We propose a new secure transmission scheme in the multi-input multi-output multi-eavesdropper wiretap channel. In this channel, the N _A -antenna transmitter adopts transmit antenna selection (TAS) to choose the antenna that maximizes the instantaneous signal-to-noise ratio (SNR) at the receiver to transmit, while the NB-antenna receiver and the N _E -antenna eavesdropper adopt maximal-ratio combining (MRC) to combine the received signals. We focus on the practical scenario where the channel state information (CSI) during the TAS process is outdated. In this scenario, we propose a new transmission scheme to prevent the detrimental effect of the outdated CSI on the wiretap codes design at the transmitter. To thoroughly assess the secrecy performance achieved by the proposed scheme, we derive new closed-form expressions for the exact secrecy outage probability and the probability of non-zero secrecy capacity for arbitrary SNRs. We also derive new compact expressions for the asymptotic secrecy outage probability at high SNRs. Notably, in the analysis, we take spatial correlation at the receiver into consideration. Apart from the advantage of our scheme over the conventional TAS/MRC scheme, we demonstrate that the outdated TAS reduces the secrecy diversity order from N _A N _B to N _B . We also demonstrate that antenna correlation improves the secrecy performance at low SNR but deteriorates the secrecy performance at medium and high SNRs, by affecting the secrecy array gain only.

We design new secure on-off transmission schemes in wiretap channels with outdated channel state information (CSI). In our design, we consider not only the outdated CSI from the legitimate receiver but two distinct scenarios as well, depending on whether the outdated CSI from the eavesdropper is known at the transmitter. Under this consideration, our schemes exploit useful knowledge contained in the available outdated CSI, and based on this, the transmitter decides on whether to transmit or not. We derive new closed-form expressions for the transmission probability, the connection outage probability, the secrecy outage probability, and the reliable and secure transmission probability to characterize the achievable performance. Based on these results, we present the optimal solutions that maximize the secrecy throughput under dual connection and secrecy outage constraints. Our analytical and numerical results offer detailed insights into the design of the wiretap coding parameters and the imposed outage constraints. We further show that allowing more freedom on the codeword transmission rate enables a larger feasible region of the dual outage constraints by exploiting the tradeoff between reliability and security.

We design a novel artificial-noise-aided secure ON-OFF transmission scheme in a wiretap channel. We consider a practical scenario, where the multi-antenna transmitter only obtains partial channel knowledge from the single-antenna receiver through limited training and feedback but has no channel knowledge about the single-antenna eavesdropper. In the design, we first propose a three-period block transmission protocol to capture the practical training and quantization features. We then characterize the statistics of the received signal-to-noise ratios at the receiver and the eavesdropper. Under the secrecy outage constraint, we exploit the ON-OFF scheme to perform secure transmission and derive a closed-form expression for the secrecy throughput. Moreover, we investigate the optimization problem of maximizing the secrecy throughput by proposing an iterative algorithm to determine the optimal power allocation between the information signal and artificial noise, as well as the optimal codeword transmission rate. Furthermore, we define the net secrecy throughput (NST), which takes the signaling overhead into account and address the problem of optimally allocating the block resource to the training and feedback overhead. Numerical results clearly demonstrate how the optimal signaling overhead changes with the number of transmit antennas, and there exists an optimal number of antennas that maximizes the NST.

Physical layer security is a promising way to secure the wireless communications in the Internet of Things. Motivated by the fact that the limited feedback resources in the IoT network would degrade the secrecy advantage of the multiple-antenna technique, we attempt to investigate the problem of how many transmit antennas should be utilized to perform secure communications. In particular, we consider the heterogeneous IoT downlink network and design a multiuser secure transmission scheme. In this scheme, the zero-forcing beamforming technique is adopted to serve the IoT legitimate users, and the remaining spatial freedoms are utilized to send artificial noise (AN) for confusing the passive eavesdroppers. Given the secrecy outage constraints, we derive the closed-form expression for the network secrecy throughput and formulate a non-convex optimization problem with multiple parameters, e.g., the number of transmit antennas, the wiretap codes, the feedback bits allocation strategy, and the power allocation ratio between the information bearing signal and the AN. To effectively tackle this problem, we develop an optimization framework involving the block coordinate descent (BCD) algorithm and the one-dimensional search method. Simulation results validate the effectiveness of our proposed optimization framework and show that the optimal number of transmit antennas increases as the secrecy outage constraints become stricter, or the feedback resources become scarcer.

In this paper, we consider the secure communications in a cognitive untrusted relay network, where the secondary source intends to communicate with a secondary destination through an untrusted relay in the presence of the direct source-destination link. Specifically, we first examine the connection outage probability (COP) and secrecy outage probability (SOP) to investigate the reliability and security performance in two cases, where the secondary destination exploits the maximum ratio combining (MRC) scheme or the selection combining (SC) scheme. To characterize the tradeoff between reliability and security, we then investigate the effective secrecy throughput (EST) performance by including the COP and SOP in a unified manner. In order to gain additional insights from the performance evaluation, we also provide the asymptotic expressions for the COP, SOP, and EST in high signal-to-noise ratio region. It is demonstrated that the interference temperature constraint incurred from the primary network enables a tradeoff between reliability and security of the secondary network. Moreover, the resulting analysis shows that using the untrusted relay to forward the transmitted message is unnecessary when the secondary destination employs the SC scheme and the untrusted relay operates in half-duplex mode.

Physical layer security is an emerging technique to protect the wireless communications in the Internet of Things (IoT). Motivated by the fact that a single IoT terminal usually occupies a very small fraction of feedback resources, we propose a novel secure transmission design with feedback compression to improve the feedback resources utilization for secure communications. Specifically, we first introduce a multi-period one-feedback (MPOF) scheme to exploit the channel temporal correlation existing in the IoT scenarios, making the IoT terminal convey its channel knowledge to the central controller in a more efficient manner. Under this MPOF scheme, we then put forward a virtual quantizer model and design a generalized fixed-rate secure on-off transmission scheme, where the central controller adaptively adjusts the transmission parameters in one feedback interval. By averaging the total secrecy throughput of one feedback interval over all the coherence periods thereof, we further characterize the secrecy throughput of our proposed transmission scheme and facilitate the joint optimization design of the feedback interval length, the wiretap codes and the power allocation ratios. To handle this non-convex problem, we develop an efficient approach involving the block coordinate descent (BCD) algorithm and the one-dimensional search method. Numerical results show that when the channel temporal correlation is high, our proposed secure transmission design achieves a significantly higher secrecy throughput than the conventional design constrained by the same amount of feedback resources.