Search

In this article, we propose an efficient inexact gradient descent method to obtain the channel capacity-achieving input distribution of visible light communication (VLC). Under both a peak and an average optical power constraints, finding the channel capacity of VLC is formulated as a mixed continuous-discrete optimization problem without an analytical expression of the objective function. To solve this challenging problem, we first adopt the numerical integration method to approximate the objective function and its gradient. Here, we prove that the gaps between the original functions and the approximations can be arbitrarily small. Then, based on the approximated functions, we describe the method, and theoretically show that the obtained solution sequence converges to the channel capacity-achieving discrete distribution of VLC. We also provide simulation results to verify the effectiveness and optimality of the proposed method. More importantly, the simulations numerically reveal that on-off keying (OOK) modulation achieves the capacity of VLC channel at low signal-to-noise ratio (SNR), while pulse amplitude modulation (PAM) achieves the capacity at high SNR.

In this paper, we consider a common unicast beamforming network where Alice utilizes the communication to Carol as a cover and covertly transmits a message to Bob without being recognized by Willie. We investigate the beamformer design of Alice to maximize the covert rate to Bob when Alice has either perfect or imperfect knowledge about Willie's channel state information (WCSI). For the perfect WCSI case, the problem is formulated under the perfect covert constraint, and we develop a covert beamformer by applying semidefinite relaxation and the bisection method. Then, to reduce the computational complexity, we further propose a zero-forcing beamformer design with a single iteration processing. For the case of the imperfect WCSI, the robust beamformer is developed based on a relaxation and restriction approach by utilizing the property of Kullback-Leibler divergence. Furthermore, we derive the optimal decision threshold of Willie, and analyze the false alarm and the missed detection probabilities in this case. Finally, the performance of the proposed beamformer designs is evaluated through numerical experiments.

The influence of Fe(III) on the interaction between roxarsone (ROX) and humic acid (HA) was investigated by multi-spectroscopic techniques. The fluorescence quenching experiment indicated that the fluorescence intensity of HA-ROX was quenched by Fe(III) through a static quenching process. Synchronous fluorescence spectra provided further information concerning the competitive combination between ROX and Fe(III) for HA. The results of the dialysis equilibrium experiment confirmed the existence of Fe(III) (0.05-0.1 mmol/L) promotes the combination of HA and ROX. Binding mechanisms were further characterized by FTIR spectroscopy, and the carboxyl functional group is involved in the binding process of HA/Fe/ROX. In addition, acidic and neutral conditions are more conducive to the combination of ROX and HA/Fe than alkaline conditions. The above discussion is of great significance in understanding the environmental fate of ROX under the coexistence of Fe(III) and HA.

It is of great importance to power the nodes of the rechargeable wireless sensor network to detect events continuously in the area of interest. This paper proposes a joint data transmission and on‐demand charging algorithm based on the space correlation. The new algorithm optimises the event detection, data forwarding and node charging jointly to improve the charging efficiency. First, the active nodes participating in the event detection are selected using an improved iterative node selection method to reduce the number of nodes working concurrently. Then, the greedy data transmission scheme based on grid partition is proposed to transmit the observed data to the sink node. Finally, the nodes in the networks are charged using the on‐demand charging method based on grid partition, which greatly decreases the charging frequency and energy loss of the mobile charger. The simulation results demonstrate that the proposed method has superior performance in the distance travelled by the mobile charger, the energy utilisation, the average energy consumption of the mobile charger and the node charging latency.

Tri(2-chloroethyl) phosphate (TCEP) with the initial concentration of 5 mg/L was degraded by UV/H2O2 oxidation process. The removal rate of TCEP in the UV/H2O2 system was 89.1% with the production of Cl− and PO43− of 0.23 and 0.64 mg/L. The removal rate of total organic carbon of the reaction was 48.8% and the pH reached 3.3 after the reaction. The oxidative degradation process of TCEP in the UV/H2O2 system obeyed the first order kinetic reaction with the apparent rate constant of 0.0025 min−1 (R2=0.9788). The intermediate products were isolated and identified by gas chromatography-mass spectrometer. The addition reaction of HO• and H2O and the oxidation reaction with H2O2 were found during the degradation pathway of 5 mg/L TCEP in the UV/H2O2 system. For the first time, environment risk was estimated via the “ecological structure activity relationships” program and acute and chronic toxicity changes of intermediate products were pointed out. The luminescence inhibition rate of photobacterium was used to evaluate the acute toxicity of intermediate products. The results showed that the toxicity of the intermediate products increased with the increase of reaction time, which may be due to the production of chlorine compounds. Some measures should be introduced to the UV/H2O2 system to remove the highly toxic Cl-containing compounds, such as a nanofiltration or reverse osmosis unit.

The aggregated visible light communication (VLC) and radio frequency (RF) system, which can be viewed as a heterogeneous multi-input-multi-output system, can improve data rate compared to the conventional RF communication systems. In this paper, we first develop optimal power allocation schemes for the aggregated VLC-RF systems for the single and the multi-light-emitting diode scenarios under different dimming control setups. Moreover, we study the energy efficiency maximization problem of the considered system with the minimum rate requirement, transmitted power constraint, and the dimming control consideration which is non-convex. By using the Dinkelbach-type algorithm, we tackle this problem by solving a sequence of convex problems which converges to the global solution. Finally, the effect of critical parameters, such as total power threshold, dimming level, and bandwidths, are revealed by some selected numerical results.

Most nodes of Mine Internet of Things (Mine IoT) are powered by batteries, and wireless charging using mobile chargers (MCs) is an effective way to make nodes work sustainably. A novel hybrid charging method combining the controlled and opportunistic MCs (C&O charging) is proposed in this study. Workers (such as the repairmen and gas inspectors) carrying portable chargers are proposed to be opportunistic MCs to provide an incidental charging service for the surrounding rechargeable Mine IoT nodes while doing its own work to reduce the payload of controlled MCs. The hybrid charging model based on the incidental charging ability of the opportunistic MC is constructed and the scheduling strategy of the controlled MC and the queueing management scheme of the charging request are also proposed. The simulation results indicate that the power demands of the majority of the nodes in the maintenance areas can be met or partially met by opportunistic MCs and the charging time of C&O charging is greatly decreased compared to that of only using controlled MCs.

Tri(2-chloroisopropyl) phosphate (TCPP), a common organophosphate flame retardant, was frequently detected in the environment and posed threats to human health. In this work, the main component of ilmenite FeTiO

The protective mechanism of extracellular polymeric substance (EPS) secreted by a harmful cyanobacteria against tannins allelochemicals was explored in this study. The binding properties of soluble EPS (SEPS) and bound EPS (BEPS) of Microcystis aeruginosa to tannic acid (TA) were investigated via fluorescence spectroscopy. The results suggested that TA interacted with the proteins in SEPS and BEPS mainly with binding constants of 5.26 and 7.93 L/mol, respectively; TA interacted with the humic acids in SEPS and BEPS mainly with binding constants of 5.12 and 5.24 L/mol, respectively. Thermodynamic experiments confirmed that the binding was mainly controlled by the hydrophobic force. Combined with Fourier transform infrared spectroscopy, it was found that the amine, carbonyl, carboxyl, and hydroxyl groups in EPS were the main functional groups contributing to the interaction of TA with EPS. The existence of EPS reduced the toxicity of TA to algal cells, with the 96 h inhibition rate of 40 mg L

This study provides a new thinking for the efficient utilization of permanganate (Mn (VII)) in eutrophic water treatment. Eutrophic water contained a large amount of extracellular polymeric substances (EPS) with reduction and chelation; this study used phenol as typical organic matter and cyanobacteria EPS as a representative EPS to explore the mechanism by which EPS influences the oxidation of phenol by Mn(VII) at pH 5.0-9.0. The results showed that under the condition of pH 5.0-7.0, adding 0.2-10 mg/L EPS to the Mn(VII) system could effectively improve the oxidation efficiency of Mn(VII) for phenol. EPS promoted the continuous formation and stability of in situ EPS-MnO

In this work, natural sepiolite (SEP) was disaggregated and activated by hydrochloric acid to obtain a novel adsorbent with a great potential for defluoridation of groundwater. The adsorbent was ch...

Visible light communication (VLC) is seen as a supplement for fifth-generation (5G) wireless communication in short-range high data rate communication applications [1] . A reliable VLC system relies on an accurate estimate of the position and orientation of the receiver, which corresponds to the six-dimensional positioning problem mentioned in [2] . In this paper, we investigate the simultaneous position and orientation estimation (SPO) problem using received signal strength (RSS), for a visible light system containing multiple LEDs and multiple photodiodes (PDs) (MLMP). Although in general, the position and orientation of the receiver can be represented by a vector and a rotation matrix, respectively, the constraints imposed by the rotation matrix make the numerical optimization in the estimation process cumbersome, e.g, the commonly used constrained optimization method is often very complex and non-robust. Therefore, in this paper, we design two SPO algorithms using the principle of optimization on manifolds, which alleviates the constraints from the rotation matrix. In addition, we propose an initialization algorithm, based on the direct linear transformation (DLT) principle, to obtain an initial estimate in closed-form for the iterative algorithms. To evaluate the performance of the proposed RSS-based SPO algorithms, we derive the Cramer-Rao bound (CRB). In particular, the orientation error component of the CRB corresponds to the intrinsic CRB or the CRB on manifolds, which measures the error in the estimated rotation matrix in a physically meaningful way. Finally, computer simulations show an asymptotic tightness between the performance of the proposed algorithms and the theoretical lower bound, demonstrating the effectiveness of the proposed solutions.

Although multicomponent transition-metal oxides have been widely studied as electrocatalysts in both the oxygen reduction reaction (ORR) and the oxygen evolution reaction (OER), precise control of ...

In our previous study, we prepared the granules by embedding artemisinin into alginate-chitosan using microcapsule technology. These granules can release artemisinin sustainably and have a strong inhibitory effect on the growth of both single Microcystis aeruginosa and mixed algae. To safely and effectively use artemisinin sustained-release granules to control algal blooms, the ecotoxicity was studied by assessing their acute and chronic toxicity to Daphnia magna (D. magna) and Danio rerio (D. rerio), along with their antioxidant activities. The results showed that the 48-h median effective concentration (EC50) of pure artemisinin to D. magna was 24.54 mg/L and the 96-h median lethal concentration (LC50) of pure artemisinin to D. rerio was 68.08 mg/L. Both values were classified as intermediate toxicity according to the Organization for Economic Co-operation and Development (OECD). The optimal algae inhibitory concentration of artemisinin sustained-release granules (1 g/L) had low acute toxicity to both D. magna and D. rerio. The sustained-release granules had higher chronic toxicity to D. magna than to D. rerio. Partial indices of D. magna were inhibited by granules when the concentrations were larger than 0.1 g/L. Low granule concentration had an inductive effect on antioxidant enzyme activities in D. magna and D. rerio. With the increase of the exposure concentration and time, the enzyme activity presented a trend of first increasing and then decreasing, and the overall changes were significant. The change trend and range of enzyme activity indicated that the granules could cause serious oxidative stress to D. magna and D. rerio, and the changes were consistent with the results of toxicity experimentation.

In biomass utilisation, size reduction caused by grinding is a critical procedure because it greatly changes the particle size and shape of the biomass sample, which greatly influences its energy conversion efficiency. Raw biomass is difficult to grind due to its fibrous nature and therefore needs to be pretreated before grinding. Traditional torrefaction pretreatment consumes more energy and changes the chemical composition of biomass. In this study, we select wheat straw with abundant yield in northern China as the biomass sample for testing and introduce a new freezing pretreatment method into the grinding procedure. This method incorporates the freezing of wheat straw with liquid nitrogen. We evaluate the performance of freeze grinding in terms of grindability, energy consumption, chemical composition and energy conversion efficiency, and compare it with that of torrefaction grinding to investigate the effects of pretreatment methods on wheat straw. Particle size and shape distribution are used to characterise grindability. Moreover, we improve sample preparation, image acquisition and segmentation methods to reduce the impact of particle agglomeration. Experimental results demonstrate that freezing as pretreatment improves the grindability and energy conversion efficiency, whilst reducing energy consumption and maintaining chemical composition of wheat straw.

Pt-based alloys have been widely used as high-performance electrocatalysts for the oxygen reduction reaction (ORR). These catalysts contained high Pt contents, typically 60–70 at.% or higher, since further decreasing Pt contents would significantly lower the desired electro-catalytic activity. High-entropy alloys (HEAs) with built-in stability through their low free-energy phases provided a promising route to prepare Pt-based alloy catalysts with less than 50 at.% Pt while maintaining high ORR activity and stability in various environments. In this work, we reported a top-down de-alloying synthetic method to controllably incorporate five immiscible metals in one nanoscale solid phase. By predetermining four elements (Al, Cu, Ni and Pt) and alternating the fifth element including Pd, V, Co, Mn, etc., a series of nanoporous HEAs (np-HEAs) with Pt content of ~20–30 at.% were obtained. Among these quinary alloys, the Al-Cu-Ni-Pt-Mn np-HEA exhibited the best ORR catalytic activity and electrochemical cycling durability, well exceeding the commercial Pt/C catalysts. We expected these nanostructured HEA would offer a new class of alloy catalysts in many electrochemical reactions.

In this paper, we investigate a novel simultaneous lightwave information and power transfer (SLIPT) in visible light communication (VLC) systems, where a photo diode (PD) and a solar panel are utilized as the information receiver and the energy harvester, respectively. By systematically analyzing both the information receiver and the energy harvester, we obtain the explicit expressions to characterize the illumination-rate-energy region. Based on the derived expressions, we investigate the downlink unicast transmission of multi-LED multi-user SLIPT VLC networks, and study the total transmit power minimization problem under the rate requirements, the minimum energy harvesting requirements, and dimming control constraints. To solve such non-convex problem, we exploit the semidefinite relaxation (SDR) technique and relax the problem into a convex problem, which can be efficiently solved via interior-point methods. Moreover, for the sake of users’ fairness, we further investigate the beamformer design to maximize the minimal rate under both minimum energy harvesting and dimming control constraints. Finally, the numerical results are provided to evaluate the proposed SLIPT system.

To investigate the relationship between cyanobacteria decomposition and the remobilization of heavy metals in Taihu Lake, the indoor simulation experiments were conducted. The areas of Taihu Lake that undergo harmful algal blooms mostly caused by excessive cyanobacteria have serious problems of heavy metal pollution. The results showed that cyanobacteria decomposition can release heavy metals into the water and change the total contents and chemical speciation of Cr, Ni, Cu, Zn, As, Cd, Hg, and Pb in sediment due to the change of physical and chemical properties in overlying water and sediment. The decomposition rate of cyanobacteria with sediment was clearly faster than that without sediment, and decomposition changed the pH and dissolved oxygen (DO) concentrations in overlying water. The cyanobacteria decomposition reduced the oxidation-reduction potential (ORP), and increased organic matter (OM) and total organic carbon (TOC) in the surface sediment. According to ecological risk assessment, the cyanobacteria decomposition increased the degree of heavy metal pollution and the potential ecological risk in sediment.

The colloid of ferrous sulfide modified by humic acid (HA-n-FeS) shows good reduction and immobilization efficiency for variable-valence heavy metals in wastewater. The removal efficiency of HA-n-FeS for halogenated organic pollutants, however, remains unclear, especially in the absence and presence of oxygen. This study addressed this issue by exploring the effect and mechanism of dissolved oxygen on the degradation of tetrabromobisphenol A (TBBPA) by the HA-n-FeS colloid in water. The results showed that the removal efficiency of different concentrations of TBBPA (5,10, and 20 μm) by the HA-n-FeS colloid was 33.16%, 20.48%, and 22.37% in the absence of oxygen, respectively. When TBBPA reacted with the HA-n-FeS colloid, the concentration of Fe(II) and S(-II) remained stable. The adsorption of HA-n-FeS was the main mechanism of removing TBBPA in the absence of oxygen. In the presence of oxygen, the removal efficiency of TBBPA by the HA-n-FeS colloid was 82.37%, 56.80%, and 43.78% (for the above-mentioned TBBPA concentrations), respectively. In addition, the removal capacity of TBBPA by HA-n-FeS was 39.63, 52.21, and 89.75 mg/g, respectively. The concentration of Fe(II) and S(-II) decreased rapidly in time. Among them, the HA-n-FeS colloid removed part of the TBBPA through chemical adsorption. The main way of chemical adsorption was pore adsorption and functional group (olefin CC, phenolic hydroxyl group O-H, alcohol group C-O) combination. Besides, the HA-n-FeS colloid degraded part of the TBBPA into BPA through reduction, in which 17.72% of TBBPA was removed by the reduction of HA-n-FeS colloid. Fe(II) was the main contributor to the reductive degradation of TBBPA. Furthermore, active species (

Permanganate (Mn (VII)) is an environmentally-friendly mild oxidant in the field of advanced oxidation treatment, however, manganese colloids are produced as byproducts, which is difficult to separate from water, resulting in secondary pollution. This study used potassium methyl silicates (PMS) as surface modifiers to improve the aggregation of colloidal particles by increasing the hydrophobicity of the colloidal surface, and then explored the oxidation of bisphenol A (BPA) by Mn (VII) under the influence of potassium methyl silicate and the solid-liquid separation performance of the reaction system. The results showed that PMS and sodium silicate (SS) substantially enhanced the degradation of BPA by Mn (VII), and the promotion effect of potassium methyl silicate was greater than that of sodium silicate. PMS provided not only enough adsorption sites for MnO

Ferrous sulfide (FeS) and humic acid (HA) are typical black substances in black bloom water. Based on the strong reduction ability of FeS and the photosensitivity of HA, the transformation of toxic organic pollutants by the combination of FeS and HA (HA-FeS) is not clear. In order to explore this issue, the stability of HA-FeS was analyzed by measuring the hydrodynamic diameter and zeta potential of HA-FeS, and then the removal mechanism and possible degradation pathway of tetrabromobisphenol A (TBBPA) by HA-FeS under continuous illumination were discussed. The results showed that the hydrodynamic diameter of FeS was reduced and the stability of FeS was improved, and it was easily suspended after FeS combined with the HA in the water. The combination of HA and FeS promoted the removal of TBBPA in water, no matter it was in the presence or absence of light. Besides, compared with the absence of light, the removal efficiency of TBBPA was improved by HA-FeS with continuous light. There were two reasons for the increase in the removal efficiency of TBBPA by HA-FeS. On the one hand, Fe

Environment-friendly algaecides based on allelopathy have been widely used to control harmful algal blooms. In this research, micro nano scale artemisinin sustained-release algal inhibitor was prepared, the optimal preparation conditions were explored and the inhibitory mechanism of artemisinin algaecides was perfected. The results showed that when the particle size of artemisinin sustained-release microspheres (ASMs) was 2/10000 of artemisinin sustained-release granules (ASGs), the inhibitory effect was more remarkable. The optimal concentration of ASMs was 0.2 g L-1, and the inhibitory effect reached 99% on the 10th day; The algae density and chlorophyll-a both showed a downward trend, indicating that ASGs and ASMs could promote the degradation of chlorophyll-a; The inhibition rate of ASGs was faster than that of ASMs on the 4th day, and the inhibitory effect of ASMs was more significant after the 5th day. The activities of superoxide dismutase (SOD), peroxidase (POD), catalase (CAT) increased rapidly at first and then decreased, which indicated that ASGs and ASMs caused oxidative damage to Microcystis aeruginosa (M. aeruginosa) and inhibited the activity of antioxidant enzymes. Furthermore, the content of the oxygen free radical (O2-) and malondialdehyde (MDA) continued to rise after the 5th day, the protein, nucleic acid and conductivity in the culture medium increased. These results showed that lipid peroxidation occurred in the algal cell membrane, and the permeability of the membrane increased. In summary, the ASMs had significant continuous inhibitory effect while the ASGs had better short-term effect. The main inhibitory mechanism of artemisinin algaecides is the irreversible damage of cell membrane.

Environment-friendly algaecides based on allelopathy have been widely used to control harmful algal blooms. In this research, micro and nano scale artemisinin sustained-release algal inhibitor was prepared, the optimal preparation conditions were explored, and the inhibitory mechanism of artemisinin algaecides was turned perfect. The results showed that when the particle size of artemisinin sustained-release microspheres (ASMs) was 2/10,000 of artemisinin sustained-release granules (ASGs), the inhibitory effect was more remarkable. The optimal concentration of ASMs was 0.2 g L

Background Low coverage whole genome sequencing is a low-cost genotyping technology. Combining with genotype imputation approaches, it is likely to become a critical component of cost-efficient genomic selection programs in agricultural livestock. Here, we used the low-coverage sequence data of 617 Dezhou donkeys to investigate the performance of genotype imputation for low coverage whole genome sequence data and genomic selection based on the imputed genotype data. The specific aims were: (i) to measure the accuracy of genotype imputation under different sequencing depths, sample sizes, MAFs, and imputation pipelines; and (ii) to assess the accuracy of genomic selection under different marker densities derived from the imputed sequence data, different strategies for constructing the genomic relationship matrixes, and single- vs multi-trait models. Results We found that a high imputation accuracy (> 0.95) can be achieved for sequence data with sequencing depth as low as 1x and the number of sequenced individuals equal to 400. For genomic selection, the best performance was obtained by using a marker density of 410K and a G matrix constructed using marker dosage information. Multi-trait GBLUP performed better than single-trait GBLUP. Conclusions Our study demonstrates that low coverage whole genome sequencing would be a cost-effective method for genomic selection in Dezhou Donkey.

As the positioning accuracy of a visible light positioning (VLP) system is highly susceptible to changes in the orientation of the receiver, accurate knowledge of the receiver orientation is required. In practice, the orientation of the receiver is estimated with an external orientation estimation device. However, these devices generally suffer from drift and misalignment, causing an uncertainty in the measured orientation that will degrade the performance of standard positioning algorithms. In this paper, we derive a novel positioning algorithm that takes into account the effect of the orientation uncertainty. To this end, we need to cope with the non-linear relationship between the received signal strength (RSS) and the orientation uncertainty, which makes the likelihood function of the RSS, required to derive the maximum likelihood (ML) estimator, hard to obtain. To solve this issue, we consider the first and second-order Taylor series expansion of the RSS. Although the accuracy of the second-order approximation is better than the first-order approximation, the first-order approximation results in a closed-form expression for the likelihood function, while this is not possible with the second-order approximation. Because of this, we derive the ML estimator using the first-order approximation, and employ the multivariate gradient descent algorithm to obtain the position estimate. Computer simulations show that the proposed algorithm outperforms state-of-the-art VLP algorithms subject to orientation uncertainty. (C) 2019 Elsevier B.V. All rights reserved.

In this paper, we focus on the fundamental issues of non-orthogonal multiple access (NOMA) visible light communication (VLC) networks: achievable rates and optimal power allocation schemes for both static and mobile users. First, we derive both a lower bound and an upper bound of the achievable rates with closed-form expressions for static users in NOMA VLC networks. With the derived lower bound, we minimize transmit power under the minimum rate requirements and individual light emitting diodes (LED) power constraints, which turns out to be NP-hard. By exploiting the semidefinite relaxation (SDR) technique, the optimal power allocation scheme can be obtained by solving a convex semidefinite program (SDP). Second, we develop an optimal power allocation scheme for mobile users. Due to users’ movement, the estimated channel state information (CSI) may be inaccurate. We first characterize the CSI uncertainties as ellipsoidal regions, and derive a lower bound of the achievable rate expression. Then, we study the transmit power minimization problem for mobile users, which is non-convex. By applying $\mathcal {S}$ -lemma and SDR, the transmit power minimization problem can be reformulated as a convex SDP. Simulation results are presented to verify the effectiveness and robustness of the proposed power allocation schemes.

In this paper, we consider the two key problems in the physical-layer security of nonorthogonal multiple access (NOMA) visible light communication (VLC) networks: investigating a closed-form achievable security rate and studying the optimal security beamforming design. Specifically, under the dimming control, practical power, and successive interference cancellation constraints, we derive both the outer and inner bounds of the security capacity region with closed-form expressions, which are evaluated via numerical results. Then, based on the proposed security-rate expression, we investigated the optimal security beamforming design to minimize the total LED power, and to maximize the minimum secrecy rate, respectively. Both the problems are nonconvex. We apply different relaxation techniques to efficiently solve them. The simulation results demonstrate the efficacy of the proposed security beamforming design schemes in the NOMA VLC networks.

In this paper, we investigate the channel capacity region of interference channel and develop both centralized and distributed interference management schemes for visible light communication (VLC) networks. For a typical multiuser and multi-LED scenario, we derive both discrete inner and outer bounds of the channel capacity region, and such a proposed inner bound is numerically shown to be the highest among the existing inner bounds. Moreover, with continuous input signals, we develop the channel capacity region bounds in a closed form, termed $\left ({{\alpha,\beta,\gamma } }\right)$ (ABG) inner bound and ABG outer bound, which are tight for the large amplitude-to-variance ratio. Then, based on the derived ABG inner bounds, we investigate a centralized beamforming design problem to minimize the total transmit power under three practical constraints: peak optical power, average optical power, and average electrical power. By utilizing semidefinite relaxation technique, we reformulate this NP-hard problem as a convex semidefinite program and obtain the optimal beamformers. Furthermore, to reduce the cost of channel station information exchange, we propose a distributed coordinated interference management scheme by adopting the alternating direction method of multipliers method. Finally, numerical results are presented to evaluate the performance of the proposed interference management schemes in VLC networks.

In this paper, we investigate the design and implementation of machine learning (ML) based demodulation methods in the physical layer of visible light communication (VLC) systems. We build a flexible hardware prototype of an end-to-end VLC system, from which the received signals are collected as the real data. The dataset is available online, which contains eight types of modulated signals. Then, we propose three ML demodulators based on convolutional neural network (CNN), deep belief network (DBN), and adaptive boosting (AdaBoost), respectively. Specifically, the CNN based demodulator converts the modulated signals to images and recognizes the signals by the image classification. The proposed DBN based demodulator contains three restricted Boltzmann machines (RBMs) to extract the modulation features. The AdaBoost method includes a strong classifier that is constructed by the weak classifiers with the k-nearest neighbor (KNN) algorithm. These three demodulators are trained and tested by our online open dataset. Experimental results show that the demodulation accuracy of the three data-driven demodulators drops as the transmission distance increases. A higher modulation order negatively influences the accuracy for a given transmission distance. Among the three ML methods, the AdaBoost modulator achieves the best performance.

Sulfamerazine (SMR) as a persistent organic pollutant in waste streams is of growing environmental concern. This study explores the extraction SMR from water into an acetic acid (AA) solution using granular activated carbon (GAC), and removal of SMR by ozonation in AA solution. Systematic experiments have shown that GAC can be used as an adsorbent to transfer sulfamerazine from water to AA solution. SMR removal efficiency is 99.5% in 10% AA aqueous solution, which is better than in water. The removal rate of SMR in the AA solution decreased as the initial molar ratio of SMR and O3 increased. The removal rate of SMR decreased with Fe3+ present in the reactive system. The removal of SMR is dominated by indirect ozonation in water, while the SMR removal is an effect of both direct and indirect ozonation in AA solution. It is a very efficient process for the degradation of SMR in micro polluted water when using combined GAC adsorption–desorption in AA solution and ozonation of the resulting solution.