Intelligent reflecting surface (IRS) is a promising technology to extend the wireless signal coverage and support the high performance communication. By intelligently adjusting the reflection coefficients of a large number of passive reflecting elements, the IRS can modify the wireless propagation environment in favour of signal transmission. Different from most of the prior works which did not consider any cooperation between IRSs, in this work we propose and study a cooperative double-IRS aided multiple-input multiple-output (MIMO) communication system under the line-of-sight (LoS) propagation channels. We investigate the capacity maximization problem by jointly optimizing the transmit covariance matrix and the passive beamforming matrices of the two cooperative IRSs. Although the above problem is non-convex and difficult to solve, we transform and simplify the original problem by exploiting a tractable characterization of the LoS channels. Then we develop a novel low-complexity algorithm whose complexity is independent of the number of IRS elements. Moreover, we analyze the capacity scaling orders of the double-IRS aided MIMO system with respect to an asymptotically large number of IRS elements or transmit power, which significantly outperform those of the conventional single-IRS aided MIMO system, thanks to the cooperative power gain brought by the double-reflection link and the spatial multiplexing gain harvested from the two single-reflection links. Extensive numerical results are provided to show that by exploiting the LoS channel properties, our proposed algorithm can achieve a desirable performance with low computational time. Also, our capacity scaling analysis is validated, and the double-IRS system is shown to achieve a much higher rate than its single-IRS counterpart as long as the number of IRS elements or the transmit power is not small.

In the existing cellular networks, it remains a challenging problem to communicate with and control an unmanned aerial vehicle (UAV) swarm with both high reliability and low latency. Due to the UAV swarm's high working altitude and strong ground-to-air channels, it is generally exposed to multiple ground base stations (GBSs), while the GBSs that are serving ground users (occupied GBSs) can generate strong interference to the UAV swarm. To tackle this issue, we propose a novel two-phase transmission protocol by exploiting cellular plus device-to-device (D2D) communication for the UAV swarm. In Phase I, one swarm head is chosen for ground-to-air channel estimation, and all the GBSs that are not serving ground users (available GBSs) transmit a common control message to the UAV swarm simultaneously, using the same cellular frequency band. Both the swarm head and other swarm members can utilize the high power gain from multiple available GBSs' transmission, to combat the strong interference from occupied GBSs, while some UAVs may fail to decode the message due to uncorrelated ground-to-air channels. In Phase II, all the UAVs that have decoded the message in Phase I further relay it to the other UAVs in the swarm via D2D communication, by exploiting the less interfered D2D frequency band and the proximity among UAVs. In this paper, we aim to characterize the reliability performance of the above two-phase transmission protocol, i.e., the expected percentage of UAVs in the swarm that can decode the common control message, which is a non-trivial problem due to the complex system setup and the intricate coupling between the two transmission phases. Nevertheless, we manage to obtain an approximated expression of the reliability performance of interest, under reasonable assumptions and with the aid of the Pearson distributions. Numerical results validate the accuracy of our analytical results and show the effectiveness of our proposed protocol over other benchmark protocols. We also study the effect of key system parameters on the reliability performance, to reveal useful insights on the practical design of cellular-connected UAV swarm communication.

Unlike passive eavesdropping, proactive eavesdropping is recently proposed to use jamming to moderate a suspicious link's communication rate for facilitating simultaneous eavesdropping. This paper advances the proactive eavesdropping research by considering a practical half-duplex mode for the legitimate monitor (e.g., a government agency) and dealing with the challenging case that the suspicious link opportunistically communicates over parallel fading channels. To increase eavesdropping success probability, we propose cognitive jamming for the monitor to change the suspicious link's long-term belief on the parallel channels' distributions and thereby induce it to transmit more likely over a smaller subset of unjammed channels with a lower transmission rate. As the half-duplex monitor cannot eavesdrop the channel that it is simultaneously jamming to, our jamming design should also control the probability of such “own goal” that occurs when the suspicious link chooses one of the jammed (uneavesdroppable) channels to transmit. We formulate the optimal jamming design problem as a mixed integer nonlinear programming (MINLP) and show that it is non-convex. Nevertheless, we prove that the monitor should optimally use the maximum jamming power if it decides to jam, for maximally reducing the suspicious link's communication rate and driving the suspicious link out of the jammed channels. Then we manage to simplify the MINLP to integer programming and reveal a fundamental trade-off in deciding the number of jammed channels: jamming more channels helps reduce the suspicious link's communication rate for overhearing more clearly but increases own goal probability and thus decreases eavesdropping success probability. Finally, we extend our study to the two-way suspicious communication scenario and show that there is another interesting trade-off in deciding the common jammed channels for balancing bidirectional eavesdropping performances. Numerical results show that our optimized jamming-assisted eavesdropping schemes greatly increase eavesdropping success probability as compared with the conventional passive eavesdropping.