A Review of Indoor Optical Wireless Communication

Abstract

Indoor Optical Wireless Communication (OWC) provides a promising solution for high-capacity, low-latency, and electromagnetic interference-resistant wireless communication. Over the past decade, there has been extensive research addressing key challenges in indoor OWC. This article provides an overview of the current development status, key technologies, and challenges faced in the field of indoor OWC. Furthermore, at the end of this overview, an experimental demonstration of an indoor non-line-of-sight (NLOS) OWC system utilizing a spatial light modulator (SLM) for beam steering is demonstrated, which is expected to inspire research on related technologies.

1. Introduction

With the emergence of fifth-generation mobile networks (5G) wireless communication applications, wireless communication is facing unprecedented higher bandwidth demands [1]. Technologies such as streaming video, virtual reality (VR), and augmented reality (AR) require higher communication rates. However, the available bandwidth resources of radio frequency (RF) are limited, presenting a significant challenge in meeting the requirements of future communication services. In order to address this issue, free space optics (FSO) communication, also known as optical wireless communication (OWC), has been proposed as a complementary solution to RF wireless communication, due to its advantage of utilizing a large amount of unlicensed spectrum resources [2,3]. Furthermore, OWC has better resistance to electromagnetic interference, thus offering enhanced stability [4]. Additionally, due to its weaker penetration capability, optical beams are typically confined to a certain area, making OWC more secure and reliable [5].

In numerous application scenarios, such as indoor wireless communication, underwater communication, data center optical wireless interconnects and vehicle communication [6], research on indoor OWC is crucial, as the majority of communication loads are generated in indoor environments [7]. Depending on the wavelength of the optical source used, indoor OWC can be classified into visible-light communication (VLC) and near-infrared (NIR) communication. There has been extensive research on VLC in the past, aiming to improve the spectral efficiency of VLC. In recent years, research on NIR communication has also emerged significantly. Due to the different characteristics of the optical sources used, the communication applications using the two indoor OWC systems are distinct. Indoor VLC, typically combined with illumination, utilizes light sources with divergent radiation characteristics and has a wider signal coverage range. In contrast, indoor NIR communication employs light beams with reduced divergence, in which successful communication relies on the high directionality of the light beam [8]. Therefore, the two communication systems have different implementation schemes, as well as distinct key technologies and challenges.

In this paper, we review the development status of indoor VLC and indoor NIR communication, aiming at innovations and improvements in technology, as well as highlighting the technical challenges. Subsequently, we provide a detailed overview of some key technologies in the two indoor OWC systems, along with a summary of research on the implementation of these key technologies. Moreover, for the implementation of indoor NIR communication, we demonstrate an NLOS-OWC system utilizing a spatial light modulator (SLM) to control the light beam and achieve successful transmission of more than 160 Gbit/s 16-QAM signal experimentally. Finally, we outline the future trends in indoor OWC research.

2. Indoor OWC Research

In the indoor OWC research field, various review articles have comprehensively introduced different aspects. In the field of VLC, Saeed Ur Rehman et al. discussed the history, standardization efforts, and applications of VLC, and outlined various technologies used in the uplink, along with their limitations [9]. Luiz Eduardo Mendes Matheus et al. provided an overview of the comprehensive research status of VLC and the key concepts and challenges related to this emerging field [10]. Pedro A. Loureiro focused on introducing the digital signal processing techniques used in VLC systems, analyzed different modulation formats, and categorized them into single-carrier and multi-carrier modulation schemes [11]. This paper mainly outlines the relevant technologies based on the characteristics of LED light sources and the current research status in the context of VLC.

VLC has already established a broader research foundation and numerous application cases. In contrast, near-infrared optical wireless communication (NIR-OWC) is still in its nascent stages as a relatively new research field. Consequently, there are relatively fewer review papers related to NIR-OWC. Wang K et al. summarized the latest developments in indoor NIR-OWC technology and systems [12]. This paper primarily outlines the key technologies for near-infrared beam control, the current development status, and the challenges faced in the context of NIR-OWC.

2.1. VLC Research

VLC utilizes light sources with wavelengths ranging from 400 nm to 700 nm and air as the transmission medium to transfer data, with communication bandwidth exceeding 320 THz. In indoor applications, VLC is often integrated with illumination equipment to enable simultaneous illumination and data transmission.

Light emitting diodes (LEDs) are the key devices driving the development of indoor VLC, as they are not only cost-effective but also can be integrated into various environments. The modulation bandwidth of LEDs can support data transmission for indoor VLC, making them a common medium for VLC applications. The output beam of an LED exhibits a divergent characteristic [13], meaning that the LED emits energy over a wide emission angle. Therefore, indoor VLC systems based on LEDs have a larger coverage area, which is advantageous for applications without the establishment of beam steering. However, due to the limited modulation bandwidth of LEDs, the visible light spectrum resources have not been fully utilized. Currently, the main approach to improving the communication rate of indoor VLC systems based on LEDs is to enhance spectral efficiency. Various modulation techniques or multiplexing technologies have been proposed, along with research on novel LED devices, to further increase the data transmission rate of indoor VLC systems.

A typical approach to enhancing spectral efficiency is wavelength division multiplexing (WDM), which involves using a combination of multi-color LEDs for illumination and assigning each color LED for different signal transmission. By individually modulating LEDs of different colors, multiple wavelengths can be used simultaneously for data transmission, and the data rate can be linearly scaled with the number of devices, enabling high-speed data transmission. White illumination can be achieved by combining three or more different colors of LEDs. In 2019, R. Bian et al. modulated light at four wavelengths in the visible light spectrum, corresponding to the colors red, green, blue, and yellow [14]. Combining orthogonal frequency division multiplexing (OFDM) with adaptive bit loading and utilizing four single-color low-cost LEDs as light sources, data transmission at a rate of 15.73 Gb/s over a 1.6 m link was achieved through WDM [14].

Color shift keying (CSK) is also a modulation format used for indoor VLC based on LEDs, where data transmission is achieved by modulating the intensity of different colored LEDs to represent digital signals. Current detectors installed in electronic devices at the receiver side can capture changes in color intensity and convert them into corresponding bit information. CSK enhances data throughput [15,16]. Recently, there has been an increasing amount of research on encoding methods under CSK modulation to achieve efficient spectrum utilization [17,18]. Additionally, for optical wireless channels with delay spread, research has shown that the four-color CSK modulation format is more efficient and reliable than the traditional three-color CSK [19].

In addition to fully utilizing available visible light resources, some studies focus on improving light sources to achieve higher transmission rates or lower implementation costs. Innovative designs of LED materials and structures aim to overcome the bandwidth bottleneck of LEDs [14]. For example, discussions have been conducted around micro-LEDs (µLEDs) with electrical-to-optical (E-O) bandwidth exceeding 1 GHz. Furthermore, organic light-emitting diodes (OLEDs) have also garnered significant attention. While the transmission rate of OLEDs is constrained by high capacitance, they offer flexible structures and lower manufacturing costs [15].

2.2. NIR Light Communication Research

NIR light communication is a communication technology widely studied in indoor OWC following VLC. Compared to VLC, NIR light communication is more compatible with optical fiber communication systems and benefits from the development of mature optical fiber communication devices and technologies. NIR light communication uses lasers as light sources. In early indoor NIR OWC research, emission wavelength lay between 850 nm and 950 nm [20]. With the maturity of optical fiber communication devices, recent research on NIR light communication has focused on the 1550 nm wavelength range [7,21]. Owing to the mature development of optical fiber communication technology, OWC systems can leverage approximately 20.9 THz of communication bandwidth within the S + C + L band wavelength range [22]. Compared to VLC, NIR light communication has fewer available bandwidth resources. However, lasers offer higher modulation bandwidth compared to LEDs, enabling NIR light communication to achieve higher transmission rates, which can reach over Tb/s [23].

Due to the collimation and directionality of lasers, the loss of the emitted beam in the free space transmission is minimal, enabling communication over longer distances. However, as NIR light cannot penetrate obstacles, it is still favored for indoor short-range applications. In contrast to VLC using LEDs as light sources, NIR light communication utilizes narrow laser beams with limited coverage. In complex indoor environments, establishing a line-of-sight (LOS) link between the transmitter and receiver may be challenging due to the presence of obstacles, while non-line-of-sight (NLOS) links typically experience higher transmission loss and cannot be directly used for communication. Increasing transmitted power to utilize diffuse links for communication is impractical, as the transmitted power must fall within the power range permitted for human eye safety (according to IEC 60825 [24] and ANSIZ136 standards [25], the maximum power for λ > 1.4 μm is 10 mW). Beam steering technology is a common method to establish NLOS NIR light communication links between the transmitter and receiver [26]. Beam steering allows the optical beam to transmit along a specified path. Furthermore, to meet the demands of high-speed wireless communication for indoor mobile users, real-time and high-speed control of the beam’s transmission direction is required. By incorporating positioning technologies, the transmission link between the transmitter and users can be reestablished, enabling coverage of all areas within the indoor environment.

Considering the high directionality of lasers, individual transmission links can be established for user terminals located at different positions to provide communication services. These different transmission links are orthogonal and do not interfere with each other. Therefore, communication services for multiple users can share the same spectrum of resources. Given the limited bandwidth resources of NIR optical communication, this multiple-access communication scheme is advantageous.

In addition to the limitation imposed by safety constraints on transmit power, NIR optical communication also faces challenges such as high equipment costs, which hinder the realization of large-scale applications. However, NIR optical communication is favored in specific application scenarios where high data rates per link are required. In recent years, indoor NIR optical communication has continuously made breakthroughs in transmission rates.

L. Chen et al. successfully transmitted a 20 Gbit/s PAM4 signal over a 1.1 m free-space link based on a waveguide grating router (WGR) with a bandwidth limit of 6 GHz [27]. T. Koonen achieved a synchronous communication speed of up to 112 Gbit/s per beam using PAM-4. With the use of an 80-port arrayed waveguide grating router (AWGR), they provided a total wireless throughput exceeding 8.9 Tbit/s [23].

Z. Li proposed a multi-user accessible indoor infrared OWC system based on passive diffractive wave-controlled antennas for point-to-multipoint indoor optical wireless communication [28]. By combining spectrum scanning technology with time division multiple access (TDMA), this system allows a single transmitter to serve 5 mobile user terminals. Each signal beam can achieve a data rate of 12.5 Gbit/s, enabling free-space optical wireless transmission over a distance of 3.6 m.

We have made a comparison of the two indoor OWC systems concisely and clearly, which is given in Table 1.

For short-range OWC, the IEEE 802.15.7 standard [29] defines physical and media access control (MAC) layer parameters, including frame structure design, the selection of modulation and coding modes, and so on. According to the standard-defined operating modes, a broad class of optical transmitters such as LEDs can achieve a maximum data rate of 96 Mbit/s, limited by the modulation bandwidth of the optical transmitter and the spectral efficiency of the selected modulation format. Furthermore, the standard defines operating modes for CSK, which can achieve higher data rates at the same optical clock rate, indicating that CSK can more fully utilize available optical spectrum resources. The IEEE 802.11 [30] series wireless standards are widely used for indoor wireless communication. In addition, it also provides definitions for light communication with wavelengths in the range of 800 nm to 1000 nm. Currently, the 802.11 suite is widely available commercially, while VLC and NIR light communication are still in the early stages of entering the market or being researched in laboratories.

Table 1. Comparison of indoor OWC system.

	VLC	NIR-OWC
Light sources	LED	LD
Operating wavelength	400–700 nm	1460–1625 nm
Bandwidth	320 THz	20.9 THz
System complexity	Low	High
Coverage area	3 m	<10 m
System cost	Low	High
Power consumption	High	Low (<10 mW)
Maximum achievable rate	46.4 Gbit/s [31]	>112 Gbit/s [23]
Infrastructure	Shared LED illumination	Fiber
Safety	Penetrate eyes; power << 1 mW	Does not penetrate eyes; power < 10 mW
Direction	low directional	high directional

Table 1. Comparison of indoor OWC system.

	VLC	NIR-OWC
Light sources	LED	LD
Operating wavelength	400–700 nm	1460–1625 nm
Bandwidth	320 THz	20.9 THz
System complexity	Low	High
Coverage area	3 m	<10 m
System cost	Low	High
Power consumption	High	Low (<10 mW)
Maximum achievable rate	46.4 Gbit/s [31]	>112 Gbit/s [23]
Infrastructure	Shared LED illumination	Fiber
Safety	Penetrate eyes; power << 1 mW	Does not penetrate eyes; power < 10 mW
Direction	low directional	high directional

3. Indoor OWC Key Technologies

Indoor VLC has abundant available bandwidth resources, but its limited modulation bandwidth restricts the full utilization of these resources. In addition to innovative materials and designs for light sources, some multiplexing or modulation techniques are key to improving the data rate of indoor VLC communication. In contrast, indoor NIR optical communication can achieve higher data rates but relies on highly directional beams. Therefore, beam control techniques are the focus of research in indoor NIR optical communication.

3.1. Key Technologies in VLC

Despite limited modulation bandwidth and low output power, considering cost, lifespan, and security issues, the LED remains the light source of choice in many indoor VLC research studies. In indoor VLC systems, the direction of the LED beam is typically not controlled, but rather the strong divergence of the LED output beam is utilized to provide a larger signal coverage area. In order to overcome the limitations on system performance imposed by the small bandwidth of LED devices, in addition to researching new LED devices, many advanced modulation formats are being applied in indoor VLC. Furthermore, considering the unique characteristics of LED light sources, some multiplexing and novel modulation techniques have been proposed to more effectively utilize the visible light spectrum resources. By employing these key technologies, the achievable transmission rates in VLC are further enhanced.

3.1.1. WDM Based on Multi-Color LEDs

Traditional illumination applications typically use blue LEDs coated with phosphors to generate white light. The modulation bandwidth of single-color LEDs is limited, making it difficult to meet the growing bandwidth demands of future high-speed indoor wireless access applications. Additionally, the existence of a nanosecond-level relative delay between excitation and emission in phosphors [14] may not be easily observed in illumination applications but has a more pronounced impact on communication.

Therefore, using multi-color LED chips for mixed emission is a more promising solution for achieving indoor VLC. WDM based on multi-color LEDs can effectively utilize the spectral resources of visible light, and separately modulating multiple LEDs can significantly increase transmission bandwidth. For example, utilizing a three-color element LED composed of red, blue, and green can constitute three individual channels [15]. Each color LED is modulated with a different signal, as shown in Figure 1. The receiving end typically consists of optical filters, optical concentrators, photodiodes (PDs), and optional trans-impedance amplifiers (TIAs). Optical filters are used to select the wavelengths of light to be detected. Additionally, optical concentrators such as lenses are necessary to maximize received optical power. The role of the PD is to convert the optical intensity signal into an electrical signal. The achievable information rate per channel in a WDM system is lower than when all LEDs transmit the same signal, as the irradiance of each color LED decreases when the total illuminance is constant, resulting in reduced received optical power and signal-to-noise ratio. However, the use of a WDM system with multi-color LEDs increases the total achievable information rate, effectively utilizing the spectral resources of LEDs. The WDM channels are expanded to five (including five different color LEDs), resulting in a total information rate of 10.72 Gb/s [16]. It should be noted that due to variations in the efficiency of different color LEDs and the optimal wavelength response of different PDs, channels of different colors typically exhibit different performance, allowing for flexible modulation formats and rate allocation for different channels [16,19]. Since different color LEDs have different spectral emission characteristics, as shown in Figure 2, and the bandwidth of the modulated electrical signals is typically much smaller than the visible light carrier frequency, the optical filters at the receiving end should be specially designed to reduce the introduction of crosstalk interference in WDM channels [17].

3.1.2. Multi-Input-Multi-Output (MIMO)

Due to the typical low output power of LEDs [33], multiple LEDs can be used for illumination in indoor environments, meaning that there can be multiple transmitters in a room. Additionally, there are typically multiple receiving devices communicating simultaneously in the room. By exploiting this characteristic, MIMO can be applied in indoor VLC to provide additional benefits. MIMO technology establishes channels between multiple transmitters and multiple receivers, effectively increasing spectrum efficiency and transmission data rates through spatial multiplexing (SMP) [34]. Furthermore, the transmission link in indoor VLC can be divided into LOS and NLOS links based on whether the transmitted signal directly reaches the receiver. LOS links typically occupy higher power but are susceptible to communication obstacles. On the other hand, scattering systems often have multiple NLOS links reaching the receiver, making them more resilient to obstacles, but NLOS links generally experience higher losses and introduce multipath fading. To address this issue, MIMO’s spatial diversity gain can be utilized to mitigate the effects of multipath fading, thereby improving channel reliability. A typical indoor MIMO-VLC system is illustrated in Figure 3.

Three commonly used MIMO schemes are repetition coding (RC), SMP, and spatial modulation (SM) [35,36,37,38,39,40]. RC is the simplest coding scheme for transmission diversity, achieved by transmitting the same signal from different transmitters. RC does not improve spectral efficiency but can provide diversity gain. Taking a system consisting of two transmitters and one receiver as an example, an indoor VLC system using RC is illustrated in Figure 4, where the transmitted signals from the two transmitters are identical. The received signal, after passing through different channels from Tx1 and Tx2 to the receiver, can be represented as

$$R_i=\left(H_1+H_2\right)\times S_i+n,$$

(1)

where $S_i$ represents the transmission signal at the ith moment, $H_1$ and $H_2$ are the channel responses, and $n$ represents the received noise. Considering the phase relationship between the received signal and the channel response, sending the same signal in multiple channels does not always constructively improve the signal-to-noise ratio. Additionally, the transmission distances for different transmission paths may vary, leading to multipath delays. The frequency power fading introduced by the relative delay of different paths cannot be compensated by RC. To combat multipath effects, Alamouti space-time block coding (STBC) can be used as an alternative, but STBC requires signal estimation to obtain channel information. As shown in Figure 4, STBC encoding involves a special design of the transmitted signals. Assuming that Tx1 sends signals $S_0$ and $S_1$ at two consecutive different moments, the transmitted signals of Tx2 at the same two moments are $-S_1^{\ast }$ and $S_0^{\ast }$, where $\ast$ denotes complex conjugate operation. The received signals at two consecutive moments can be represented as:

$$R_1=H_1{S}_0-H_2{S}_1^{\ast }+n_1,$$

(2)

$$R_2=H_1{S}_1+H_2{S}_0^{\ast }+n_2,$$

(3)

At the receiver end, the channel is estimated, and then the received signal is subjected to special decoding. The prediction of the transmitted signal can be given by

$$\widehat{S_0}=H_1^{\ast }{R}_1+H_2{R}_2^{\ast }=\left({\left|H_1\right|}^2+{\left|H_2\right|}^2\right)S_0+H_1^{\ast }{n}_1+H_2{n}_2^{\ast },$$

(4)

$$\widehat{S_1}=-H_2^{\ast }{R}_1+H_1{R}_2^{\ast }=\left({\left|H_1\right|}^2+{\left|H_2\right|}^2\right)S_1-H_2^{\ast }{n}_1+H_1{n}_2^{\ast },$$

(5)

The STBC scheme is not affected by the phase of the channel response and can mitigate the impact of power fading on system performance.

SMP involves parallel transmission of different data from all transmitters, where different signals utilize the same spectrum resources for transmission, thereby improving spectrum efficiency. In the SMP scheme, there exist multiple orthogonal channels between the transmitters and receivers. Assuming the number of transmitters and receivers are $N_t$ and $N_r$, respectively, the MIMO channel matrix $\mathit{H}$ has dimensions of $N_t\times N_r$. The SMP scheme requires obtaining responses from each channel to achieve signal demultiplexing. The received signal vector can be represented as:

$$\mathit{r}=\mathit{H}\mathit{s}+\mathit{n},$$

(6)

where $\mathit{s}$ represents the transmission signal vector, and $\mathit{n}$ represents the noise vector. Assuming that the receiver has knowledge of the channel information, different detection methods can be used to obtain estimations of the transmission signal, such as maximum likelihood (ML) estimation [36] and zero-forcing (ZF) detection [37,38]. With the ZF method, an estimation of the transmission signal can be obtained, as given by:

$$\widehat{\mathit{s}}=\mathit{W}\mathit{r},$$

(7)

where $\mathit{W}={\left({\mathit{H}}^{\mathit{T}}\mathit{H}\right)}^{-1}{\mathit{H}}^{\mathit{T}}$ represents the pseudo-inverse of the channel matrix. The SMP scheme requires that there be low coherence between different channels; otherwise, signals from different channels may cause interference.

SM is a modulation format that considers the spatial dimension, utilizing spatial information to represent additional bits, thereby improving spectral efficiency. Spatial information is reflected in the transmitter array. In the SM scheme, only one transmitter is activated at a time, with each transmitter corresponding to a dedicated bit sequence, as shown in Figure 5. The signals transmitted by the transmitters are modulated onto conventional constellation symbols, and when the receiver detects the activated transmitter, the corresponding bit is added to the transmitted sequence. Therefore, SM modulates in both the time and spatial dimensions, enhancing spectral efficiency. As only one transmitter is active at any given time, this scheme also avoids inter-channel interference (ICI). Since the channels between the receiver and transmitters typically have different channel responses, the receiver can detect the activated transmitter. SM is also affected by channel correlation, and reducing the distance between the transmitter and receiver, as well as the transmitter’s emerging angle, can decrease correlation [37].

3.1.3. CSK

In OWC systems, signals are typically represented by the intensity of light. However, in indoor VLC systems in which communication is combined with illumination, fluctuations in illumination intensity can pose risks to eye health and may even induce nausea or epilepsy [41]. Therefore, it is important to minimize variations in total illumination intensity. The IEEE 802.15.7 standard introduces a high-speed VLC specification, where the physical layer employs CSK [42]. CSK is a special modulation technique in VLC that utilizes different chromaticity coordinates in a chromaticity diagram to represent different binary sequences [43], as shown in Figure 6. CSK modulates by adjusting the output intensity of the different color LEDs in an LED to change the chromaticity value of the output light, thereby transmitting different bit information. In CSK modulation, the changes in light intensity occur at rates higher than what the human eye can perceive, thus avoiding adverse effects and providing higher data rates [44].

3.2. Beam-Steered NIR Optical Communication Technologies

NIR light communication offers a higher link power budget than visible light communication, allowing for higher data rates; however, successful communication relies on the high directionality of the beam. Free-space narrow NIR beams can serve a single-user device, avoiding congestion by eliminating sharing among devices. Controlling the beam to point it in the desired direction is crucial for indoor NIR OWC systems. In complex indoor spaces, NLOS links between the transmitter and user devices need to be established, considering that LOS links may be blocked. However, NLOS links typically experience higher losses because the light undergoes scattering when it hits objects (e.g., walls) and propagates in different directions. Therefore, schemes for establishing NLOS links based on beam control need to be investigated. Moreover, to meet the demands of mobile communications, high-speed, real-time beam control techniques need to be explored.

The concept of reconfigurable optical wireless was first proposed in 1999, utilizing a holographic material to record a scanned beam encoded with a code-division multiple access (CDMA) code, thereby generating a hologram. When incident light with the correct spatial code illuminates the hologram, the beam is steered in a specific direction. The system uses a spatial light modulator to generate and access the spatial codes. Since then, reconfigurable beams have been extensively studied. There are two main types of devices used to achieve beam control: active beam controllers and passive beam controllers.

3.2.1. Passive Beam-Steering Devices

Passive beam controllers do not require local power and separate device control channels for each beam, leading to lower system management costs. Laser tuning times are shorter, resulting in faster beam control. Passive beam control uses wavelength-dependent beam-shaping elements, such as gratings, and can handle multiple beams simultaneously.

Since diffraction gratings can spatially separate wavelengths, using a tunable high-bandwidth light source allows light with different wavelengths to be diffracted to different spatial locations by the diffraction grating, thereby achieving beam control. In indoor NIR light communication, the diffracted narrow NIR beam can be used to transmit signals. The grating equation can be given as

$$m\lambda =d\left(n_{1}sin{\theta }_i\pm n_{2}sin{\theta }_m\right),$$

(8)

where $m$ is the diffraction order, $\lambda$ is the wavelength of the beam, $d$ is the grating constant, $n_1$ and $n_2$ correspond to the refractive indices of the incident medium and the transmission or reflection medium, and ${\theta }_i$ and ${\theta }_m$ are the incident angle and diffraction angle, respectively. The diffracted light energy is mainly concentrated in the 0th order, but the 0th order does not carry information. Blazed gratings can shift the interference null spectrum and concentrate it on a certain order spectrum, so two orthogonal blazed gratings can usually be used to achieve beam control, as shown in Figure 7.

Under two-dimensional (2D) beam control, a narrow beam from a single-mode optical fiber is steered to an angle of ψ after the first grating. The other grating diffracts the beam in the orthogonal direction and operates at a lower diffraction order, with the total wavelength tuning range Δλ being smaller than its free spectral range (FSR). Therefore, when the wavelength of the beam is tuned, the beam is steered in a 2D direction sequentially. The $FSR$ can be given by

$$FSR={\displaystyle \frac{{\lambda }_m}{m+1}},$$

(9)

where ${\lambda }_m$ is the wavelength corresponding to the order m.

Based on the above theory, CW Oh et al. designed a passive beam control module based on cascaded reflection gratings, which can achieve simultaneous coverage of multiple users [45]. This beam control method does not require a local power supply and can be controlled remotely in real time by simply tuning the wavelength of the signal. With a scanning angle of 5.61° × 12.66°, free-space transmission of at least 37 Gbps per beam was achieved over a distance of up to 2 m.

In addition to free-space diffraction gratings, in-fiber diffraction gratings can also be used. Free-space diffraction gratings have limited diffraction efficiency and a bulky configuration, whereas diffraction gratings using fibers can achieve efficient, compact, and fiber-compatible laser beam control. Wang G et al. implemented in-fiber diffraction based on a 45° tilted fiber grating (TFG) and realized all-optical fiber laser beam control for full-duplex wireless optical communication among multiple users [4]. Utilizing orthogonal frequency-division multiplexing signals with a bandwidth of 2.4 GHz, full-duplex wireless transmission over 1.4 m free space was achieved, with a data transmission rate of up to 12 Gb/s per beam.

Another passive beam control approach employs an AWGR with a large number of output fiber ports [23,46,47]. Compared to the method using diffraction gratings, the AWGR-based beam control approach does not require the alignment of two gratings and does not require highly efficient and low-polarization-dependent gratings. These gratings are rearranged into a 2D fiber array. As shown in Figure 8, a lens is placed after the 2D fiber array, and the lens is used to collimate the beams emitted from the fibers. Beams with different wavelengths are emitted from different fibers in the 2D array and are steered to specific 2D angular directions, thereby achieving beam control. Koonen, T used an 80-port AWGR to achieve simultaneous communication of up to 112 Gbit/s per beam, providing a total wireless throughput of over 8.9 Tbit/s [23].

Tunable metasurfaces are widely used for beam control, but it is impractical to control each resonator, resulting in very limited beam steering. Therefore, P. C. Wu et al. proposed using passive metasurfaces to control the beam, similar to the method using gratings, applying wavelength tuning to change the beam direction [48]. Subsequently, J Huang et al. designed a polarization beam splitter based on a passive gap-surface plasmon metasurface (GSPM) to control 2D infrared beam steering and established a 20 Gbps beam-guided infrared wireless link over 1.2 m of free space [49].

3.2.2. Active Beam-Steering Devices

Although passive beam controllers do not require power, they use wavelength tuning, which requires a precisely controlled and expensive tunable laser source. In contrast, active beam controllers can change the state of the device through a separate control channel, thereby changing the direction of the beam. Common active beam controllers include MEMS-based steering mirrors [50,51,52,53], silicon-integrated optical phased arrays (OPAs) [54,55,56,57,58], and SLMs [23,59].

In MEMS-based reflective beam steering, the main actuation mechanism is shown in Figure 9. The beam from the laser is reflected by two fast steering mirrors (FSMs). The first FSM acts as a control actuator, and the second FSM is used to interfere with the beam direction, resulting in a shift in the position of the beam spot on the detector surface.

There are two sources of jitter in the system: the shaker on which the control actuator is mounted and the disturbance actuator FSM2. The shaker vibrates in the vertical direction, while the FSM2 vibrates in two directions. By measuring the horizontal and vertical displacements of the laser beam center on the position-sensitive device (PSD) plane and transmitting them to computer 1 in the form of voltage, the digital signal processor runs the feedback and adaptive controller and sends actuator commands to FSM1. Computer 2 is used to send disturbance commands to FSM2, and the coordinates of the laser beam center on the sensor also correspond to the beam deflection caused by the rotation of the mirror.

Currently, there are increasing research applications of MEMS-based steering mirrors for beam control [49,50,51]. Wang X et al. designed an adaptive controller to effectively suppress laser beam jitter, and the schematic diagram is shown in Figure 10. The beam deflection angle is adjusted by controlling the FSMs driven by piezoelectric ceramics [2]. Čierny O et al. designed a beaconless alignment and tracking method using a single MEMS steering mirror for precise beam pointing and jitter injection to infer fine tracking information [51].

The structure of a 2D multi-beam steering OPA is shown in Figure 11 [54]. Silicon-integrated OPAs can steer a beam to different directions in the far field by changing the relative phase delay between adjacent phased array elements [55,56,57]. Compared with other active control devices, OPAs can not only provide agile, low-divergence, and wide-steering-range beams [56] but also be compatible with CMOS circuits. In recent years, benefiting from the rapid development of silicon photonics integration technology, OPAs have been further studied due to their integration and mass production using advanced CMOS manufacturing equipment.

Wang K designed and fabricated a 1 × 4 integrated phased array, which achieved data transmission of up to 12.5 Gbit/s over 1.4 m of free space using silicon-integrated waveguide devices [6]. Y. Li et al. used 2D beam control of a non-uniform space OPA chip to achieve data transmission of a 32 Gbit/s NRZ signal over a distance of 54 m [59].

Among active control devices, SLMs can control beams without mechanical movement and do not require steering the beam to the target position by rotation. SLMs can be used to steer a beam in a certain direction by controlling the phase on a diffractive surface. Compared with other methods, SLMs offer higher repeatability and tolerance to environmental changes.

SLMs are generally composed of liquid crystal materials. By changing the voltage applied to the SLM, the deflection of the liquid crystal molecules can be controlled, thereby changing the information such as the phase and polarization of the incident light. SLMs do not limit the modulation bandwidth [60] and are also highly resistant to wavelength drift. Feng used a pure phase-type LCOS SLM for beam control and achieved data transmission with narrow beam manipulation. The angular coverage of the system was expanded to 3 degrees, and no moving parts were required, realizing a long-distance free-space optical wireless communication system with three-degree angular coverage [60].

Z Cao used an SLM to modulate the wavefront of a beam after it passed through a scattering medium, causing the beam to focus on the target position after passing through the scattering medium [26]. As shown in Figure 12, when a beam is incident on a diffuse reflection plane, the light will scatter in all directions, and the detector can only detect a small part of the light. In order to transmit data, the phase of the incident beam is changed using an SLM, so that the beam can be focused in a certain direction, thereby maximizing the light intensity at the detector. Based on this principle and device, Z Cao designed a reconfigurable beam shaping system and NLOS OWC system.

4. NLOS-OWC System Enabled by Beam-Steering with SLM for Multi-User Indoor Access

Reconfigurable beam systems based on SLMs play a significant role in NLOS-OWC. The complex indoor environment makes it difficult to establish an LOS link for NIR communication, hence the research on NLOS-OWC is meaningful. In general, the light impinging on the scattering object is scattered in many different directions, resulting in only a small amount of scattered light being detected at the receiver. To this end, wavefront shaping algorithms can be employed, which are used to focus the beam after passing through the scattering medium. The wavefront of the beam is shaped using an SLM, so that the light impinging on the scattering object is scattered in a specific direction, thus enhancing the light intensity in the desired direction and establishing an NLOS link. Therefore, wavefront shaping technology is of great significance for data transmission in NLOS links. This scheme can both control the direction of the beam and maximize the light intensity received by the detector at the receiving end.

Additionally, single-user wireless communication technology may not effectively address the high-density user demands, leading to issues such as network congestion and communication delays. Therefore, the study of indoor multi-user wireless communication technology is highly essential.

Inspired by the work of Cao et al., our previous research investigated an SLM-based multi-user NLOS-OWC system [61]. In this system, spatial division of input modes is utilized to enable multiple wavelengths to serve multiple users, achieving wavelength-division multiplexing. This system successfully transmitted 2 × 160 Gbit/s 16-QAM signals over a free-space distance of 0.125 m. The application scenario of this system is depicted in Figure 13. Transmitters are typically installed centrally in a room, while the NLOS multi-user communication system can be established across multiple rooms connected to transmitters via fiber optic transmission. The indoor NLOS OWC system supporting multi-user communication based on the proposed wavelength manipulation strategy has been experimentally verified.

The system setup diagram is shown in Figure 14, with two users positioned at an offset of ±10° from the main reflection angle. In this system, SLM was employed to achieve beam control, manipulating two different wavelengths of light into two different directions. This enables different users to utilize distinct wavelengths for data transmission, thereby achieving wavelength division multiplexing.

The spatial partitioning strategy employed in the system was experimentally validated to demonstrate the performance difference between using all pixels and half of the pixels for modulation, as shown in the experimental results in Figure 15. It was observed that when modulation uses all pixels, it exhibits a difference of 3.5 dB under the hard decision forward error correction (HD-FEC) threshold when compared to modulation that uses half of the pixels.

Furthermore, since each user utilizes a different wavelength, crosstalk can occur between these wavelengths. Therefore, this study also investigates the impact of various wavelength differences on communication performance. As shown in Figure 16, the larger the wavelength difference is, the larger the power difference can be observed. For the wavelength difference of 10 nm, the power difference is remarkable, which can reach 14.95 dB.

5. Future Challenges and Conclusions

Indoor OWC has gained significant research attention. However, its practical implementation faces several challenges. For indoor VLC, the limited modulation bandwidth of LEDs restricts the achievable communication rate. Besides the previously mentioned key technologies, adopting widely used high-order modulation formats is an effective means to improve spectral efficiency, such as various single-carrier modulation (SCM) and multi-carrier modulation formats [62,63]. Research and optimization of these modulation formats have been extensively conducted in other fields but are still necessary in the context of OWC. Although various modulation and multiplexing techniques have been proposed to improve the utilization efficiency of the visible light spectrum, high-order modulation or excessive multiplexing channels may introduce system complexity. Additionally, due to the fact that VLC typically employs intensity modulation with direct detection (IM/DD), the receiver sensitivity is relatively low, and the limited illumination intensity also limits the received power, resulting in a lower signal-to-noise ratio (SNR). Moreover, considering the scenario of mobile users, the frequency-selective fading effect and frequency offset may severely affect the performance of these frequency-multiplexed or modulated schemes. On the other hand, as previously mentioned, µLEDs have the potential to break through the E-O bandwidth bottleneck, making them a strong candidate for the next-generation high-capacity VLC transmitter. Therefore, research on µLED materials, structures, and fabrication processes will proceed in the future [64]. Furthermore, VLC systems based on semiconductor lasers (LDs) are also being researched [65], as LDs have higher modulation bandwidth, higher energy efficiency, and better beam directionality.

For indoor NIR light communication, high-speed rates can be easily achieved, while beam steering is a key challenge, as NIR light communication realizes point-to-point communication. Although the use of highly directional beams can provide broadband access to different users, the complexity and cost of the beam steering system become the bottleneck for the practical application of NIR light communication, especially in the case of multiple users. As demonstrated in our previous work, using an SLM to generate multiple beams is an effective solution for beam steering and multi-user communication. Furthermore, for mobile users, the fast re-acquisition of the receiver and the high-speed automatic alignment of the beam are challenging issues that need to be addressed. For the accurate localization of the receiver, light detection and ranging (LiDAR) is a promising solution [66], with high distance or speed resolution. The integration of positioning systems into OWC architecture is an area that requires further research.

Despite the challenges that need to be addressed, indoor OWC is promising to play a pivotal role in future high-speed indoor wireless communication due to its unique advantages.