This paper studies the secrecy performance of light-fidelity (LiFi) networks under the consideration of random device orientation and partial knowledge of the eavesdroppers’ channel state information. Particularly, the secrecy capacity and secrecy outage probability are analysed for the case of a single eavesdropper, as well as for the case of multiple eavesdroppers. Moreover, a machine learning based access point (AP) selection algorithm is presented with the objective of maximising the secrecy capacity of legitimate users. Our results show that optimising the AP selection while taking into account the random behaviour of the optical channel results in a significant enhancement in the achievable secrecy performance. In fact, using the derived realistic secrecy expressions as the basis for AP selection results in up to 30% secrecy capacity enhancement compared to the limited assumption of fixed orientation.
This paper studies the secrecy performance of light-fidelity (LiFi) networks under the consideration of random device orientation and partial knowledge of the eavesdroppers’ channel state information. Particularly, the secrecy capacity and secrecy outage probability are analysed for the case of a single eavesdropper, as well as for the case of multiple eavesdroppers. Moreover, a machine learning based access point (AP) selection algorithm is presented with the objective of maximising the secrecy capacity of legitimate users. Our results show that optimising the AP selection while taking into account the random behaviour of the optical channel results in a significant enhancement in the achievable secrecy performance. In fact, using the derived realistic secrecy expressions as the basis for AP selection results in up to 30% secrecy capacity enhancement compared to the limited assumption of fixed orientation.
Radio frequency (RF) signals propagate through most materials that we are surrounded by while light is blocked by many of these materials. This feature makes wireless networks based on light (which are also referred to as LiFi networks) inherently more secure. However, it can also lead to sudden link failure if the legitimate data link is blocked because of user movements or changes in device orientation. In this paper, the secrecy capacity has been analysed with the consideration of imperfect channel state information, random device orientation and probability of link blockage for the case of a single eavesdropper. It has been found that the secrecy capacity almost doubles in a standing activity as opposed to a sitting activity and that the density of blocking objects degrades the secrecy capacity in single access point networks. It is evident that environmental factors and user behaviour have a significant impact on the secrecy performance and, thus, need to be considered for robust physical layer security (PLS) design in LiFi networks.
In this paper, we study the physical-layer security in a 3-D multiuser visible light communication (VLC) network. The locations of access points (APs) and mobile users are modeled as two 2-D, independent and homogeneous Poisson point processes at distinct heights. Using mathematical tools from stochastic geometry, we provide a new analytical framework to characterize the secrecy performance in multiuser VLC networks. Closed-form results for the outage probability and the ergodic secrecy rate are derived for networks without AP cooperation. Considering the cooperation among APs, we give tight lower and upper bounds on the secrecy outage probability and the ergodic secrecy rate. To further enhance the secrecy performance at the legitimate user, a disk-shaped secrecy protected zone is implemented in the vicinity of the transmit AP. Based on the obtained results, it is shown that cooperating neighboring APs in a multiuser VLC network can bring performance gains on the secrecy rate, but only to a limited extent. We also show that building an eavesdropper-free protected zone around the AP significantly improves the secrecy performance of legitimate users, which appears to be a promising solution for the design of multiuser VLC networks with high security requirements.
Abstract This paper investigates the physical layer security of indoor visible light communication (VLC) systems in the presence of randomly located eavesdroppers using a stochastic geometry framework. Unlike traditional LOS-only or perfect-CSI assumptions, this work incorporates a more realistic setting where the eavesdropper’s channel state information (CSI) is unknown, and both line-of-sight (LOS) and first-order non-line-of-sight (NLOS) reflections are accounted for in the channel modelling. The analysis further integrates device orientation randomness and evaluates security under eavesdroppers employing selection combining (SC) and maximum ratio combining (MRC). Closed-form expressions for the channel gain, SNR distribution, and secrecy outage probability (SOP) are derived by exploiting the spatial statistics of a bounded Poisson point process (PPP). Results show that incorporating NLOS components significantly enhances legitimate SNR and improves secrecy performance, while eavesdropper collaboration using maximum ratio combining presents the worst-case security threat. Monte Carlo simulations validate the analytical derivations, demonstrating close agreement across various system parameters. The study offers an accurate and practical secrecy analysis framework for next-generation VLC deployments in realistic indoor environments.
In order to improve the security performance of multiuser visible light communication (VLC) and facilitate the secure application of optical wireless communication technology in Internet-of-Things, we investigate the physical-layer security in a multiuser VLC system with non-orthogonal multiple access (NOMA). When the light-emitting diode (LED) transmitter communicates with multiple legitimate users by downlink NOMA, both single eavesdropper and multi-eavesdropper scenarios are considered. In the presence of single eavesdropper, based on transmission characteristics of the optical wireless channel, with known instantaneous channel state information (CSI) of the NOMA legitimate channels and statistical CSI of the eavesdropper channel, an exact expression of secrecy outage probability (SOP) is derived, which acts as a benchmark of the security performance to guide selecting or optimizing parameters of the LED transmitter and the photodiode receiver of NOMA legitimate users. In the multi-eavesdropper case, based on the spatial distribution of legitimate users and eavesdroppers, the SOP is obtained via a stochastic geometry theory, so as to guide the NOMA legitimate users to keep away from the area with high eavesdropper density. For typical parameters of the indoor LED transmitter and the PD receiver, simulation results show that the SOP performance improves with the increasing of LED transmission power or transmission signal-to-noise ratio (SNR) in both scenarios. Specifically, in the single eavesdropper case, enlarging the channel condition difference of user groups or deviating the eavesdropper from the given user group can improve the SOP performance, and for a given NOMA legitimate user, the SOP eventually settles around 0.2 while the semiangle at half illuminance of the LED varies between 15° to 60°. In the multi-eavesdropper case, we can get a better SOP performance when reducing the eavesdropper density or the semi-angle at half illuminance of the LED for a given eavesdropper density.
In this paper, we analyze the outage probability and secrecy capacity of a non-orthogonal multiple access (NOMA) system in the presence of an eavesdropper. In order to enhance spectral efficiency, a base station communicates with two users simultaneously in the same frequency band by superpimposing the transmit signals to the users in the power domain. Specifically, the user with the worse channel conditions is allocated higher power such that it is able to directly decode its signal from the received superimposed signal. At the user with the better channel conditions, the interference due to NOMA is processed by successive interference cancelation. Given these system settings and accounting for decoding thresholds, we analyze the outage probability of the NOMA system over Rayleigh fading channels. Furthermore, based on the locations of the users and eavesdropper, the secrecy capacity is analyzed to assess the level of security provided to the legitimate users in the presence of an eavesdropper. Here, the decoding thresholds of legitimate users and eavesdropper are also included in the analysis of the secrecy capacity. Through numerical results, the effects of network parameters on system performance are assessed as well as the the superiority of NOMA in terms of secrecy capacity over traditional orthogonal multiple access.
This paper investigates the impact on the achievable secrecy performance of multiple-input multiple-output systems by dealing with realistic propagation conditions. Specifically, we focus on the κ-μ shadowing fading model, which has proven to be more accurate in characterizing mm-wave scenarios than traditional Rice and Rayleigh ones. By considering transmit antenna selection and maximal ratio combining at the receiver ends, we study two different scenarios, namely: (i) the transmitter has knowledge of the channel state information (CSI) of the eavesdropper link, and (ii) the transmitter does not know eavesdropper's CSI. Based on these assumptions, we derive novel analytical expressions for the secrecy outage probability (SOP) and the average secrecy capacity (ASC) to assess the secrecy performance in passive and active eavesdropping scenarios, respectively. Moreover, we develop analytical asymptotic expressions of the SOP and ASC at high signal-to-noise ratio regime. Some useful insights on how to obtain noticeable improvements in the secrecy performance are also provided.