4 October 2021 in IEEE Transactions on Network Science and Engineering
Reconfigurable wireless network can flexibly provide efficient spectrum access service and keep stable operation in highly dynamic environment. In this paper, a primary-prioritized recurrent deep reinforcement learning algorithm for dynamic spectrum access based on cognitive radio (CR) technology is proposed. The spectrum Markov state is modeled to capture the evolution behavior to achieve the priority queuing of the primary users and the secondary users. According to the spectrum access strategies of the secondary users under different optimal criteria, we can obtain the best tradeoff benefits of spectrum access fairness and throughput. Furthermore, we proposed a learning-based algorithm for dynamic spectrum access, which allows the secondary users to modify their parameters to select the optimal access policy to maximize network throughput utilization. The Dueling Deep $Q$ -Network (Dueling DQN) with prioritized experience replay combined with recurrent neural network is used to improve the convergence speed. Extensive experimental results demonstrate that the proposed RDRL scheme outperforms the existing Dueling DQN and DQN schemes in terms of convergence speed and channel throughput.
25 December 2023 in IEEE Journal on Selected Areas in Communications
Human-centric communication metaverse relies on the convergent integration of multiple existing technologies such as 5G and beyond networks, virtual reality, augmented reality, and digital twins, and thus their security vulnerabilities and vulnerability to interference attacks may also be inherited by the metaverse. In particular, existing security policies may be inefficient for communication interference problems encountered in multi-device collaborative computing in a metaverse 5G environment and lack adaptability to metaverse applications. In this paper, we propose a novel intelligent game anti-interference collaborative computing model that accurately describes the interference relationships among source devices, cooperating devices, and interferers in metaverse collaborative computing.We model the offensive and defensive confrontation between multiple metaverse devices as a Stackelberg game, where the source device is the leader, the collaborative computing device acts as the sub-leader, adjusts its antijamming strategy according to the source device’s strategy to improve the source device’s communication anti-jamming performance, and the jammer acts as the follower. We design an intelligent Stackelberg Game-theoretic Policy-based Learning (SGPL) algorithm for jamming resistance in metaverses over 5G and Beyond Networks, where the leaders (co-computing devices) update their training parameters using the total derivatives of the objective function, while the followers (i.e., jammers) update their training parameters using an independent gradient dynamics strategy. Loops are eased and convergence is accelerated by introducing differential dynamics into the leader training network to reflect the interaction structure of the critic and actor network layers. Finally, numerical results demonstrate the effectiveness of the proposed SGPL algorithm in metaverse anti-jamming countermeasures. The proposed SGPL algorithm has the potential to be generalized to other metaverse applications with multi-user attack and defense characteristics.
10 August 2022 in Expert Systems with Applications
8 August 2022 in IEEE Transactions on Emerging Topics in Computational Intelligence
Intelligent computation offloading and resource allocation for mobile users (MUs) in ubiquitous edge Internet of Things (IoT) systems is a worthy research hotspot. To improve the latency and energy consumption of MUs in ubiquitous edge IoT systems, we propose a deep reinforcement learnin (DRL)-driven intelligent joint optimization strategy for computation offloading and resource allocation that includes relay selection, offloading decisions, and resource allocation. Specifically, according to the limited coverage of base stations and the extremely high deployment costs in actual environments, we introduce the virtual backbone architecture to provide users with efficient multi-hop offload services through a connected dominating set (CDS). Then, we propose a CDS-based deep reinforcement learning algorithm to search for the shortest path from the MUs to the multi-access edge computing (MEC) server. Furthermore, based on the highly coupled relationship between the offloading decision and resource allocation, we design a DLIO algorithm to solve for the joint optimization of computation offloading and resource allocation. The experimental results demonstrate that our proposed optimization algorithm outperforms the state-of-the-art methods in terms of total system cost, success rate, and acceptance number.
28 February 2023 in Expert Systems with Applications
3 November 2025 in ACM Transactions on Autonomous and Adaptive Systems
In the era of Industry 5.0, Unmanned Aerial Vehicles (UAVs) equipped with multiple sensors play a vital role in industrial tasks such as patrol and surveillance, offering distinct advantages of high mobility, accurate perception, and autonomous operation. However, traditional path planning methods for UAVs struggle with challenges related to interpretability and adaptation to dynamic industrial environments. To address these challenges, this paper proposes a novel LLM-based approach to UAV swarm path planning in autonomous and adaptive industrial systems. The proposed method aims to minimize the time and computational consumption of path planning while improving the task completion rate of UAVs. Specifically, we first propose the multi-step deep thinking movement decision samples generation algorithm (MSDTMD-SG) to generate deep thinking training samples for LLMs at different ends and fine-tune them. Second, we design a scene memory and replay learning mechanism, enabling damaged UAVs to store perceived information and generate training samples via MSDTMD-SG for continuous LLM learning and optimization. Finally, extensive experiments demonstrate that the proposed method exhibits strong adaptability to dynamic environments, achieves the highest task completion rate among all methods, and maintains competitive system consumption.