With the uninterrupted revolution of communications technologies and the great-leap-forward development of emerging applications, the ubiquitous deployment of Internet of Things (IoT) is imperative to accommodate constantly growing user demands and market scales. Communication security is critically important for the operations of IoT. Among the communication security provisioning techniques, physical layer security (PLS), which can provide unbreakable, provable, and quantifiable secrecy from an information-theoretical point of view, has drawn considerable attention from both the academia and the industries. However, the unique features of IoT, such as low-cost, wide-range coverage, massive connection, and diversified services, impose great challenges for the PLS protocol design in IoT. In this article, we present a comprehensive review of the PLS techniques toward IoT applications. The basic principle of PLS is first briefly introduced, followed by the survey of the existing PLS techniques. Afterwards, the characteristics of IoT are identified, based on which the challenges faced by PLS protocol design are summarized. Then, three newly-proposed PLS solutions are highlighted, which match the features of IoT well and are expected to be applied in the near future. Finally, we conclude the paper and point out some further research directions.

The Viterbi algorithm is widely adopted in digital communication systems because of its capability of realizing maximum-likelihood signal sequence detection. However, implementation of the Viterbi algorithm requires instantaneous channel state information (CSI) to be available at the receiver. This is difficult to satisfy in some emerging communication systems such as molecular communications, underwater optical communications, etc, where the underlying channel models are highly complex or completely unknown. ViterbiNet, developed in the prior literature, is a promising framework to cope with this challenge, where deep learning (DL) techniques are combined with the Viterbi Algorithm to enable near-optimal signal detection without CSI. This paper offers a non-trivial variation of ViterbiNet based on generative adversarial networks (GAN). Specifically, a novel architecture using GAN is designed to directly learn the channel transition probability (CTP) from receiver observations, which is the only part of the Viterbi algorithm that is channel-dependent. With the learned CTP, the classical Viterbi algorithm can be implemented without modifications. To make the proposed architecture applicable to time-varying channels, we further develop two methods to fine-tune the learned CTP online. In the first method, pilots within each frame are exploited to update the CTP learning network; In the second method, a decision-directed approach is devised to generate training data in real-time, which is utilized to re-train the learning network. By combining these two approaches, the receiver is able to track the dynamic channel conditions without being trained from scratch. Numerical simulations demonstrate the superiority of the proposed design compared to existing methods.

This paper develops a fountain-coding-based scheme to secure communications over fading wiretap channels, in which the transmission between the legitimate nodes is overheard by a malicious eavesdropper. The transmitter first splits its message into K source packets and generates a potentially infinite number of fountain-coded packets, each of which is the XOR of distinct source packets chosen according to the channel feedback. Based on the characteristics of fountain-coded transmissions, a sufficient number of independent coded packets have to be successfully received to recover the original source message. Therefore, secrecy is guaranteed if the legitimate receiver can accumulate the required number of coded packets before the eavesdropper does. To realize this, we propose to exploit outage prediction and channel feedback to optimize the fountain encoding procedure. To be more specific, upon the reception of any packet, the receiver predicts the conditional outage probability of the next slot's transmission based on the current channel state and notifies the transmitter to adjust the structure of the fountain code accordingly. In this manner, the encoder design matches well with the time-varying channel conditions of the legitimate link such that the receiver can achieve a high accumulation rate for the fountain packets. Meanwhile, because of the mismatch between the fountain encoder and the eavesdropper's channel, it is extremely difficult for the eavesdropper to collect enough independent fountain-coded packets, and transmission secrecy is thus guaranteed. Simulation results demonstrate that the proposed scheme outperforms the existing candidate solutions in terms of the intercept probability, the decoding delay at the legitimate receiver, and the achievable throughput, as well as the quality-of-service violating probability.

This paper proposes a novel technique called constellation overlapping to secure two-way untrusted relaying systems. We first consider the symmetric case where two terminal users adopt the same modulation type. For this case, a truncated-channel-inversion-based approach is developed to make the signals from different users experience the same equivalent channel, thereby realizing full constellation overlapping at the relay. Consequently, an error floor is created at the untrusted relay, and data confidentiality is protected. Then, the asymmetric case is investigated, where different modulations are employed at the two users and the constellation overlapping cannot be realized. To address this issue, we develop a constellation virtualization method, which exploits the intrinsic user-cooperation mechanism to “virtualize” two identical constellations at the two users. In this manner, the constellation overlapping effect can still be achieved, without sacrificing much reliability performance. We optimize the truncation threshold to maximize the sum rate for end-to-end information exchange, and design the bit-to-symbol mapping criterion and the signal detection method as well. Unlike the existing physical layer security approaches that guarantee secrecy with additional resource consumptions, this method realizes security enhancement by letting different users mutually encrypt their signals over the air, which is more resource-efficient.

This paper studies the privacy-preserving for downlink transmission in multiuser relay networks, where a source communicates with multiple users via a relay employing the amplify-and-forward protocol. Within any scheduling unit, only one user (desired user) is chosen for data reception, and the other users (undesired users) are viewed as potential eavesdroppers due to the broadcast nature of wireless medium. To prevent information leakage, we propose a physical-layer cooperative privacy preserving scheme, whose key idea is to schedule a cooperating user in addition to the desired user to deliver artificial noise (AN). By exploiting the characteristics of channels, the cooperating user carefully designs the AN transmitted during two time slots such that the AN can be canceled out at the desired user, but cannot be removed at the undesired users. As a result, the end-to-end signal-to-noise-ratio of any undesired user is heavily degraded, while that of the desired user is not seriously affected, thus preserving the data confidentiality of the desired user. To maximize the instantaneous secrecy rate, an opportunistic user selection criterion is developed. The lower bound of the ergodic secrecy rate (ESR) as well as the approximate upper bound of the secrecy outage probability is derived. The asymptotic performance of ESR is also analyzed via extreme value theory. Furthermore, to motivate users with heterogeneous channel conditions to participate in cooperation and guarantee the fairness among users, a user-grouping-based selection method is proposed. To evaluate the performance of this method, a novel concept called system fairness factor is introduced and studied. Theoretical analysis and simulation results show that, thanks to the proposed cooperative privacy preserving mechanism, the system ESR grows with the increasing number of users, and much higher secrecy rate and lower secrecy outage probability can be achieved compared with the existing schemes in the literature.

This letter studies the secure transmission protocol design for two-way cooperative systems, where two terminal users communicate via an untrusted relay. The relay acts as both a helper and a potential eavesdropper, from which the users' messages need to be kept secret. A constellation-rotation aided approach is proposed to prevent the relay from decoding these messages. By rotating the signal constellation with a proper angle, the complex-valued symbol can be fully represented by its real component. Then, either user utilizes only one dimension of the signal to deliver its information, with the other dimension carrying the artificial noise (AN). In this manner, the received signal-to-interference-plus-noise ratio (SINR) at the relay is degraded due to the injection of AN, thereby increasing the difficulty of eavesdropping. Moreover, the desired signal and the AN lie in different directions at the terminal users, which makes the signal detection free-of-interference. The approximate expression for the upper bound of the users' symbol error rate (SER) is derived based on which a criterion is developed to choose the rotation angle that can optimize the system SER performance. Simulation results demonstrate the superiority of our scheme compared to the competing candidates.