This paper considers secrecy enhancement mechanisms in visible light communication (VLC) systems with spatially distributed passive eavesdroppers (EDs) under the assumption that there are multiple LED transmitters and one legitimate receiver (UE). Based on certain amplitude constraints, we propose an optimal beamforming scheme to optimize secrecy performance. Contrary to the case where null-steering is made possible by using knowledge of the ED locations, we show that the optimal solution when only statistical information about ED locations is available directs the transmission along a particular eigenmode related to the intensity of the ED process and the intended channel. Then, a sub-optimal LED selection scheme is provided to reduce the secrecy outage probability (SOP). An approximate closed-form for the SOP is derived by using secrecy capacity bounds. All analysis is numerically verified by Monte Carlo simulations. The analysis shows that the optimal beamformer yields superior performance to LED selection. However, LED selection is still a highly efficient suboptimal scheme due to the complexity associated with the use of multiple transmitters in the full beamforming approach. These performance trends and exact relations between system parameters can be used to develop a secure VLC system in the presence of randomly distributed EDs. Index TermsPhysical layer security, visible light communication, beamforming, stochastic geometry, secrecy outage probability.

This paper considers physical layer security enhancement mechanisms that utilize simultaneous beamforming and jamming in visible light communication (VLC) systems with a randomly located eavesdropper under the assumption that there are multiple light-emitting diode (LED) transmitters and one intended user. When an eavesdropper with an augmented frontend receiver is present, the jamming is very useful for preventing the eavesdropper from wiretapping the information since it is not possible to extract only the information component from the received signal if the jamming signal is random. Thus, in this paper, an optimization problem is formulated with a focus on the signal-to-interference-plus-noise ratio for the legitimate link, and it is solved by a heuristic method called the concave-convex procedure. Then, a ternary scheme is proposed, which is less complicated than the full (joint) scheme, and it is optimized by adopting a formulation based on an assignment problem, the solution of which is effectively obtained by the so-called tabu search procedure. Additionally, the problem of maximizing the average secrecy rate is investigated by utilizing a continuous LED model, which significantly relaxes the complication that rises from calculating the expectation with respect to the location of the eavesdropper. Our analysis and simulation results show that the proposed simultaneous beamforming and jamming strategies (both joint and ternary) are good proxies for maximizing the average secrecy rate by utilizing the statistical information on the eavesdropper's random location.Index Terms-Physical layer security, visible light communication, beamforming, jamming, average secrecy rate. I. INTRODUCTIONOver the past decade, as the number of mobile devices connected to the Internet has increased, with primary user activities including data-intensive HD video streaming and cloud-based service access, the capacity demand on the radio access network has been steadily increasing. To satisfy this demand, wireless providers are deploying additional access infrastructures that rely on new cells and WiFi endpoints. However, it has proved challenging to improve data rate and reduce latency given the limited range of available radio frequency (RF) spectrum. Moreover, a large number of access points deployed in congested public areas cause high interference among themselves, which results in a degradation in the performance of the communication network [1]- [3].

Abstract-This paper considers secrecy enhancement mechanisms in visible light communication (VLC) systems with spatially distributed passive eavesdroppers (EDs) under the assumption that there are multiple LED transmitters and one legitimate user equipment (UE). Based on certain amplitude constraints, we propose a beamforming scheme to improve secrecy performance. Contrary to the case where null-steering is made possible by using knowledge of the ED locations, the proposed beamforming when only statistical information about ED locations is available directs the transmission along a particular eigenmode related to the intensity of the ED process and the intended channel. Then, a LED selection scheme that is less complicated than beamforming is provided to reduce the secrecy outage probability (SOP). An approximate closed-form for the SOP is derived by using secrecy rate bounds. All the analysis is numerically verified by MonteCarlo simulations. The analysis shows that the beamformer yields superior performance to LED selection. However, LED selection is still a highly efficient alternative scheme due to the complexity associated with the use of multiple transmitters in the full beamforming approach. These performance trends and exact relations between system parameters can be used to develop a secure VLC system in the presence of randomly distributed EDs.Index Terms-Physical layer security, visible light communication, beamforming, stochastic geometry, secrecy outage probability.

This paper proposes zero-forcing (ZF) beamforming strategies that can simultaneously deal with active and passive eavesdroppers in visible light communication (VLC) systems. First, we propose a ZF beamforming scheme that steers a transmission beam to the null space of active eavesdroppers' (AEDs) channel, while simultaneously considering the SNRs for a legitimate user (UE) and passive eavesdroppers (PEDs) residing at unknown locations. To find an eigenmode related to the optimal beamforming vector, we adopt an inverse free preconditioned Krylov subspace projection method. For unfavorable VLC secrecy environments, the proposed ZF beamformer appears to be incapable of effectively coping with the PEDs due to the strict condition that the data transmission must be in the null space of the AEDs' channel matrix. Hence, an alternative beamforming scheme is proposed by relaxing the constraint on the SNRs of the AEDs. The related optimization problem is formulated to reduce the secrecy outages caused by PEDs, while simultaneously satisfying the target constraints on the SNRs of the UE and the AEDs. To simplify the mathematical complexity of the approach, Lloyd's algorithm is employed to sample the SNR field, which in turn discretizes the problem, thus making it tractable for practical implementation. The numerical results show that both the exact and relaxed ZF beamforming methods achieve superior performance in the sense of secrecy outage relative to a benchmark ZF scheme. Moreover, the proposed relaxed ZF beamforming method is shown to cope with PEDs better than the exact ZF beamforming approach for unfavorable VLC environments.

This paper proposes a novel cooperative beamforming and jamming scheme to deal with passive and active eavesdroppers (EDs) in indoor visible light communication (VLC) networks. An ED in VLC systems can augment its front-end receiver by implementing possible device modifications; thus, jamming is very useful for curbing such an enhanced ED since it would be impossible to distinguish between the information and jamming signals. In contrast to the traditional artificial noise strategies for VLC that can only deal with either passive or active EDs, we propose a combined scheme of beamforming and jamming that significantly improves secrecy performance when both types of EDs exist. The proposed scheme is designed to maximize the signal-to-interference-plus-noise ratio (SINR) of the legitimate receiver, entirely suppress the SINRs of the active EDs, and restrict the average SINR of the passive EDs. We apply an inverse free preconditioned Krylov subspace projection method and the convex-concave procedure to obtain the suboptimal beamforming weight and jamming intensity vectors. Also, an optimal power splitter coefficient is found through the golden section search method. The numerical results verify that the proposed scheme shows superior performance compared to the three benchmarks: zero-forcing beamforming, artificial noise scheme, and enhanced zero-forcing beamforming.

This article highlights challenges associated with securing visible light communication (VLC) systems by using physical layer security (PLS) techniques. Motivated by the achievements in PLS studies for radio frequency (RF) communication, many PLS techniques for VLC systems were also rigorously investigated by tailoring the RF techniques to the VLC environment. However, careful consideration of the inherent differences between RF and VLC systems is still needed. By disregarding these differences, an eavesdropper could be given an opportunity to wiretap the VLC systems, even when PLS techniques are employed to protect them. Crucially, the fact that it is often not possible to know the number and locations of eavesdroppers in real VLC systems may allow eavesdroppers to devise various cooperative eavesdropping methods. By examining a few examples of the possible eavesdropper threats that can occur in VLC systems, this article offers novel insights into the vulnerabilities of state-of-the-art PLS schemes for VLC systems. Although the focus of the paper is mostly on these weaknesses, some potential solutions are also briefly proposed with a view to stimulating discourse in the community.

This paper studies a novel electrophoretic molecular communication (EMC) framework utilizing a piecewise constant electric field. EMC is a particular type of molecular communication that exploits electric fields to induce the movement of charged particles to enhance communication performance. Our previous work proposed an EMC framework utilizing a time-varying electric field that exponentially changes; however, the field with such a complicated shape might be challenging to be implemented in practice. Thus, this paper proposes a new EMC approach exploiting a piecewise constant electric field that can be readily implemented via, e.g., an on/off switch method. We formulate two optimization problems to design the electric field based on different objectives: minimizing a mean squared error and minimizing a bit interval. The solutions of each, such as optimal on-off timings and corresponding strengths of the constant electric fields, are obtained through the Lagrange multiplier approach and the geometric programming, respectively. The Monte Carlo simulation results verify that the proposed piecewise constant electric field significantly reduces the bit error rate relative to the constant field benchmark while performing less well, but not significantly, than the exponential field benchmark.

This paper investigates electrophoretic molecular communication (EMC) operating in circular duct channels. EMC utilizes the time-varying electrophoretic force that can controllably induce the movement of charged particles to enhance communication performance. In circular duct channels, where the memory component is high, intersymbol interference (ISI) must be considered. Thus, this paper presents a method to design an electric field under the framework of the calculus of variations, simultaneously reducing the ISI and strengthening the information signal reception. The numerical results show that the proposed electric field can significantly reduce the bit error rate compared to the constant field benchmark.