Takaya Yamazato

Visible light communication (VLC) evolved from early Japanese studies showing that white LEDs can provide both illumination and data transmission, and has grown into a broader “VLC family” spanning optical wireless communication (OWC), optical camera/image sensor communication (OCC/ISC), display/screen-camera communication (DCC/SCC), and underwater OWC (UOWC). In parallel, LiFi reframed VLC from point-to-point links to networked optical wireless systems. This paper reviews research trends in Japan across these domains. We first adopt a transmitter-centric perspective, from camera-compatible, high-speed architectures to multi-functional co-designs and dynamic, machine-learning-integrated emitters. We then take a reception-centric view of OCC/ISC— including rolling-shutter camera reception and DCC/SCC—followed by an overview of UOWC as an emerging application domain.

The increasing proliferation of internet-connected devices, together with the widespread reliance on wireless communication in daily activities, are limiting the capabilities of the radio frequency (RF) spectrum to meet data traffic demands. In particular, the rise in connectivity requirements to realize next generation networks, coupled with the growth in utilization of artificial intelligence (AI) and its associated transmission of massive amounts of data, contribute to the need to leverage the optical spectrum to complement the already-congested RF bands. Optical wireless connectivity encompasses a range of solutions that can support data transfer in various use cases, ranging from intra-chip interconnects to deep-space communication. They benefit from advantages in terms of operation within a broad unlicensed spectrum, minimization of RF interference, and inherent security. On the other hand, optical wireless communication (OWC) systems might experience a decline in performance as a result of factors like atmospheric conditions and link blockage. This requires customizing system designs according to the considered use cases and investigating new solutions for realizing optical wireless connectivity. In this article, we present an overview of OWC, encompassing both its conceptual frameworks and the scientific and technological advancements that hold the potential for shaping its future in next generation networks.

Image sensor communication (ISC) employing a propeller-LED transmitter encodes data along rotating light trails. We present an analytical framework that (i) constructs a single-LED, single-blink light trail model that maps optical power to pixel values, and (ii) integrates a probabilistic noise model to derive a closed-form bit-error rate (BER) using the Q-function. Trimodal pixel-value histograms motivate an adjacent-only inter-symbol interference (ISI) model in which the decision at segment j depends on adjacent segments. Applying a hardest-pair midpoint threshold yields per-segment BER and a general BER after marginalization. We further provide practical sufficiency conditions under which adjacent-only ISI is adequate, and validate its tightness against Monte Carlo simulations and experimental results. Using the analytical BER, we select the control angle that maximizes throughput while satisfying a target BER reliability constraint.

Accurate site-specific radio propagation characterization, including both the Power Delay Profile (PDP) and Angle of Arrival (AoA), is critical for deploying mobile communication systems. However, conventional methods are inefficient because PDPs measured with an omnidirectional antenna fail to capture the AoA. Consequently, separate, time-consuming measurements using mechanically rotated directional antennas or complex calibrated array antennas are required. This paper proposes a novel and efficient method for simultaneously estimating both PDP and relative AoA using data obtained from a single moving omnidirectional antenna. The proposed technique leverages the phase information inherent in the channel data, specifically by processing the phase difference between spatially adjacent profiles acquired during the antenna's movement. This approach eliminates the need for mechanical rotation or specialized array hardware. Experimental results demonstrate the feasibility and effectiveness of the proposed method, highlighting its potential to drastically reduce the time and complexity required for comprehensive channel characterization.

Coarse-to-fine partitioning (CFP) algorithm is proposed for angle-of-arrival (AoA) estimation with a controllable liquid crystal display (LCD) in front of a photodetector (PD). By electrically switching each pixel of the LCD in transparent or opaque, the position of the aperture through which the incident light passes is controlled to estimate the AoA of the incident light to the PD. The size of the aperture needs to be a certain size to maintain the accuracy of the AoA estimation, which consists of a region of several pixels. In the conventional algorithm, with a fixed size of the aperture equal to the size of resolution unit, the entire surface of the LCD was exhaustively searched horizontally and vertically. In our CFP algorithm, the entire surface is partitioned into two or three equal-area regions, and the intensities observed at the PD when pixels consisting of each of such regions are alternately switched to transparent mode are compared. Among them, the region with the highest intensity observed is chosen as the next candidate region through which the incident light passes. This process is iteratively applied, partitioning the candidate region from coarse to fine, until the region reaches the resolution size. For the LCD consisting of 2560 × 1440 pixels with the resolution unit size of 10 × 10 pixels, to estimate the AoA, our CFP algorithm requires 30 searches while the conventional algorithm requires 400 searches. Experimental results show that although intensity measurements are repeated fifteen times for each search to stabilize the intensity measurement values in both of the conventional algorithm and our CFP algorithm, our algorithm estimates it with the error less than one degree. Furthermore, by adding another PD into the system, from the AoA estimation for the two PDs with our CFP algorithm, the distance estimation to the light source from 200 mm to 900 mm is experimentally shown with a relative error of 4.5$\%$.

We propose an estimation method of the relative position of a light emitting diode (LED) from a rolling-shutter (RS) camera with a cross filter (CF). The CF is set in front of the camera lens so that incident light can be scattered into two straight lines extending about $\pm \pi /4$ from the horizontal row of the CMOS RS image sensor (IS). With the geometry of a right-angled isosceles triangle, our method makes it possible to estimate the coordinates of the LED in the IS plane when the scattered light is captured at two points in a scanline. With the CF, our method continues to update the estimated coordinate of the LED as long as the row captures scattered light at two points. Naturally, our method makes it possible row by row when multiple LEDs are used or when they are moving. In contrast, without the CF, estimation is not possible until the scanline reaches the row which captures the whole LED, and the estimated coordinates are not updated once it is done for that frame. Experiments verify that our estimation method works with the estimation error less than five pixels, and extends the region of a frame in which estimated coordinates are output, even in a moving environment, while the conventional method (without CF) does the estimated coordinates only once per frame.

A secure method is proposed for parallel-transmission (PT) image-sensor (IS) based visible light communication, where $M$ luminance signals in one-dimensional order produced by inverse discrete Fourier transform (IDFT) are mapped in a two-dimensional plane transmitted in free space between liquid crystal display (LCD) and IS. Each position in the one-dimensional order is preassigned to a region on the two-dimensional plane, which is known to both transmitters and receivers. In the secure method, the one-dimensional order of the $M$ DFT-related luminance signals is shuffled, and only the legitimate receiver can restore the original order. To share the original order secretly between authorized transmitter/receiver pairs, we use three integers $e$, $d$, and $n$ based on RSA cryptosystem. When a transmitter chooses an integer $I$ between 0 and $M!-1$, it uses the $I$th order among $M!$ ways to rearrange the $M$ DFT-related luminance signals. For the transmitter to tell $I$ as the ciphered integer $u$, it calculates the modulo-($M+1$) of $u$ to divide $u_{0}$, $u_{1}$, $\ldots$, $u_{H-1}$. Then, it inserts a new luminance signal called “mark” at the $u_{h}$th position in the rearranged order over $H$ consecutive frames. Although all receivers detect the position of “mark” every frame, only the legitimate receiver correctly calculates $I$, from the mark-detected positions $v_{0}$, $v_{1}$, $\ldots$, $v_{H-1}$, using the private integer $d$. Receivers without $d$ must search the original order among $M!$ possible permutations every frame even though they detect all the positions of “mark” correctly. Experimental results show that the luminance value $L$ of “mark” exists that receivers can distinguish it from $M$ DFT-related signals, in the transmission distance of 1.5 meters to seven meters and that the probability is extremely low that receivers without $d$ demodulate the information bits correctly.

This study aims to provide adaptability to changes in the communication distance of rolling shutter based visible light communication (RS-VLC). A key issue in RS-VLC is the trade-off between the data rate and communication distance. To resolve this issue, we propose a superimposed data transmission method in which two sets of data, namely short- and long-range data, are superimposed and transmitted simultaneously. Short-range data are transmitted in long-bit sequences with a small number of iterations, whereas long-range data are transmitted in short-bit sequences with a large number of iterations. These two datasets can be separately decoded at the receiver by superimposing and transmitting them at different amplitudes. The proposed method can decode all the superimposed data at short ranges, thus suppressing data rate degradation. As the communication distance increases, the proposed method decodes only the available data, allowing for data reception over longer ranges, albeit at a lower data rate. We further extend the proposed method by adding a third set of data, namely mid-range data, for further flexibility in communication distance and data rate. The experimental results show that the proposed method achieves data reception over a longer distance (up to 150 cm) while suppressing throughput degradation at short-ranges, that is, it mitigates the trade-off between the data rate and communication distances compared to conventional $M$-ary pulse amplitude modulation systems.

This paper proposes an image sensor communication (ISC) system based on light trails using rotary LEDs. The light trail refers to an optical phenomenon where light appears elongated and forms a trajectory due to the movement of luminous objects. In this study, we developed a propeller LED transmitter (P-Tx) that rotates LEDs with a propeller to generate and transmit light trail signals. According to the LED's rotation radius, the proposed system segments light trails to achieve high-capacity data transmission at short distances and high-trust transmission at long distances. In addition, we perform the rotation coordinates system conversion for the captured light trails to accurately and simply detect the signal LEDs' coordinates. We experimentally evaluate the proposed system's communication performance and data rate. The experimental results demonstrate that our system is capable of meeting communication requirements at various distances and exhibits significant potential in enhancing the data transmission rate of ISC.

This paper presents the successful reception of visible light communication (VLC) signals, transmitted from an LED array, by an image sensor that utilizes a rolling shutter while driving at 40 km/h. The rolling shutter image sensors used in commercial cameras, such as those found in smartphones and dashcams, capture images line by line at a fast rate, allowing for VLC signal reception in vehicular environments. By analyzing the relationship between the signal reception rate for each line and the frame rate, we demonstrate that parallel VLC signals transmitted from the LED array can be received even while in motion. Notably, to our knowledge, this is the first time that an automobile moving at 40 km/h has successfully received a VLC signal.