1
Introduction to Intelligent Surfaces

Kaitao Meng1, Qingqing Wu2, Trung Q. Duong3,4, Derrick Wing Kwan Ng5, Robert Schober6, and Rui Zhang7

1State Key Laboratory of Internet of Things for Smart City, University of Macau, Macau, China

2Department of Electronic Engineering, Shanghai Jiaotong University, Shanghai, Country

3Queen's University Belfast, United Kingdom

4Memorial University of Newfoundland, Canada

5School of Electrical Engineering and Telecommunications, University of New South Wales, Sydney, NSW, Australia

6Institute for Digital Communications, Friedrich-Alexander University of Erlangen-Nuremberg, Erlangen, Germany

7Department of Electrical and Computer Engineering, National University of Singapore, Singapore

1.1 Background

In the forthcoming era of Internet of Everything (IoE), worldwide mobile data traffic is expected to grow at an annual rate of roughly 55% between 2020 and 2030, eventually reaching 5016 exabytes by 2030 (Andrews et al. 2014). According to a recent report, by 2025, the number of connected devices will increase to more than 30 billion globally. Also, considering the rapid emergence of new wireless applications such as smart cities, intelligent transportation, and augmented/virtual reality (Nikitas et al. 2020), it is foreseen that the fifth generation (5G) may encounter capacity and performance limitations in supporting the accommodating these low-latency, high-capacity, ultra-reliable, and massive-connectivity wireless communication services. In addition to supporting these high-quality wireless communications, next-generation wireless networks are required to provide several other heterogeneous services, e.g., extremely high-accuracy sensing (Meng et al. 2023) and low-latency computing capabilities. Specifically, the representative key performance indicators advocated by the sixth generation (6G) are summarized as follows (Letaief et al. 2019; Tataria et al. 2021; Jiang et al. 2021):

• The peak data rates under ideal wireless propagation conditions are higher than terabits per second for both indoor and outdoor connections, which is 100–1000 times that of the state-of-the-art 5G;
• The energy efficiency of 6G is 10–100 times that of 5G to achieve green communications;
• Five times more the spectral efficiency of 5G is pursued by utilizing the limited frequency spectrum more efficiently;
• Connection density could be 10 times that of 5G, about to satisfy the high demand for massive connectivity in IoE and enhanced mMTC;
• Reliability is larger than % to support more enhanced ultra-reliable and low-latency communication (URLLC) compared to 5G;
• Shorter than latency is required to support numerous enhanced URLLC;
• Centimeter (cm)-level positioning accuracy in three-dimensional (3D) space is required to fulfill the harsh demands of various vertical and industrial applications, instead of requiring meter (m)-level positioning accuracy in two-dimensional space (2D) space as in 5G;

However, even with the existing technologies such as massive multiple-input-multiple-output (M-MIMO) and millimeter wave (mmWave)/terahertz (THz), the abovementioned performance requirements for IoE services may not be fully realized due to the following reasons:

• First, dense deployment of active nodes such as access points (APs), base stations (BSs), and relays can shorten the communication distance, thereby enhancing network coverage and capacity, which, however, incurs higher energy consumption and backhaul/deployment/maintenance costs.
• Second, installing substantially more antennas at APs/BSs/relays to take advantage of the huge M-MIMO gains inevitably results in increased hardware/energy costs and signal processing complexity, as well as exacerbates more severe and complicated network interference issues (Lu et al. 2014).
• Third, migrating to higher frequency bands, such as mmWave and THz frequencies, is able to harness their larger available unlicensed bandwidth (Niu et al. 2015). Yet, it requires the deployment of more active nodes and more antennas to compensate for the associated severer propagation attenuation over distance.
• Fourth, the diffraction and scattering effects of high-frequency radio are weakened, such that propagating electromagnetic waves can be easily blocked by obstacles such as urban buildings. As a result, the effective coverage radius of APs/BSs decreases while the potential number of blind spots increase. Thus, it will be difficult to ensure universal coverage and wireless services exploiting traditional cellular technologies.

Taking into account the above limitations and issues, it is highly imperative to develop disruptively new and innovative technologies to realize spectrum- and energy-efficient and cost-effective capacity growth of future wireless networks.

1.2 Concept of Intelligent Surfaces

The fundamental challenge to achieve high-throughput and ultra-reliable wireless communication arises from time-varying wireless channels caused by user mobility (Neely 2006). The conventional approaches to this challenge are mainly to utilize various modulation, coding, and diversity techniques to counteract for channel fading, or to adapt to the channel through adaptive power/rate control and beamforming techniques (Chen and Laneman 2006). Traditionally, the fading wireless channel is treated as uncontrollable block-box and becomes one of the main limiting factors for performance improvement.

Motivated by the above, the advancing radio environment reconfiguration technique has recently emerged as a promising new paradigm to achieve smart and highly controllable ratio propagation channels for next-generation wireless communication systems. This has been achieved by the recently proposed intelligent surfaces, such as intelligent reflecting/refracting surfaces (IRSs) (Wu and Zhang 2019; Zheng et al. 2022; Huang et al. 2022), also called reconfigurable intelligent surfaces (RISs) (ElMossallamy et al. 2020; Liu et al. 2021), or large intelligent surfaces (LISs) (Hu et al. 2018; Jung et al. 2021). Generally speaking, an intelligent surface is a planar surface comprising a large number of passive reflecting/refracting elements, each of which is able to induce a controllable amplitude and/or phase change to the incident signal independently. More specifically, intelligent surfaces can be realized by adopting metamaterial or patch-array-based technologies (Liu et al. 2021). With a dense deployment of smart surfaces in a wireless network and smartly coordinating their reflection/refraction, the signal propagation/radio between transmitters and receivers can be flexibly reconfigured to achieve the desired realizations and/or distributions. This paradigm serves as a new approach to fundamentally address the wireless channel fading impairment and interference issues, and it is possible to achieve a dramatic improvement in wireless communication capacity and reliability.

As shown in Fig. 1.1, a typical architecture of an intelligent surface consists of three layers and a smart controller. On the right-most layer, a large number of metal patches are printed on a dielectric substrate to directly interact with the incident signal. Behind this layer, a copper plate is adopted to avoid signal energy leakage. Finally, the left-most layer is the control board responsible for adjusting the reflection/refractionamplitude/phase shift of each element, triggered by a smart controller connected to the intelligent surface (Jian et al. 2022). Practically, a field-programmable gate array (FPGA) can be implemented as a controller, which also acts as a gateway, communicating and coordinating with other network components (such as BS, AP, and user terminals) through an out-of-band wireless link to achieve low power consumption rate information is exchanged with them.

Figure 1.1 Architecture of intelligent surfaces.

Source: Adapted from (Wu et al. 2021).

In the following, we further highlight the main differences between intelligent surfaces and other related technologies such as active relaying and backscatter communication. First, compared to active wireless relays that assist source–destination communication through signal regeneration and retransmission, intelligent surfaces neither demodulate nor generate any information sources but act as passive array (or weakly active array, such as active intelligent surfaces; Long et al. 2021) to reflect/refract the received signals. Additionally, active relays generally operates with a half-duplex (HD) protocol and are therefore generally less spectral efficient than that of intelligent surfaces operating in full-duplex (FD) mode. Second, a tag reader in backscatter communication requires to perform interference cancellation at its receiver to decode the radio frequency identification tags' message (Wang et al. 2016). Actually, intelligent surfaces can also modulate its information during reflection, but which are primarily designed to facilitate existing communication links.

1.3 Advantages of Intelligence Surface

Benefiting from the appealing ability to reconfigure wireless channels, it is envisioned that intelligent surfaces...