LA-DRQN is proposed, a novel hierarchical decision-making framework that synergistically integrates Learning Automata at the strategic level for macro-policy selection with a Deep Recurrent Q-Network (DRQN) at the tactical level for fine-tuning computation offloading ratios.
The growing smart devices (SDs) in the Industrial Internet of Things (IIoT) generate complex computations that strain the performance and energy of local processing. Mobile Edge Computing (MEC) addresses this by providing nearby computing resources for low-latency offloading. However, achieving efficient computation offloading under massive device concurrency and densely distributed computation offloadings remains a key challenge. To address this, this paper constructs a multi-server MEC system model for IIoT and introduces Mean-Field Game (MFG) theory to model the offloading competition among SDs. This effectively reduces the dimensionality and complexity of multi-agent interactions. A novel Mean-Field Computation Offloading (MFCO) algorithm is proposed, which combines MFG with Rainbow Deep Q-Network under a Multi-Agent Deep Reinforcement Learning framework. By incorporating advanced components such as distributional value estimation, prioritized experience replay, multi-step learning, and dueling architecture, each SD acts as an autonomous agent, optimizing its policy based on local observations and mean-field approximations. Further enhancements include Boltzmann exploration, adaptive learning rates, and a mean Q-network structure, which improve convergence speed and training stability. Extensive simulations on a large-scale IIoT platform (100 SDs, 9 MEC servers) demonstrate that MFCO reduces computation latency and improves long-term rewards while maintaining robust server performance.
Xinmin Cheng, Chengquan Yu, Lu Gao et al.· IEEE Transactions on Green C...· 0 citations
Networks with highly dynamic data transmission demands and network topologies are common in real world. A fundamental problem in such networks is achieving scalable traffic allocation to maximize long-term total throughput under link capacity constraints. However, state-of-the-art (SOTA) works lack scalability. This is primarily due to two reasons in large-scale networks: first, they require solving constrained optimization problems online, which leads to high decision latency; second, they rely on reinforcement learning algorithms for policy optimization, which are inefficient in exploration and challenging to train effectively. To address these issues, we propose the Fast Networked Control (FNC) policy framework, which firstly utilizes parallelizable neural network modules to process the state and generate raw decisions, followed by basic operations such as normalizations and comparisons, which do not require iteration or optimization, to obtain decisions that satisfy the constraints. Hence, FNC policy avoids solving constrained optimization problems and supports parallel execution, significantly reducing decision latency. Furthermore, this policy preserves gradient flow and supports backpropagation, which enable us to design an imitation learning algorithm to efficiently train the policy in an end-to-end manner. Experiments in large-scale networks show that our FNC policy achieves an average 8% improvement in demands satisfaction and 10 times reduction in decision latency versus SOTA works.
Zhaoxing Yang, Guiyun Fan, An-Jie Cao et al.· IEEE Transactions on Network...· 0 citations
As the Multimedia Internet of Things (M-IoT) evolves, the orchestration of numerous resources that offer support for high-bandwidth, low-latency applications arises as a key challenge. Architecturally, the edge-cloud framework alleviates structural concerns, but the linked nature of compute and data transfer poses problems of resource management. Approaches that tackle task offloading and adaptive transmission that think independently of each other tend to have problems such as user-server cross-region overloads or network congestion. This paper presents JATO, a framework to jointly tackle the problems of adaptive task offloading and transmission optimization using Deep Reinforcement Learning. JATO offers a mono-faceted solution, learning a policy to simultaneously determine the best offloading target and the transmission quality. The framework was implemented for evaluation with a combination of different edge devices in a testbed alongside a simulation environment. JATO recorded a result of 0.9321 as the holistic score of the overall framework endpoint, a score significantly better than that of all the other frameworks that were used as functional baselines. JATO was able to resource optimally with a network lag of 131.65 milliseconds and a network freeze of 0.09% with the resources utilized. This is evidence that offloading and rate control in combination provides better resource elasticity for M-IoT systems.
G. Purnama, Irma Amelia Dewi, A. Langi et al.· Journal of ICT Research and...· 0 citations
Autonomous driving struggles to anticipate events due to the limitations of first-person perception. By offloading sensor data, environmental awareness can be shared between road users and greatly improve perception of obstacles, leading to safer and more versatile trips. To this end, vehicular networks require stable, high-performance connectivity across heterogeneous Radio Access Technologies (RAT), yet existing selection mechanisms react to degradation rather than anticipating it. We present a predictive quality-of-service (QoS) framework that jointly forecasts latency and packet delivery rate using dual-output recurrent neural networks, enabling proactive RAT selection across 5G NR, C-V2X, and DSRC. Where most frameworks tend to stop at link continuity, our proposal operates as a hierarchical closed-loop system that also maximizes throughput: a model-based selector chooses the optimal RAT, then a discrete-time Markov chain adapts packet size to channel conditions, then a queue manager enforces per-RAT capacity constraints, leading to new QoS measurements and online retraining, closing the loop. To study scalability, we introduce per-RAT contention models grounded in each RAT’s scheduling standard, and evaluate a two-pass contention-aware selection policy under increasing population sizes. Field trial data collected on an urban platform using V2X hardware across three simultaneous RATs validates the approach. The predictive QoS scheme achieves 75.2% of transmissions above 99% PDR versus 63.7% for the reactive baseline, with 40–60% fewer RAT handovers. Upscaled simulations show that contention-aware selection yields measurable reliability gains at medium and high densities, where shared-medium RATs face meaningful resource pressure. Dynamic packet sizing further improves per-vehicle throughput by up to 12.6% by exploiting high-reliability windows to increase payload.
Lucas Bréhon–Grataloup, Rahim Kacimi· International Conference on...· 0 citations
With the advancement of autonomous driving and smart navigation, Internet of Vehicles (IoV) systems face stringent requirements for real-time data delivery and processing reliability. Traditional metrics cannot fully capture information timeliness due to network dynamics and packet loss. Existing approaches also struggle with the coupling between task offloading and resource allocation, lacking adaptability in dynamic IoV environments. To address these issues, we propose a joint optimization scheme using a deep Q-network (DQN). Specifically, we build an IoV system model incorporating V2V and V2I communication, and formulate an optimization problem to minimize the average age of information (AAoI) under delay, bandwidth, computing, and energy constraints. We then design a mixed-action DQN algorithm with dual-network architecture, experience replay, and an action mask mechanism to enhance training stability and environmental adaptability. Simulation results show that our DQN-based scheme achieves the lowest AAoI among Random, Greedy, A2C, and DDQN, with reductions of 29.5%, 8.9 %, 7.1 %, and $\mathbf{7. 6 \%}$, respectively. It also exhibits superior delay and energy performance, confirming its effectiveness for dynamic IoV task offloading and resource allocation.
Chao He, Wanting Wang, Dongfeng Fu et al.· 2026 International Conferenc...· 0 citations
The rapid deployment of fifth-generation (5G) mobile communication systems has significantly transformed wireless connectivity by supporting ultra-high data rates, ultra-low latency, massive machine-type communications, and heterogeneous Internet of Things (IoT) applications. However, the unprecedented growth in mobile traffic, dynamic user mobility, and diversified quality-of-service requirements have introduced substantial challenges in traffic prediction, congestion management, spectrum utilization, and network resource allocation. Conventional optimization techniques often fail to adapt to the highly dynamic and nonlinear characteristics of modern 5G environments. Deep learning has emerged as an effective paradigm for intelligent traffic optimization by learning complex spatial-temporal traffic patterns from large-scale network data and enabling proactive decision-making. Advanced architectures such as Long Short-Term Memory networks, Convolutional Neural Networks, Graph Neural Networks, Autoencoders, and Deep Reinforcement Learning provide enhanced capabilities for traffic forecasting, dynamic routing, load balancing, network slicing, edge intelligence, and energy-efficient resource management. This paper investigates the integration of deep learning algorithms into 5G traffic optimization frameworks, presents a comprehensive review of recent developments, proposes an intelligent AI-driven optimization architecture, evaluates major performance metrics, and discusses implementation challenges, scalability issues, and future research directions toward autonomous next-generation mobile networks.
Kamal N, Thirupathi Sundararajulu, A. R et al.· International journal of com...· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
