in the implementation of ultra-massive MIMO.

To further interpret the working principle of an RHS, we assume that the RHS has $K$ feeds and $N$ RHS elements. The object wave propagating in the direction $(\theta,\phi)$ is denoted as $\Psi_{obj}$, while the reference wave generated by feed $k$ is denoted as $\Psi_{ref}$. The RHS can construct a holographic pattern on the metasurface to record the interferogram $\Psi_{intf}$ between the reference wave and the object wave based on the holographic principle, i.e., $\Psi_{intf}=\Psi_{obj}\Psi_{ref}^{\ast}$ [11]. When the interferogram $\Psi_{intf}$ recorded on the RHS is excited by the propagating reference wave, the leaked wave $\Psi_{intf}\Psi_{ref}$ is proportional to $\Psi_{obj}|\Psi_{ref}|^{2}$, indicating that its direction is exactly $(\theta,\phi)$.

Remark 1. The physical structure and operating mechanism of RHS are different from those of another representative metasurface called RIS. Specifically, an RHS’s feeds are attached to the edge of the metasurface. Hence, RHSs can be integrated with transceivers conveniently. In contrast, the feeds of an RIS are set outside the metasurface owning to the reflection characteristic [5]. An extra link is required between the transmitter and the RIS to control the phase shifts of each RIS element. In addition, the RHS utilizes the method of series feeding, while the RIS utilizes the method of parallel feeding where all RIS elements are excited by the incident signals at the same time. Due to their different physical structure and operating mechanism, RHSs are more likely to serve as transmit/receive antennas directly, while RISs are widely used as relays.

The RHS utilizes an amplitude-controlled method to represent the information contained in the interferogram $\Psi_{intf}$ by tuning each RHS element based on the phase of the reference wave at each RHS element. Specifically, the phase of the reference wave changes in the process of propagation. At each RHS element, the phase of the reference wave is determined by the product of the propagation vector on the RHS and the distance vector from feed $k$ to the $n$-th RHS element $\mathbf{r}_{n}^{k}$ [11], which is a-priori fixed. If the reference wave is in phase with the object wave at an RHS element, the RHS element will be tuned to radiate much energy of the reference wave into free space. Otherwise, the RHS element will be detuned and not radiate energy into free space.

To map the phase difference between the object wave and the reference wave to the radiation amplitude of the RHS elements, the real part of the interference $\text{Re}[\Psi_{intf}]$ (i.e., the cosine value of the phase difference) is considered. Since the value of $\text{Re}[\Psi_{intf}]$ decreases as the phase difference grows, satisfying the amplitude control requirements, $\text{Re}[\Psi_{intf}]$ is a direct representation of the radiation amplitude distribution on the RHS. Hence, the holographic pattern $\bm{m}$, i.e., the radiation amplitude of each RHS element which can generate the object wave with the direction of $(\theta,\phi)$ is parameterized mathematically by

$$m(\mathbf{r}_{n}^{k},\theta,\phi)=\frac{\text{Re}[\Psi_{intf}(\mathbf{r}_{n}^{k},\theta,\phi)]+1}{2},$$

(1)

where $\text{Re}[\Psi_{intf}]$ is normalized to [0,1] to avoid negative values. This formula represents the basic principle for amplitude-controlled holographic beamforming. The effectiveness of this formula to achieve holographic beamforming will also be verified in Section V.

As shown in Fig. 2, the kernel of the designed RHS element with controllable radiation amplitude is a complementary electric-LC (cELC) resonator connected with PIN diodes¹¹1The RHS element with controllable radiation amplitudes can also be fabricated by loading varactor diodes and liquid crystals [6]. [7]. For symmetry, two PIN diodes are connected across the gaps separating the central metal patch from the micro-strip line. By controlling the biased voltages applied to the PIN diodes, the cELC resonator’s mutual inductance together with the radiation amplitude of the RHS element can be changed. Specifically, when two PIN diodes are in the OFF states, the RHS element radiates the energy of the reference wave into free space. When the PIN diodes are in ON states, the reference wave’s energy is hardly radiated²²2The design requirement of the radiation efficiency of an RHS element is related to the size of the RHS to guarantee that most of the input energy from the feed can be radiated into free space. For example, for a one-dimensional RHS with 16 RHS elements, the radiation efficiency of an RHS element with PIN diodes in the OFF and ON states is required to be 30%-40% and lower than 15%, respectively [7]..

The size of the RHS element is $0.82\times 1.7\times 0.11$ $\text{cm}^{3}$. The aimed working frequency is set as 12 GHz for satellite communications. To cover the working frequency, we choose MACOM MADP-000907-14020 with low insertion loss for PIN diodes [12]. The whole RHS element is composed of five layers given as below.

• •

  The top layer is the micro-strip line etched with the cELC.

• •

  The second layer is the radio frequency (RF) substrate, which is 0.8-mm-thick F4B, for the propagation of the reference wave.

• •

  The third layer is the ground plane.

• •

  The fourth layer is the spacer dielectric layer, which is 0.2-mm-thick F4B.

• •

  The bottom layer is the direct-current (DC) bias feed line which can apply biased voltages to the PIN diodes through a via extending through the cELC, the RF substrate, the ground plane, and the spacer dielectric layer.

Moreover, the radiation characteristics of an RHS element such as its radiation efficiency and resonant frequency are determined by its geometric parameters. Considering that the geometric parameters of the RHS element are coupled with each other and cannot be adjusted independently through a single dimension, a parameter optimization procedure based on full-wave analyses is also applied to achieve the desired radiation characteristics.

Based on the simulation of the RHS element, we conduct full-wave analyses of RHSs with different sizes utilizing the CST Microwave Studio.

As shown in Fig. 4, the designed one-dimensional RHS prototype consists of 16 RHS elements and the size of the RHS is $15.2\times 1.7\times 0.11$ $\text{cm}^{3}$. The radiation amplitude of each RHS element can be controlled based on a field-programmable gate array (FPGA). Specifically, the bias voltage applied to each PIN diode can be changed by controlling the FPGA, and thus, the ON/OFF state of the PIN diode can be controlled. In addition, we utilize SubMiniature version A (SMA) connectors attached to the edge of the RHS to feed signals into the RHS and introduce a trapezoidal evolutionary structure to the edge of the RHS for impedance matching.

As shown in Fig. 4, a vector network analyzer (VNA) is utilized to measure the beam pattern of the RHS. Specifically, port 1 of the VNA is connected to the RHS, while port 2 of the VNA is connected to a standard horn antenna. The RHS is placed on an antenna rotating platform. To measure the transmit beam pattern of the RHS, the relative signal strength (i.e., $S_{21}$ parameter) received at the horn antenna corresponding to different angles is measured and normalized. To validate the transceiver reciprocity of the RHS, the receive beam pattern is obtained by measuring and normalizing the value of $|S_{12}|$ corresponding to different angles.

The hardware modules of the RHS-aided point-to-point communication platform are shown in Fig. 5 and their functions are introduced below.

• •

  Transmitter: The transmitter (Tx) is implemented by utilizing a universal software radio peripheral (USRP) LW-N321, which can realize RF modulation/demodulation and baseband signal processing. The output port of the USRP is connected to a frequency converter to up convert the signal frequency to 12 GHz. The frequency converter then sends the up-converted signal to the RHS.

• •

  Receiver: The receiver (Rx) is also a USRP LW-N321, whose input port is connected to a frequency converter to down convert the received signal frequency. The frequency converter is connected to a standard horn antenna, which receives the signal emitted from the RHS.

• •

  Host computer: The host computer can control the Tx and Rx via a software program. The received signal is also processed by the host computer. The FPGA can also transform the input from the host computer into different bias voltages applied to the PIN diodes of each RHS element.

• •

  Ethernet switch: The ethernet switch connects the host computer, the Tx, and the Rx, and thus, the transmitted and received signals can be acquired.

To avoid environmental scatterings, the RHS prototype is deployed in a microwave anechoic chamber. To measure the far-field beam pattern of the RHS, the horn antenna acting as the receive/transmit antenna is placed 2 m away from the RHS⁶⁶6The Rayleigh distance of the RHS $d_{Ray}$ is $2D^{2}/\lambda$, where $D$ is the maximum dimension of the RHS, i.e., 15.2 cm, and $\lambda$ is the wavelength. Since the working frequency of the RHS is 12 GHz, $d_{Ray}$ is 1.8 m.. In the deployed RHS-aided point-to-point communication platform, the transmit power of the Tx USRP is 2 dBm, and the gain of the horn antenna is 20 dBi.

Fig. 6 shows the normalized far-field beam pattern of the RHS. For convenience, we set the holographic pattern of the RHS generating the beam with the direction of $0^{\circ}$ according to (1). Specifically, when the theoretical radiation amplitude of an RHS element is less than 0.5, the PIN diodes of the RHS element will be set in ON states and vice versa. It can be seen that the measured main lobe’s direction of the beam pattern matches the object direction, indicating that holographic beamforming can be achieved based on (1). We also observe that the transmit beam pattern and the receive beam pattern are almost the same, validating the transceiver reciprocity of the RHS. Moreover, the power consumption of the PIN diode in the implemented RHS is 0.01 W, which is far smaller than that of phase-shifting circuits. Hence, compared with traditional phased arrays, the RHS also provides a powerful solution to reduce power consumption in practice⁷⁷7More discussion and comparison about the performance of RHSs and traditional phased arrays can be found in [8] and [13], to which the interested readers are referred for further information..

Fig. 7 depicts the graphical interface of the Tx and Rx signals on the host computer when transmitting a real-time data stream. Specifically, the data stream generated by the vector source of the Tx USRP is a vector beginning with a 1 followed by a sequence of 0. We observe that the received signal demodulated by the Rx USRP is the same as the transmit signal. This validates that our RHS-aided point-to-point communication platform supports real-time data transmission. Moreover, from the received spectrum at the Rx, it can be seen that the signal-noise ratio (SNR) of the RHS-aided point-to-point communication is about 20 dB under the experimental environment.

Low-Earth-orbit (LEO) satellite communication networks are being developed with the promise of providing high-capacity backhaul or data relay services for terrestrial networks. However, the high mobility of LEO satellites and the severe path loss put stringent requirements on antenna technologies in terms of accurate beam steering and high antenna gain. Traditional antennas integrated with UTs such as dish antennas and phased arrays either require heavy mechanics or costly phase shifters, making their implementation in practical systems prohibitive.

Since RHSs are ultra-thin and lightweight antennas and can achieve beamforming with low hardware cost and low power consumption, RHS-enabled holographic radio can be utilized in integrated terrestrial-satellite networks to overcome the shortfalls in traditional antennas. Since existing algorithms optimizing traditional complex-valued analog beamformer do not work well for real-valued holographic beamformer, a holographic beamforming scheme generating multiple directional beams toward the satellites needs to be developed for sum rate maximization. In addition, considering that the high mobility of LEO satellites leads to time-varying beam patterns, the RHS-aided multi-satellite communication protocol design is also worth exploring.

ISAC, where the radar and communication systems share the common spectrum, is one of the most promising candidates to mitigate the spectrum congestion issue [15]. However, the high power consumption and hardware cost of traditional phased arrays restrict the implementation of ultra-massive MIMO, leading to an insufficient ISAC performance.

Benefitting from the capability of achieving high directive gain with low hardware cost, RHS-enabled holographic radio can be applied in ISAC systems to enhance sensing and communication performance. In ISAC systems, the amplitude-controlled holographic beamforming design and the estimation method for target parameters are coupled with each other, which should be simultaneously optimized in the sensing scheme. Moreover, to effectively detect targets and serve communication users, a holographic beamforming optimization algorithm needs to be carefully designed while considering the trade-off between the sensing and communication functionalities.

Wireless SLAM, which estimates the position of the user and builds up the map of an unknown environment is a promising technique to empower location-based services. It utilizes antennas to estimate the time of arrival and the angle of arrival based on the amplitudes of received multipath components (MPCs). Hence, the accuracy of the SLAM system is determined by the directive gain of the antennas.

Considering that the RHS is composed of numerous RHS elements, leading to a superior beam-steering capability and high directive gain, the RHS can be utilized in wireless SLAM systems. By adjusting the amplitude of each RHS element properly, the amplitudes of MPCs can be enhanced, such that the accuracy of the wireless SLAM system can be improved.