In scenarios with extremely harsh channel conditions and severely limited communication resources, the reliability and effectiveness of semantic communication require urgent enhancement to satisfy the increasing demands of 6G technology. To address this issue, we propose an importance prioritized framework that integrates both message importance and feature importance to identify critical semantics for reliable and efficient semantic transmission. Considering service personalization and task intelligence, we analyze the message importance by factoring the receiver’s preferences and the communication tasks requirements. Specifically, a large AI model is introduced to quantify message importance, while an importance-based metric for semantic accuracy is established to evaluate the overall reliability of semantic communication. To safeguard significant messages in harsh channel conditions, an unequal error protection strategy based on message importance is employed. Furthermore, we propose a novel approach for feature importance analysis based on loss variation to accurately identify critical features. A feature importance prediction network is designed for algorithm deployment. Additionally, a semantic compression strategy based on feature importance is utilized to prioritize the transmission of essential features in limited communication resources scenarios. Extensive experimental results demonstrate substantial performance advantages of our framework and methods, especially in low signal-to-noise ratio and communication resource shortages, providing a reliable and efficient solution for semantic communication in adverse communication environments.

Unmanned aerial vehicles (UAVs) acting as aerial base stations are regarded as an effective solution for emergency communication, due to their high mobility and low cost. In this paper, we consider the differentiated communication requirements in disaster relief scenarios, where the uplink demands high throughput and the downlink requires low latency and high reliability. To simultaneously capture the timeliness and reliability requirements of downlink transmissions, we propose a novel QoS evaluation metric based on finite blocklength theory. Building on this, we formulate a system utility maximization problem and solve it by jointly optimizing UAV trajectory and ground node access control. To address this problem, we propose a novel algorithm named PW-QMIX, which integrates a priority-based heuristic access control policy with a QMIX-based trajectory optimization scheme, effectively tackling the challenges posed by the high-dimensional joint action space. Furthermore, to tackle the challenge of dynamic observation caused by UAV movement and sensing limitations, we design a weight generation network (WGN) and incorporate it into the input layer of QMIX. The WGN dynamically generates weight matrices based on observed nodes, enhancing UAV’s adaptability to local observation variations. Simulation results validate the significance of jointly optimizing uplink and downlink communications. Moreover, compared with state-of-the-art algorithms, the proposed PW-QMIX demonstrates significant advantages in convergence and scalability.

Unmanned aerial vehicles (UAVs) significantly augment mobile edge computing (MEC) networks with their flexible deployment. In this paper, we investigate a priority-driven multi-UAV cooperative MEC system, in which the task priority are jointly determined by the task queue and task type. The system aims to maximize the task priority gain, subject to the constraints on offloading decision, UAV trajectory design and task scheduling. To solve this problem, we develop a priority scheduling insert based heterogeneous Q-mixing networks (PSI-HQMIX) framework, where the PSI scheme dynamically updates the position of tasks within the queues and the HQMIX algorithm is used to obtain the optimal offloading decisions and trajectories. Simulation results demonstrate that the proposed algorithm outperforms benchmark algorithms in terms of the achieved average priority gains and convergence.

The 6G wireless communication demands intelligent and versatile interaction between humans and machines that can deal with various intelligent tasks. Semantic communication that focuses on transmitting the meanings rather than the data is expected to be one of the promising technologies to achieve this goal. However, most existing semantic communication systems optimize the whole system under a single objective, lacking the scalabi-lity to dynamic scenarios. For a dynamic scenario with changing channel conditions and background knowledge, we propose a layered semantic communication system (LSCS), which takes advantage of layered coding architec-ture at the semantic and syntactic levels. In addition, a symbolic attention-based denoising network is designed at the receiver to recover transmitted meanings. Simulation results demonstrate that the proposed LSCS can adapt to dynamic scenarios and achieve superior performance over benchmarks under different channel conditions, especially in the low signal-to-noise ratio (SNR) region.

This paper considers low-complexity hybrid analog-digital precoding for millimeter wave (mmWave) multiple-input multiple-output (MIMO) systems with finite alphabet inputs. We analyze the mutual information with finite alphabet inputs in the extreme signal-to-noise ratio (SNR) region. At low SNR, we find out that it is optimal to transmit signals through its strongest sub-channel, and thus formulate the analog precoding problem as a single-variable optimization problem. An iterative coordinate descent algorithm is developed to obtain the near-optimal hybrid precoder. At high SNR, the mutual information degrades when the Gaussian scheme is directly applied into the practical systems with finite alphabet inputs. Therefore, a modified fully-digital precoder is proposed to improve the mutual information. Numerical results validate that the proposed schemes achieve a good trade-off between mutual information and computational complexity.

Devices in the Internet of Things (IoT) are usually limited in computing resources and energy capacity, which means that encryption schemes with higher complexity are not suitable for them to ensure secure communication. As a promising solution to this problem, physical layer key generation suggests that shared secret keys can be generated from noisy wireless channel measurements to enhance the security of wireless communications. In this article, we propose a key generation scheme with extremely low implementation complexity, which allows physical layer key generation to be implemented on IoT nodes. First, we preprocess the channel measurements with simple moving average filtering before quantization to improve channel reciprocity. Next, a bidirectional difference quantization scheme is proposed to realize reliable quantization of channel measurements, which is ingenious in that the quantization process does not depend on quantization thresholds, and thus, the mismatched key bits caused by measurements close to quantization thresholds can be effectively avoided. Then, we propose an improved Cascade protocol to achieve lightweight and efficient information reconciliation. The simulation results show that our scheme can well balance the reliability and efficiency of key generation, and has excellent performance in terms of implementation complexity and key randomness.

In this paper, we devise a highly efficient machine learning-based channel estimation for orthogonal frequency division multiplexing (OFDM) systems, in which the training of the estimator is performed online. A simple learning module is employed for the proposed learning-based estimator. The training process is thus much faster and the required training data is reduced significantly. Besides, a training data construction approach utilizing least square (LS) estimation results is proposed so that the training data can be collected during the data transmission. The feasibility of this novel construction approach is verified by theoretical analysis and simulations. Based on this construction approach, two alternative training data generation schemes are proposed. One scheme transmits additional block pilot symbols to create training data, while the other scheme adopts a decision-directed method and does not require extra pilot overhead. Simulation results show the robustness of the proposed channel estimation method. Furthermore, the proposed method shows better adaptation to practical imperfections compared with the conventional minimum mean-square error (MMSE) channel estimation. It outperforms the existing machine learning-based channel estimation techniques under varying channel conditions.

This paper considers secure communication for amplify-and-forward (AF) relay networks with finite alphabet input. The joint optimization of power selection and beamforming design for improving the physical layer security of AF relay networks with single and multiple eavesdroppers is investigated. For the case with one eavesdropper, we transform the problem of multi-variable beamforming design into a single-variable optimization problem through semi-definite programming and solve it with one-dimensional optimization techniques. Moreover, the corresponding source power is obtained by utilizing the relation between the mutual information and minimum mean square error. Then, an iterative two-step algorithm is proposed to maximize the achievable secrecy rate. In the presence of multiple eavesdroppers, a zero-forcing beamforming scheme, where the confidential signal is nulled out in the direction of all eavesdroppers, is proposed to enhance the physical layer security. We decouple the source power and the beamforming vector by transforming the achievable secrecy rate into a single-variable function of the source power. Then, the suboptimal source power and the corresponding beamforming vector with low-complexity are derived. Numerical examples show that the proposed schemes significantly enhance the secrecy performance of the AF relay networks.

This study considers linear precoding for secure transmission in a multiuser peer‐to‐peer relay network with finite alphabet input. Under the assumption that the global channel‐state‐information is available, the achievable secrecy rate is derived. However, the computational complexity to evaluate the achievable secrecy rate grows exponentially with respect to the number of pair users. To reduce the computational complexity caused by the multiuser interference, an accurate approximation of the achievable secrecy rate is derived. Based on Karush–Kuhn–Tucker analysis, necessary conditions for the optimal precoder which maximises the approximated achievable secrecy rate are presented. In light of this, an iterative gradient method is developed to find the optimal precoder. Numerical examples demonstrate that the proposed scheme achieves significant gains in terms of the secrecy rate over schemes designed for Gaussian input.