In this paper, to enhance the communication quality of cognitive unmanned aerial vehicle networks (CUAVNs), we investigate the intelligent reflecting surface (IRS)-assisted CUAVNs, where a leading UAV (LUAV) is deployed for sensing spectrum and a group of following UAVs (FUAVs) transmit data to LUAV with the aid of IRS. Our objective is to maximize the achievable throughput of CUAVNs by jointly optimizing sensing duration, IRS passive beamforming and LUAV’s three-dimensional (3D) location, where LUAV’s 3D location is restricted by IRS, FUAVs and primary user. For IRS-assisted single FUAV case, the intractable non-convex optimization problem is resolved into three subproblems, which are solved by utilizing the bisection search method, the closed-form expression of the optimal IRS phase shift matrix and the successive convex approximation method, respectively. Finally, an efficient alternating optimization algorithm is developed to obtain a high-quality suboptimal solution. By exploiting the solutions of single FUAV, we further propose the throughput weighted sum (TWS) algorithm to solve the intricate non-convex problem in IRS-assisted multiple FUAVs case. To further reduce the complexity of TWS, the low-complexity location weighted sum (LWS) algorithm is proposed. Numerical results show that compared to the scheme without IRS assistance, the achievable throughput increases about 102% with the proposed single FUAV scheme, and over 88% with the proposed TWS-based multiple FUAVs scheme. Moreover, the performance gap between the low-complexity LWS and the TWS is less than 7%.

The data-driven deep learning methods have brought significant progress and potential to intrusion detection. However, there are two thorny problems caused by the characteristics of intrusion data: "multi-type features" and "data imbalance". The former means that forcefully and improperly transforming intrusion features from distinct metric spaces can result in semantic loss and noise. The latter indicates that the intrusion data is imbalanced in quantity and quality due to its complex spatial distribution. We propose a Hybrid Framework for Multi-type and Imbalance Data (HF-Mid) to address the above two problems. Firstly, we divide the intrusion features into equivalent and non-equivalent groups, and then embed them sequentially using Supervised Paragraph Vector-Distributed Memory (SPV-DM), which excels at modeling co-occurrence relationships, and Deep Neural Network (DNN), which is suitable for modeling non-linear relationships, thereby solving the "multitype features" problem. Secondly, we adopt a low-noise collective matrix factorization (CMF) model to fuse the two obtained features for dimensionality reduction. Finally, we employ a multiple classifier to detect intrusion. During the classifier training stage, we design a genetic algorithm-based proportional sampling method to select high-quality samples in each training batch. thus addressing the "data imbalance" problem. The experimental results demonstrate the proposed framework exhibits an overall improvement of 5.9% and 1.5% in terms of accuracy and false positive rate on average, respectively.

In this article, an effective energy-efficiency optimization problem is investigated in the cognitive unmanned aerial vehicle (UAV) communication system, where the moving following UAV (FUAV) transmits collected data to the leading UAV (LUAV) by reusing the spectrum of the ground primary user. For this scenario, a novel joint UAV trajectory and resource allocation optimization algorithm is proposed. We aim to maximize the energy efficiency of the cognitive UAV communication systems under interference, shortest step size, collision prevention, and speed constraints. However, this optimization problem is difficult to be solved, as it is nonconvex and involves strong relations among many variables. To address this issue, we first decompose the original optimization problem into four subproblems: 1) spectrum sensing duration subproblem; 2) spectrum sensing threshold subproblem; 3) transmitted power subproblem; and 4) FUAV trajectory optimization subproblem. For spectrum sensing duration subproblem and spectrum sensing threshold subproblem, their optimal solutions can be efficiently solved via a golden section search method. For the transmitted power subproblem, the closed-form expression of the optimal transmitted power is deduced. For the nonconvex FUAV trajectory optimization subproblem, we consider the shortest step size design under known starting and destination positions and then obtain an approximate solution of FUAV trajectory via the successive convex optimization (SCO) technique. Finally, an alternate iterative framework with fast convergence is designed to solve the original optimization problem. The simulation results show that the proposed algorithm has 18 times improvement in energy efficiency compared with the straight flight scheme with lower energy consumption.

In this paper, a novel three timeslots frame structure is proposed for Energy-Harvesting Cognitive Radio Networks, where the frame structure includes spectrum sensing, energy harvesting and non-orthogonal multiple access uplink transmission. Our goal is to maximize the sum-capacity of secondary users (SUs) by jointly optimizing spectrum sensing duration, energy harvesting duration and data transmission duration. To solve the challenging problem, we first derive the closed form expressions for optimal energy harvesting and data transmission durations by fixing the spectrum sensing duration, and then optimize the spectrum sensing duration by golden section search method. Finally, we obtain a sub-optimal solution through the alternate iteration of two previous steps. Simulation results show that the sum-capacity of SUs under the proposed scheme significantly increases compared to the time division multiple access (TDMA) uplink transmission protocol, especially when the transmitted power of cognitive base station (CBS) increases and the number of SUs is large enough.