Title: EAGER: Towards Securing Visible Light Communications, funded by NSF under grant CNS-1745254, 09/01/2017 – 08/31/2020.

Project Summary:

Due to the many nice features of visible light communication (VLC), such as license-free spectrum, abundant available bandwidth, high energy efficiency, and supporting extremely high (Gbps-level) transmission rate and dense spatial reuse of the spectrum, VLC has been considered to be a promising and urgently-needed small-cell solution for alleviating the RF spectrum scarcity in the 5G era. While the research on VLC devices has made significant progress in recent years, the security aspect of VLC has not been well understood so far. Contrary to the initial belief that VLC is intrinsically secure because the propagation of visible light is directive and can be confined within a closed space, recent studies have revealed that VLC is subject to eavesdroppers that are outside of the direct beam of the light, and even to eavesdroppers that are outside of the space and do not have direct line-of-sight (LOS) to the light source. Even for indoor eavesdroppers, because the light illumination that the VLC piggybacks on is diffusive in most real-world applications, an unintended receiver may easily receive the communication without being noticed – a “what you see is what you send” feature of VLC. This makes VLC even more vulnerable to eavesdropping in reality than its RF counterparts. In addition to eavesdropping, the special optical nature of visible light, such as being impermeable to obstacles, also subjects VLC to other types of attacks, including blocking and spoofing. Altogether, these attacks constitute realistic threats to many VLC applications, such as VLC access networks in indoor public space (library, open office, etc.), fixed-communication VLC links in a data center, and VLC positioning and sensing, etc. Today’s limited understanding on VLC vulnerabilities and their countermeasure may lead to the dangerous “zero-day attacks” issue when the technology is deployed in large scale in the near future.

Figure 1. Types of VLC attacks

Project Goals:

The overarching goal of this EAGER exploratory project is to obtain a comprehensive understanding on the security vulnerabilities of visible light communication (VLC), and to develop a solid mathematical framework that can be used to investigate and develop rigid and provably-secure countermeasures to these vulnerabilities. In particular, the project includes the following major goals: 

· Study of eavesdropping for VLC channels: The project will investigate performance metrics related to eavesdropping attacks on a VLC channel and will propose a friendly-jamming-assisted visible light MIMO beamforming architecture to counter such attacks from both inside and outside the room where the VLC channel resides. Compared with existing methods that can only counter known eavesdroppers, the proposed approach protects VLC against eavesdroppers of an unknown number and at unknown locations. 

· Study of blocking and spoofing attacks for VLC: The project will propose methods to effectively detect and mitigate blocking and spoofing attacks targeting visible light communications (VLC) and visible light sensing (VLS). The PI will also study the blocking-resilience capacity of the multi-user VLC MIMO system to provide differential protection for users of different priorities.

· A MIMO VLC testbed will be built to evaluate the performance of the proposed solutions.

Project Personnel

PI: Tao Shu, Ph.D.

Graduate Students

· Jian Chen

· Jing Hou

· Li Sun

· Xueyang Hu

Project Activities and Results

1. Statistical modeling and analysis on the confidentiality of indoor VLC systems

While visible light communication (VLC) is expected to have a wide range of applications in the near future, the security vulnerabilities of this technology have not been well understood so far. In particular, due to the extremely short wavelength of visible light, the VLC channel presents several unique characteristics than its radio frequency (RF) counterparts, which impose new features on the VLC security. Taking a physical-layer security perspective, this research studies the intrinsic secrecy capacity of VLC as induced by its special channel characteristics. Different from existing models that only consider the specular reflection in the VLC channel, a modified Monte Carlo ray tracing model is proposed to account for both the specular and the diffusive reflections, which is unique to VLC. A deep neural network model is also proposed to describe the spatial VLC channel response based on a limited number of channel response samples calculated from the ray tracing model. Based on these models the upper and the lower bounds of the VLC secrecy capacity are derived, which allow us to evaluate the VLC communication confidentiality against a comprehensive set of factors, including the locations of the transmitter, receiver, and eavesdropper, the VLC channel bandwidth, the ratio between the specular and diffusive reflections, and the reflection coefficient. Our results reveal that due to the different types of reflections, the VLC system becomes more vulnerable at specific locations where strong reflections exist.

The main outcomes of this research activity include the following:

· A modified Monte Carlo ray tracing method is proposed to account for both the specular and diffusive reflections in calculating VLC channel impulse response at a given location.

· A deep neural network (DNN) regression model is proposed to efficiently estimate the VLC channel impulse response as a function of the VLC link location in the communication space based on the training data set of a limited number of channel response samples calculated according to the ray tracing model.

· Based on these models, the upper bound and the lower bound of the VLC secrecy capacity are calculated considering multiple reflections under specific conditions.

· Leveraging the secrecy capacity bounds, we depict the spatial characteristics/distribution of the VLC secrecy capacity over given indoor communication space. We also study how the multiple types of reflections affect VLC secrecy capacity against a comprehensive set of factors, including the locations of the VLC transmitter, receiver, and eavesdropper, the VLC channel bandwidth, the ratio between the specular and the diffusive reflections, and the reflection coefficient.

Two papers have been published as a result of this research activity:

· Jian Chen and Tao Shu, “Statistical modeling and analysis on the confidentiality of indoor VLC systems,” IEEE Transactions on Wireless Communications, vol. 19, no. 7, pp. 4744-4757, July 2020.

· Jian Chen and Tao Shu, “Impact of multiple reflections on secrecy capacity of indoor VLC systems,” Proc. 2019 International Conference on Information and Communication Security, pp. 105-123, Dec. 2019.

Some of our results in this research have been presented as a poster “Indoor VLC fingerprinting and its application to data sniffing and anti-light spoofing” in the 2017 Auburn University Wireless Engineering Research and Education Center Advisory Board Meeting in Nov. 2017.

2.  VL-Watchdog: Spoofing detection for indoor visible light systems

Thanks to the license-free visible light (VL) spectrum and the pervasive availability of light fixtures in almost all indoor environment, visible light communication (VLC) and visible light sensing (VLS) have received an increasing amount of interest in recent years as a promising solution to offloading the crowded RF traffic. As more and more VLC and VLS systems are mounted on today’s light fixtures, how to guarantee the authenticity of the VL signal in these systems becomes an urgent problem. This is because almost all of today’s light fixtures are unprotected and can be openly accessed by almost anyone, and hence are subject to tamperment and substitution attacks. an attacker can easily replace an authentic LED by a rogue LED under his control to inject spoofed VL signal into user’s receiver. Unfortunately, most of today’s VLS applications do not have a reliable built-in signal authentication mechanism to detect these spoofed signals and hence will mistakenly accept them as authentic sensing inputs, leading to compromised sensing outcome. Similar situation also arises in VLC. As such, ensuring received VL signals are coming from the authentic transmitters (LEDs), rather than from a spoofer, is the key in ensuring the quality and correctness of the VL communication and sensing outcomes. Existing cryptography-based authentication methods are mainly applicable to data applications (i.e., frames of “0” and “1” bits), and is not directly applicable at the signal level where most of the VLS applications operate and there is no “0” and “1” bits.

In this research, by exploiting the intrinsic linear superposition characteristics of visible light, we propose VL-Watchdog, a scalable and always-on signal-level spoofing detection framework that is applicable to both VLC and VLS systems. VL-Watchdog is based on redundant orthogonal encoding of the transmitted visible light, and can be implemented as a small hardware add-on to an existing VL system. A proof-of-concept testbed was developed to verify the feasibility of VL-Watchdog. A theoretical model is also proposed to analyze the spoofing detection accuracy of VL-Watchdog under various attack and noise conditions. The effectiveness of the proposed framework was validated through extensive numerical evaluations against a comprehensive set of factors.

The main outcomes of this research activity include the following:

· An orthogonal coding based signal-level always-on VL spoofing detection framework, VL-Watchdog, is proposed.

· A theoretical model is also proposed to analyze the spoofing detection accuracy of VL-Watchdog under various attack and noise conditions. Its optimal detection threshold is also derived by analysis.

· A false-warning filter is proposed to improve VL-Watchdog’s detection accuracy by accounting for random light perturbations caused by human activities and environmental changes in realistic application scenarios.

· A proof-of-concept testbed is developed to verify the feasibility of VL-Watchdog.

· The performance of VL-Watchdog is evaluated based on extensive numerical simulations by taking into account a comprehensive set of parameters, including the number of orthogonal coding basis, the spoofing power to noise ratio, spoofing detection window size, the spoofer’s strategies in fabricating its spoofing signals, and random perturbations from the application environment.

A conference paper describing the above research has been submitted to IEEE ICC 2021, and a journal paper will be submitted soon.

· Jian Chen and Tao Shu, “Spoofing detection for indoor visible light systems with redundant orthogonal encoding (6-page conference version),” submitted to IEEE ICC 2021, under review, Oct. 2020.

· Jian Chen and Tao Shu, “VL-Watchdog: spoofing detection for indoor visible light systems with redundant orthogonal encoding (12-page journal version),” to be submitted soon.

3. Study of the value of traded target information in security games

This research activity is related to both Goal 1 and Goal 2 of the project in that it investigates the fundamental theoretical problem of how much the target information is worth (i.e., the value of target information), which underlays all attack models. Ample evidence has confirmed the importance of information in security. While much research on security game has assumed the attackers’ limited capabilities to obtain target information, few work considers the possibility that the information can be acquired from a data broker, not to mention exploring the attackers’ profit-seeking behaviors in the shrouded underground society. This work studies the role of information in security problem when the target information is sold by a data broker to multiple attackers. We formulate a novel multi-stage game model to characterize both the cooperative and competitive interactions of the data broker and attackers. Specifically, the attacker competition with correlated purchasing and attacking decisions is modeled as a two-stage stochastic model; and the bargaining process between the data broker and the attackers is analyzed in a Stackelberg game. Both the attackers’ competitive equilibrium solutions and data broker’s optimal pricing strategy are obtained. This study generates new knowledge on the value of target information and contributes to the literature by characterizing the behaviors of the attackers with labor specialization, and providing quantitative measures of information value from an economic perspective.

A conference paper documenting the above research has been published in SecureComm 2019, and a journal paper has been submitted to IEEE/ACM Transactions on Networking and is under major revision:

· Jing Hou, Li Sun, Tao Shu, and Husheng Li, “Target information trading – An economic perspective of security,” Proc. 15th EAI International Conference on Security and Privacy in Communication Systems (SecureComm 2019), vol. 2, pp. 126-145, Oct. 2019.

· Jing Hou, Li Sun, Tao Shu, and Husheng Li, “The value of traded target information,” submitted to IEEE/ACM Transactions on Networking (ToN), major revision, under review, Nov. 2020.

Broader Impacts

Lack of spectrum resource is the common challenge confronting all RF communication systems today, which are serving almost every citizen of the nation. VLC provides a promising solution for wireless traffic offloading, and can significantly alleviate the RF spectrum scarcity. However, the security aspect of VLC has not been well understood so far, which constitutes a serious threat to the technology when it is deployed in large scale in the near future. If successful, this project will provide the much needed understanding on the security vulnerabilities of VLC, and will also equip us with the new knowledge of countermeasures to these vulnerabilities. The outcome of this project will protect the interest of every wireless user in the new era of broadband wireless, and thus will generate deep impacts on the nation’s economy and social wellbeing. The project will also carry out a comprehensive education plan to broaden its impact to the society, including integrating research findings with undergraduate and graduate courses, recruiting and outreaching to minority and under-represented students, disseminating research findings through open access, and open-lab days.

The following activities have been taken to broaden the impacts of this project so far:

· The PI has presented the research outcomes in the 2017 Auburn University Wireless Engineering Research and Education Center Advisory Board Meeting in Nov. 2017.

· The PI has integrated part of the research outcomes in the course materials he is teaching at Auburn University, including COMP 4320 (Introduction to Computer Networks), COMP 5320/6320/6326 (Design and Analysis of Computer Networks), and COMP 7370/7376 (Advanced Computer and Network Security). 

· This project was also introduced to over 1000 high-school students and their parents during the 2018 Open House Engineering Day (E-day) of the Samuel Ginn College of Engineering at Auburn University. This helps to foster the high-school students' interests in taking science and technology as their future career.

· This project was introduced to over 1700 high-school students and their parents during the 2020 Open House Engineering Day (E-day) of the Samuel Ginn College of Engineering at Auburn University. This helps to foster the high-school students' interests in taking science and technology as their future career.

· One of the Ph.D. students supported by this project is a female. Therefore this project has helped to increase the diversity in the STEM areas and promote women in engineering.