3 February 2020 in IEEE Access
In communications, innovative paradigm shifts have emerged in integrating various devices into the network to provide advanced and intelligent services. However, various security threats may occur that may not always be detected using traditional cryptographic techniques. Secure authentication is of paramount importance in modern wireless systems. This paper focusses on robust authentication in a time-varying communication environment where conventional authentication mechanisms are severely limited. We propose an Adaptive Neural Network (ANN) as an intelligent authentication process to improve detection accuracy. Specifically, a Data-Adaptive Matrix (DAM) is designed to track time-varying channel features. By utilizing a convolutional neural network as an intelligent authenticator, the proposed approach integrates deep feature extraction and attack detection, hence, leading to effective physical layer security. To evaluate the system, the ANN is prototyped on a universal software radio peripheral (USRP) and its authentication performance is evaluated in a conference room environment. Experimental results show that the ANN is effective in tackling the challenges of physical layer authentication under interference conditions, and is effective in time-varying environments.
17 June 2025 in IEEE Internet of Things Journal
Low-earth orbit (LEO) satellite Internet of Things (IoT) has emerged as a promising solution to address the limitations of terrestrial IoT by providing global coverage and seamless connectivity. Among the various techniques enhancing LEO satellite IoT, beam hopping (BH) stands out as an efficient approach that dynamically adjusts beam illumination to match the varying traffic demands of diverse IoT devices. This flexibility enables optimal utilization of limited on-board resources. However, while BH allows adaptive beam illumination planning, it can also introduce severe interbeam interference, particularly when adjacent beams are simultaneously activated. To address this challenge, we propose a novel rate-splitting multiple access (RSMA)-enabled cluster-based BH (CBH) LEO satellite IoT system. By leveraging RSMA, the proposed framework supports large-scale IoT devices access, and mitigates interbeam interference introduced by CBH. Within this framework, we introduce a metric-the ratio of offered capacity to traffic demand (ROCD)–to quantify how well the required traffic sum rate aligns with the achievable sum rate for each beam. We then focus on jointly optimizing the precoding vector, common rate allocation, and CBH pattern design to maximize the worst-case ROCD among beams. To solve this problem efficiently, we decompose the original problem into three subproblems and propose a two-stage algorithm. Numerical results demonstrate that our proposed scheme improves the minimum satisfaction rate by 14.10% and 39.59% compared to the nonorthogonal multiple access and space-division multiple access baselines, achieving effective interference mitigation.
3 March 2026 in IEEE Transactions on Vehicular Technology
In this correspondence, we propose an optimized beamforming algorithm for downlink multi-user communications in low Earth orbit (LEO) satellite cell-free massive multi-input multi-output (CF-mMIMO) network. Firstly, we consider a LEO satellite CF-mMIMO network based on orthogonal frequency-division multiplexing (OFDM). In such a CF-mMIMO network, a satellite cluster formed by a core satellite (CS) and multi-satellite access points (SAPs) serves the ground terminals (GTs). Secondly, to maximize the system weighted sum rate (WSR) subject to total power budget, we adopt the Lagrangian dual transform and quadratic transformation to tackle the non-convex design problem. And the block coordinate descent (BCD) method is used to optimize the beamforming matrix and auxiliary variables alternately. Finally, the proposed strategy is compared with the conventional methods. Simulation results demonstrate that the proposed strategy effectively enhances the WSR performance, outperforming conventional minimum mean-square error (MMSE) and zero-forcing (ZF) beamforming schemes by over 25$\%$. And the correspondence verifies the feasibility of LEO satellite cluster CF-mMIMO network, and provides a new direction for the design of satellite communication system.
18 March 2026 in IEEE Transactions on Vehicular Technology
The evolution of 6 G and Internet of Things (IoT) technology is driving the proliferation of computation-intensive and latency-sensitive applications, which impose increasingly stringent requirements on IoT Devices (IDs) for powerful and timely data processing capabilities. Space-Air-Ground Integrated Network (SAGINs) have emerged as a promising candidate for providing ubiquitous computing services for processing various sensing data generated by IDs, particularly in remote areas. However, supporting these applications in SAGINs still faces two key challenges: IDs cannot afford intensive data transmission and computation due to the limited battery and computing capability, and the Ultra-Reliable Low-Latency Communication (URLLC) constraints are difficult to satisfy due to the network dynamics. To tackle these challenges, we propose an Intelligent Reflecting Surface (IRS) backscatter-enabled SAGIN architecture, where battery-efficient, radio-frequency-chain-free uplink transmissions from IDs to Unmanned Aerial Vehicles (UAVs) are enabled, significantly reducing the communication energy consumption of IDs. We aim to minimize the system energy consumption for balancing low-latency requirements against energy-saving demands under URLLC constraints. Considering that resource allocation decisions and URLLC constraints have different time scales, the Lyapunov optimization is employed to decompose the formulated problem into tractable subproblems, which are solved by integrating Deep Reinforcement Learning (DRL) with fractional programming, semi-definite programming, and the Lagrange method. Simulation results demonstrate that the proposed algorithm significantly reduces the total energy consumption by 66.67% compared to benchmark schemes, thereby enabling sustainable, high-performance computation for remote IDs.