Shao-Yu Lien

To support emerging low-altitude economic activities such as urban monitoring, unmanned aerial vehicles acting as non-terrestrial access points (NT-APs) play a pivot role in providing high data-rate communication and high-resolution sensing services in non-terrestrial networks (NTNs). To tackle optimization for matching spatial-temporally varying resources and demands, NT-APs can utilize artificial intelligence (AI) models to enable near-real-time decision-making. Nevertheless, due to the partial observability, those NT-APs as AI-agents struggle to actively perform long-horizon adaptability to the dynamic environment. Toward optimizing the overall performance, the agentic AI technology with coordination to AI-agents in different levels is a promising remedy to orchestrate communication, sensing services and resource control models. To this end, we propose a multi-AI-agent coordinated architecture to empower intelligent resource optimization in cell-free communication and sensing (C&S) services, in which the C&S service clustering and resource allocation are tackled by AI-agents at central processing units and NT-APs, respectively. By leveraging multi-agent rollout training and cross-task information exchange, AI-agents can tackle individual sub-problems such as trajectory design and beam management in a nearly-stationary environment. To enable continuous policy improvement, model-based planning is conducted using online-collected sensing results of unexplored environments. Finally, simulations on the obstacle-avoidance task of air traffic management demonstrate the effectiveness of the proposed design over the centralized evolutionary algorithm and centralized-training-and-decentralized-execution deep reinforcement learning (DRL).

Mobility management associating all user equipments (UEs) to proper base stations (BSs) [also known as handover (HO)] to achieve the designated performance optimization is one of the most crucial functions in mobile networks. Although effective mobility management has received considerable research attentions, existing schemes follow event-driving operations, in which HO decisions are made based on events of performance degradation that BSs/UEs passively suffer from or proactively foresee. However, the performance and decisions of these schemes are highly subject to the identified events, to lose generalization and better performance under unidentified events. To address this issue, we propose a foundation model (FM)-based mobility management for the sixth-generation (6G) mobile networks inherently supporting artificial intelligence (AI) computing, in which the FM generates the HO decisions for all UEs to maximize the overall throughput under the constraints of ping-pong rate and HO failure (HOF) rate by implicitly taking moving trajectories, traffic demands, channel conditions of all UEs, and available resources of BSs into account. To this end, a hierarchical model structure composed of long short-term memory (LSTM) networks with multihead attention (MHA) is adopted, which is trained by emulated datasets with augmentation. The performance evaluation results show that the proposed scheme outperforms the state-of-the-art schemes in terms of the average throughput over all UEs while satisfying the required ping-pong rate and HOF rate, and justify the robustness of the proposed FM under different network deployment scenarios.

The rapid evolution of Industry 4.0 has led to the emergence of smart factories, where multirobot system autonomously operates to enhance productivity, reduce operational costs, and improve system adaptability. However, maintaining reliable and efficient network operations in these dynamic and complex environments requires advanced automation mechanisms. This study presents a zero-touch network platform that integrates a hierarchical Open Radio Access Network (O-RAN) architecture, enabling the seamless incorporation of advanced machine learning algorithms and dynamic management of communication and computational resources, while ensuring uninterrupted connectivity with multirobot system. Leveraging this adaptability, the platform utilizes federated deep reinforcement learning (FedDRL) to enable distributed decision-making across multiple learning agents, facilitating the adaptive parameter reconfiguration of transmitters (i.e., multirobot system) to optimize long-term system throughput and transmission energy efficiency. Simulation results demonstrate that within the proposed O-RAN-enabled zero-touch network platform, FedDRL achieves a 12% increase in system throughput, a 32% improvement in normalized average transmission energy efficiency, and a 28% reduction in average transmission energy consumption compared to baseline methods such as independent DRL.

The reconfigurable intelligent surface (RIS) steering reflective beam directions toward a target mobile user equipment (UE) has been a promising technology for coverage enhancement and physical-layer (PHY) security to avoid signal eavesdroppers and jammers. However, tracking and predicting the moving trajectory of UE demand significant computing burdens unaffordable to the RIS. Performing beam steering computations on the radio access network (RAN) intelligent controller in the open RAN (O-RAN) architecture therefore emerges as a revolutionary paradigm. In this paper, we propose a PHY-secured beam steering design based on a pre-processing-based Kalman filter algorithm. Through a pre-processing phase to derive the radiation feature, the proposed design accurately tracks the position of the mobile UE and quickly recovers the position search while position tracking fails. Through practical field-try experiments, we justify the capability of O-RAN RIC-empowered RIS for PHY security.

With promising benefits of splitting deep neural network (DNN) computation loads to the edge server, split computing has been a novel paradigm achieving high-quality artificial intelligence (AI) services for the energy-constrained user equipments (UEs). To satisfy the service demands of a large number of UEs, traditional edge-UE split computing evolves toward multitier split computing involving the edge and cloud servers with different capabilities, leading to a “complex” optimization involving communication and computation. To tackle this challenge, in this article, we propose a multitier deep reinforcement learning (DRL) decision-making scheme for distributed splitting point selection and computing resource allocation in the three-tier UE-edge-cloud split computing systems. With the proposed scheme, the high-dimensional optimization can be tackled by the UEs and an edge server with different control cycles through performing local decision-making tasks in a sequential manner. Based on the policies updated by the UEs and the edge server in successive stages, the overall performance of split computing can be continuously improved, which is justified through a theoretical convergence performance analysis. Comprehensive simulation studies show that the proposed multitier DRL decision-making scheme outperforms the conventional split computing schemes in terms of the overall latency, inference accuracy, and energy efficiency to practice multitier split computing.

Cellular vehicle-to-everything (C-V2X) was introduced to support autonomous driving through 5G and beyond networks. C-V2X leverages cellular network infrastructure to integrate vehicle-to-network, vehicle-to-pedestrian, vehicle-to-infrastructure, and vehicle-to-vehicle communications. It has been suggested that Open RAN can be used to achieve the latency requirements essential to realize C-V2X as it achieves real-time optimization through the use of AI in Near real-time RAN Intelligence Controller (Near-RT RIC). The Open RAN will allow the access to historical traffic data or acquisition of data from vehicles. The data will then be transferred to Near-RT RIC for detecting network anomalies while maintaining reliable communication, which is essential for realizing autonomous driving. Open RAN also supports non-real-time RAN intelligent controller (Non-RT RIC) that allows more complex ML workflows such as policy-based feature extraction and optimization to guide vehicles when real-time acquisition is not available. Open RAN provides support for edge cloud, i.e. Open Cloud that helps to interface the Near-RT RIC with Open RAN central unit's user and control plane. Together, the Open RAN and C-V2X are considered to be the key-enabling technologies for achieving low-latency in autonomous vehicular communication networks. The issue attracted over 100 high-quality submissions from all over the world, among which 12 original contributions were eventually selected for publication. The novelty and key contributions of these articles are summarized as follows.

With the merit of global coverage, non-terrestrial networks (NTNs) with low-earth-orbit (LEO) satellite constellations have been regarded as a promising innovation to support ubiquitous connectivity. However, due to the high orbital velocity, each LEO satellite can only service a different set of ground terminals (GTs) for few minutes, and thus it is difficult to utilize machine learning (ML) based and non-ML based algorithms to tackle high-dimensional resource optimizations within such a short dwell duration. To tackle this challenge, an effective solution lies in constructing collaborations between multiple LEO satellites and GTs to jointly address the resource optimization. In this article, motivated by the mutualism advantage in symbiotic radios, we therefore propose a novel collaborative learning scheme for implementing intelligent NTNs, in which each GT with powerful computing capability operates a ground-tier learning agent to assist LEO satellites tackling GT resource allocation tasks. Each LEO satellite also operates as a space-tier learning agent and the learning model of the LEO satellite can be transferred to its successor satellite as a starting point to continue updating the learning model. Finally, comprehensive simulations are conducted with a Walker-Delta satellite constellation and a channel model composed of free-space path-loss and Rician fading, and the simulation results show that the proposed scheme outperforms DRL benchmarks with more than 5.9% improvements toward the optimal average overall throughput performance over a sufficient number of elevation angle samples from the interval of [31°, 90°].

Constructing earth-fixed cells with low-earth orbit (LEO) satellites in non-terrestrial networks (NTNs) has been the most promising paradigm to enable global coverage. The limited computing capabilities on LEO satellites however render tackling resource optimization within a short duration a critical challenge. Although the sufficient computing capabilities of the ground infrastructures can be utilized to assist the LEO satellite, different time-scale control cycles and coupling decisions between the space- and ground-segments still obstruct the joint optimization design for computing agents at different segments. To address the above challenges, in this paper, a multi-time-scale deep reinforcement learning (DRL) scheme is developed for achieving the radio resource optimization in NTNs, in which the LEO satellite and user equipment (UE) collaborate with each other to perform individual decision-making tasks with different control cycles. Specifically, the UE updates its policy toward improving value functions of both the satellite and UE, while the LEO satellite only performs finite-step rollout for decision-makings based on the reference decision trajectory provided by the UE. Most importantly, rigorous analysis to guarantee the performance convergence of the proposed scheme is provided. Comprehensive simulations are conducted to justify the effectiveness of the proposed scheme in balancing the transmission performance and computational complexity.

A potential candidate technology for the development of future 6G networks has been recognized as Reconfigurable Intelligent Surface (RIS). However, due to the variation in radio link quality, traditional passive RISs only accomplish a minimal signal gain in situations with strong direct links between user equipment (UE) and base station (BS). In order to get over this fundamental restriction of smaller gain, the idea of active RISs might be a suitable solution. In contrast to current passive RIS, which simply reflects and directs signals without any additional amplification, active RISs have the ability to enhance reflected signals by the incorporation of amplifiers inside its elements. However, with additional amplifiers, apart from the relatively complex attributes of RIS-assisted arrangements, the additional energy consumption of such technologies is often disregarded. So, there might be a tradeoff between the additional energy consumption for the RIS technologies and the overall gain acquired by deploying this potential advancement. The objective of this work is to provide a primary idea of a three-layer hybrid RIS-assisted configuration that is responsive to both active and passive RIS, as well as an additional dormant or inactive state. The single RIS structure should be capable of adjusting its overall configuration in response to fluctuations in transmit power and radio link quality. Furthermore, our fabricated passive RIS-assisted structure verifies a portion of the proposed idea, with simulations highlighting its advantages over standalone passive or active RIS-assisted technologies.