Research on Coupling Mechanism between Resonators for Advanced Antenna Design

An antenna comprising mutually coupled resonators (radiators) is typically a common approach for enhancing specific performance, such as bandwidth and radiation gain. The strategic exploitation of mutual couplings, including the coupling coefficient and phase shift, enables the realization of enhanced antenna functionalities. Recent investigations into coupling effects have garnered significant attention, particularly in circularly polarized (CP) antennas and filtering antennas. However, conventional designs often suffer complex feeding structures and lack of design theory. In this talk, I will review the related works and introduce our recent progress on the antennas by coupling effects.

Rate Maximization and Mode Selection for RDARS-assisted MIMO Communications With Perfect and Imperfect CSI

Reconfigurable distributed antenna and reflecting surface (RDARS) has been recently proposed as a promising technology. This architecture enables each element to perform flexibly either in the reflection mode, like the traditional passive reconfigurable intelligent surface (RIS), or in the connection mode, akin to the distributed antenna system (DAS). This dual capability allows RDARS to harness both reflection gain and distribution gain.  In this talk, we investigate a dynamic RDARS-aided multiple-input multiple-output communication system, where the optimal configuration of the elements operating in connection mode can provide additional selection gain. Considering the theoretical and practical significances, we address the achievable rate maximization problem by jointly optimizing the mode selection, transmit power allocation and passive beamforming under both perfect and imperfect channel state information (CSI) cases. For perfect CSI case, by investigating the inherent properties of the objective function, we propose a greedy-based alternating optimization (AO) algorithm with low-complexity and then extend the proposed algorithm to the general multi-user multi-RDARS scenario. Additionally, some interesting insights about the mode selection of RDARS in a special scenario with a single-antenna user are found. For imperfect CSI case, an efficient alternative direction method of multipliers-based AO algorithm is developed. Numerical results show that RDARS-assisted system outperforms the passive-RIS assisted system and DAS under both perfect and imperfect CSI scenarios with promising reflection, distribution and selection gains.

Enhancing Vehicular Network Efficiency: Deep Reinforcement Learning for HD Map Dissemination and Multicast Resource Allocation

The rapid evolution of autonomous driving and vehicular networks demands efficient resource allocation and timely data dissemination to ensure safety and reliability. This work addresses two critical challenges in vehicular communications: (1) optimizing high-definition (HD) map dissemination with freshness guarantees and (2) enabling intelligent resource allocation for multicast transmission in hybrid vehicle scenarios.

First, we propose a deep reinforcement learning (DRL)-based framework for HD map dissemination, jointly minimizing the Age of Information (AoI) for dynamic map updates and ensuring the Urgency of Information (UoI) for static map preloading. By integrating Lagrangian optimization with DRL, our method achieves a 10%–40% improvement in AoI control compared to existing approaches. Second, we introduce a hybrid DRL algorithm, Pretrain-Masked Proximal Policy Optimization (PM-PPO), tailored for multicast resource allocation in heterogeneous vehicular networks. Leveraging behavior cloning and action masking, PM-PPO accelerates training convergence while balancing AoI constraints and throughput maximization. Experimental results demonstrate a 15.6%–31.9% enhancement in effective traffic delivery, outperforming conventional methods.

These contributions highlight the transformative potential of DRL in addressing time-sensitive, location-dependent vehicular communication challenges, paving the way for safer and more efficient autonomous driving systems.

Fractional Fourier Domain PAPR Reduction

High peak-to-average power ratio (PAPR) has long posed a challenge for multi-carrier systems, impacting amplifier efficiency and overall system performance. This talk introduces dynamic angle fractional Fourier division multiplexing (DA-FrFDM), an innovative multi-carrier system that effectively reduces PAPR for both quadrature amplitude modulation (QAM) and Gaussian signals with minimal signaling overhead. DA-FrFDM leverages the fractional Fourier domain to balance PAPR characteristics between the time and frequency domains, achieving significant PAPR reduction while preserving signal quality. Furthermore, DA-FrFDM refines signal processing and enables one-tap equalization in the fractional Fourier domain through the simple multiplication of time-domain signals by a quadratic phase sequence. The results show that DA-FrFDM not only outperforms existing PAPR reduction techniques but also retains efficient inter-carrier interference (ICI) mitigation capabilities in doubly dispersive channels.

NOMA Empowered Integrated Sensing, Communications and Computing for Mobile Edge Networks

This presentation introduces a non-orthogonal multiple access (NOMA)-empowered integrated sensing, communications, and computing framework for mobile edge networks, addressing the critical challenges of resource optimization and interference management in emerging wireless applications. The proposed framework features a two-tier task offloading architecture, where a multifunctional AP provides task offloading services for edge computing users while performing high-precision radar sensing. To balance the utilization of the computing resources across different tiers, the AP can further offload part of the received workloads to a group of cloudlet servers. In the two-tier framework, different NOMA mechanisms are designed for alleviating different inter-functionalities interferences and thus improve the performances of both sensing and offloading. By incorporating two-tier task offloading strategies, two-tier integrated sensing and communication signal designs, and the associated allocations of the communication and computing resources, we propose an optimization to balance the tradeoffs between computational latency and sensing performance. Results validate the framework’s superiority in reducing energy consumption, enhancing task offloading performance, and improving sensing accuracy, while effectively suppressing interference from undesired directions.