Underwater visible light communication (UVLC) has gained attention for its effectiveness in undersea exploration. However, due to background noise like sunlight, UVLC has mainly been applied in deep ocean environments. To extend its use to shallow water, a scheme was proposed that applies maximal-ratio combining to an angle diversity receiver (ADR) with multiple photodiodes (PDs) positioned at different angles, focusing on optimizing PD placement. Previous studies fixed each PD’s field of view (FOV) at 90 degrees without optimization, and neglected movement in the Z direction (depth changes). This paper explores optimal FOV settings for PDs to maximize area coverage (AC). Four types of communication areas are considered, optimizing parameters for each under solar noise and receiver movement in the XYZ direction. Results show that uniform optimization, setting all FOVs equally, works well for areas with similar aspect ratios but performs poorly in distinctly vertical or horizontal areas, reducing AC significantly. In contrast, individual optimization, adjusting FOV based on PD installation angle, improves AC by over 8% compared to uniform optimization.

Lighting-constrained visible light communications (VLC) can provide both high-speed wireless communications and dimming control of the lighting state of indoor environments. Here, a dimming control method based on multi-pulse pulse position modulation (MPPM) has been investigated to simultaneously achieve high-data-rate transmissions and dimming control. MPPM data is transmitted on r slots selected from the available ones. Dimming-control-based MPPM can also be realized by changing the value of r. When the dimming level changes, the data transmission rate changes, resulting in a significant degradation of the average data transmission efficiency averaged over the data transmission efficiency of each dimming stage of VR-MPPM. This paper proposes a selective VR-MPPM (SVR-MPPM) scheme to improve this average data transmission efficiency by selecting a value of r that achieves the highest possible transmission efficiency while satisfying a wide dimming range.

This paper aims to propose a multi-color-based modulation scheme for underwater simultaneous light information and power transfer (USLIPT), which can perform wireless charging while maintaining a desired communication rate. In particular, we proposed an enhanced optimized color shift keying (EOCSK) that can adjust signal space enhancement coefficient$\alpha(0\leq\alpha\leq 1)$to balance communication and energy harvesting performances. We further analyze the spectral efficiency, averaged bit error rate (BER), and harvested energy of USLIPT using time-switching (TS) and power-splitting (PS) receivers. It was found that the average BER performance of the proposed system can be improved when the value of$\alpha$is increased. Conversely, when the value of$\alpha$is decreased, the harvested energy can be increased. Furthermore, focusing on the proposed method's maximum communication distance, it was shown that a larger$\alpha$increases the possible communication distance. Still, the amount of power supplied can be increased at the expense of a slight reduction in communication distance, indicating that the optimal$\alpha$is other than$\alpha=1$, i.e., conventional optimized CSK. For example, when the desired spectral efficiency is 1.5 b/s/Hz, the harvested energy of the proposed system with$\alpha=0.6$is more than twice that with$\alpha=1$at the expense of reduced communication distance within 1 m.

Underwater simultaneous light information and power transfer (USLIPT) has recently attracted attention for the efficient exploration and excavation of international seabed resources. In this paper,$N$parallel code shift keying (N-CSK), in which$\mathrm{N}$pseudo-noise (PN) codes are selected according to the transmitted signal, is adopted as a multiple-access scheme to study USLIPT with multiple access. In this paper, we propose an inverted N-CSK scheme to further improve the power-feeding performance of the USLIPT using N-CSK by constantly turning on$M-N$unselected LEDs, which are not used as transmit signals in N-CSK, to improve the power-feeding performance while maintaining the amount of information transmitted. The performance of the bit error rate (BER) and the power supply are evaluated by theoretical analysis. The results show that the power feeding performance degrades with decreasing$N$in N -CSK, but improves with decreasing$N$in the proposed scheme because the power feeding performance improves with increasing$N$. This is because, in the proposed method, the larger$N$is, the fewer the number of LEDs that are lit. However, no matter how large$N$is, the power-feeding performance of the proposed method is superior to that of N -CSK.

Recently, underwater visible light communications (UVLC) have attracted much attention for searching and excavating undersea resources effectively. While UVLC suffers from background noise from sunlight, the application of UVLC has been limited to the deep ocean. To realize the UVLC in shallow water, UVLC with the maximal-ratio combining scheme applied to angle diversity receiver with multiple Photodiodes (PDs) at different angles was proposed and, in particular, the optimization of PD placement has been studied. However, as the field of view (FOV) of each PD was fixed at 90 degrees, its optimization has not been studied. Therefore, this paper investigates the optimal FOV of PDs that maximize the average SNR of a square communication area of$n$centered at the initial position of the receiver. As a result, in an area of$k$, narrowing the FOV of the receiver with evenly distributed PDs to 25 degrees improved the average SNR by about 3 dB compared to the case where the FOV was set to 90 degrees. In the$i$area, narrowing the FOV of the receiver with PDs placed on the sides to 75 degrees improved the average SNR by about 0.8 dB compared to the case where the FOV was set to 90 degrees.

As a method of dimming support for visible light communications using a massive LED array, we consider a hybrid pulse width modulation/digitally controlled pulse amplitude modulation (PWM/DPAM) system. Especially, in this paper, we consider a digital color shift keying (DCSK) using an RGB-LED array as an optical intensity modulation scheme, that transmits data through the ratio of the optical intensities (i.e., color) emitted by red, green, and blue LEDs. In DCSK, since only one color is activated in each RGB-LED at a time, the color can be represented by the combination of the digitally (i.e., linearly) controlled “ON-OFF” LEDs. In general, for the dimming control system of DCSK, two schemes have been considered. One is the dimming control by PWM, which changes the duty cycle of optical transmit signals, and the other is the dimming control by DPAM, which changes the number of active LEDs in the RGB-LED array. In this paper, PWM and DPAM are combined to realize a higher spectral efficiency than PWM and a wider dimming range than DPAM. We evaluate the error performances of the proposed system, DCSK with PWM, and DCSK with DPAM from a simulation analysis under several measured light dimming levels. The results show that DPAM should be used only for low bit rate systems because the effect of inter-symbol-interference (ISI), caused by the LED frequency response, increases at a high bit rate. While, PWM is significantly robust against ISI because the PWM signal duration is limited in a symbol duration at the low dimming levels and the empty duration can mitigate the effect of ISI even if the bit rate is high. When focusing on symbol error rate performances corresponding to the dimming levels, the hybrid dimming control and the PWM dimming control can achieve lower signal energy-to-noise-ratio at the dimming level of low and high, respectively.

Graph clustering has recently attracted much attention as a technique to extract community structures from various kinds of graph data. Since available graph data becomes increasingly large, the acceleration of graph clustering is an important issue for handling large-scale graphs. To this end, this paper proposes a fast graph clustering method using GPUs. The proposed method is based on parallelization of label propagation, one of the fastest graph clustering algorithms. Our method has the following three characteristics: (1) efficient parallelization: the algorithm of label propagation is transformed into a sequence of data-parallel primitives; (2) load balance: the method takes into account load balancing by adopting the primitives that make the load among threads and blocks well balanced; and (3) out-of-core processing: we also develop algorithms to efficiently deal with large-scale datasets that do not fit into GPU memory. Moreover, this GPU out-of-core algorithm is extended to simultaneously exploit both CPUs and GPUs for further performance gain. Extensive experiments with real-world and synthetic datasets show that our proposed method outperforms an existing parallel CPU implementation by a factor of up to 14.3 without sacrificing accuracy.

We propose an efficient content-based image retrieval (CBIR) method for position estimation of mobile devices. The idea is to use videos of first-person vision associated with geographical position information as the database. When a user sends a current subjective image, the system estimates the position using CBIR. Since features extracted from images are in general high-dimensional vectors, thousands of vectors are extracted even from a single image, resulting in high processing cost. To tackle this problem, we have proposed a method in which features are compressed using LSH, and GPU is used for accelerating CBIR. Nevertheless, it suffered from performance degradation due to write conflicts among different threads. This paper presents an improved method which avoids write conflicts and modifies LSH algorithm to improve accuracy. Also, we demonstrate the efficiency and accuracy of the proposed scheme through experiments using a video dataset.

In this paper, the extended pseudo orthogonal M-sequence is proposed for the optical-wireless code-shift-keying (CSK) system. The extended pseudo orthogonal M-sequence is a new optical pseudo-noise (PN) code which consists of the modified pseudo orthogonal M-sequence sets (MPOMS) and the bi-orthogonal code. The performance of the CSK systems is enhanced by increasing the number of the optical PN codes. As the conventional optical PN code, the bi-orthogonal code and the pseudo-orthogonal extended prime code set can generate the number of the optical PN codes same as the code length. The extended pseudo orthogonal M-sequence can achieve the performance enhancement of the CSK systems because of its generation of the large number of the codes per the code length. The CSK system using the extended pseudo orthogonal M-sequence has better performance than the CSK systems using the conventional PN code.

Uncertain databases have been widely developed to deal with the vast amount of data that contain uncertainty. To extract valuable information from the uncertain databases, several methods of frequent itemset mining, one of the major data mining techniques, have been proposed. However, their performance is not satisfactory because handling uncertainty incurs high processing costs. In order to address this problem, we utilize GPGPU (General-Purpose computation on GPU). GPGPU implies using a GPU (Graphics Processing Unit), which is originally designed for processing graphics, to accelerate general purpose computation. In this paper, we propose a method of frequent itemset mining from uncertain databases using GPGPU. The main idea is to speed up probability computations by making the best use of GPU's high parallelism and low-latency memory. We also employ an algorithm to manipulate a bitstring and data-parallel primitives to improve performance in the other parts of the method. Extensive experiments show that our proposed method is up to two orders of magnitude faster than existing methods.

In this paper, M-CSK with extended prime code sequence to enhance a data transmission rate of CSK is considered. Moreover, the bit error rate (BER) performance of M-CSK/CDMA is evaluated by taking into account scintillation, background-noise, avalanche photo-diode (APD) noise, thermal noise, and signal dependence noise through theoretical analysis. It is shown that the data transmission rate of the M-CSK system is better than that of the CSK system using extended prime code sequences. However, the BER performance of the M-CSK system is degraded by increasing the number of selected extended prime code sequences.

In this paper, to increase normalized throughput of the optical wireless code-division multiple-access (CDMA) system, modified pseudo orthogonal M-sequences (MPOMS) are considered as optical pseudo-noise (PN) code. Moreover, the bit error rate (BER) performance of the optical wireless CDMA system using multi-pulse pulse position modulation (MPPM) is also derived in the presence of the scintillation, background-noise, avalanche photo-diode (APD) noise, thermal noise and signal dependence noise. Consequently, the BER performance in consideration of normalized throughput of the MPPM system with MPOMS is superior to that of the MPPM system with conventional pseudo orthogonal M-sequences in the single-user case and multi-user case.

In this paper, optical wireless multi-pulse pulse position modulation (MPPM) system with modified pseudo orthogonal M-sequence sets is proposed. Moreover, the symbol error rates of the proposed system are analyzed by taking into account the scintillation, background-noise, avalanche photo- diode (APD) noise, thermal noise, and signal dependence noise in the single-user case. Consequently, the normalized throughput in consideration of SER of the proposed system is superior to that of the conventional MPPM systems. Furthermore, the normalized throughput of the proposed system can achieve about 1.25 [bit/chip].

In this paper, the atmospheric optical code-division multiple- access (CDMA) system with Pseudo-orthogonal M-sequences is proposed. The bit error rate (BER) of the proposed system using pulse position modulation (PPM) by taking into account the effects of the scintillation, avalanche photo-diode (APD) noise, thermal noise, and multiple access interference (MAI) are analyzed. By comparing with the optical orthogonal code (OOC) and the prime code, the proposed system with pseudo- orthogonal M-sequences is better than the systems with other PN codes. But, the scintillation has an enormous effect on the performance of the proposed system.