Dynamic Path Selection in SDN Based on Reinforcement Learning and Link Utilization

Advanced deep reinforcement learning techniques for dynamic resource allocation in 5G heterogeneous networks

PPO-MS: Confidence-Aware and Collaborative Traffic Management for Multimedia Streaming in SDN

The rapid growth of multimedia streaming poses critical challenges, including bursty traffic and congestion, leading to playback delays. The existing separate prediction and control mechanisms for multimedia traffic scheduling, which are based on software-defined networks (SDN), are unable to proactively manage bursty traffic under uncertain conditions. This limitation is particularly evident in SDN-enabled backbone and multimedia-aware access networks, which typically assume centralized control and stable topologies. They lack integration of traffic prediction, traffic shaping, and real-time perception scheduling through reinforcement learning, resulting in low efficiency when exploring multiple paths in dynamic networks. To address this challenge, we propose PPO-MS (Proximal Policy Optimization-based Multimedia Scheduler), an SDN-based multimedia traffic scheduling algorithm integrating three key innovations: 1) A novel LSTM+HTB synergy where LSTM’s confidence intervals dynamically adjust HTB (Hierarchical Token Bucket) shaping parameters, enabling adaptive rate control under prediction uncertainty and overcoming the limitations of static LSTM+HTB hybrids; 2) A Deep Reinforcement Learning (DRL)-optimized path pruning method that reduces state and action spaces by generating a constrained set of $k$ disjoint candidate paths via an improved redundant tree algorithm. Unlike traditional multi-path schemes, this method tightly couples path preselection with the RL decision loop for adaptive, context-aware routing; 3) Generalized Advantage Estimation (GAE)-accelerated PPO for stable convergence in dynamic environments. In contrast to prior works (e.g., LSTM+RL for QoE or standalone tree algorithms), PPO-MS uniquely unifies these modules through confidence-aware traffic shaping and hierarchical decision-making, validated via comparative experiments. Results demonstrate that PPO-MS, through the synergistic integration of confidence-aware traffic shaping and DRL-optimized path pruning, significantly outperforms decoupled baselines. In particular, via isolation studies against simpler alternatives (e.g., mean-prediction and fixed-margin shaping), the confidence-aware shaping mechanism is validated to be superior under bursty traffic conditions. Overall, PPO-MS reduces end-to-end latency by 17.3% and packet loss by 32.4% while achieving 24.4% better load balancing during traffic bursts.

Reinforcement Learning for Efficient Link Scheduling in Multi-Link Networks

The increasing demand for higher throughput and lower latency in modern applications has driven the evolution of IEEE 802.11 with the introduction of Wi-Fi 7. The new Extremely High Throughput (EHT) amendment enhances performance with Multi-Link Operation (MLO), allowing concurrent transmissions over multiple frequency links. While MLO improves channel access, reduces latency, and boosts reliability, uneven traffic loads may still cause link congestion, starvation, or collisions, which can severely impact TCP flows. This work investigates the impact of MLO on TCP best effort and video traffic flows, and proposes a Reinforcement Learning (RL) Transmission Opportunity (TXOP) Random Discard strategy for adaptive load balancing. Simulation results demonstrate that the proposed AI-driven approach enhances TCP throughput and latency in Wi-Fi 7 multi-link networks.

Deep Reinforcement Learning for Handover Optimization in 5G Networks

Ultra-dense 5G networks require advanced traffic steering to maintain performance and balance load amid growing user and base station (gNB) densities. Traditional heuristics such as nearest-base-station and Signal-to-Interference-plus-Noise Ratio (SINR)-based selection provide simple solutions but struggle to adapt to dynamic user mobility, diverse traffic, and fluctuating radio conditions at the mobility-control level, often leading to inefficient handovers and degraded network quality. We propose a deep reinforcement learning (DRL) framework to dynamically tune a global handover hysteresis margin that governs handover triggering decisions, optimizing handover success, reducing failures, and enhancing throughput and fairness. Implemented in Python using Stable Baselines3 and NumPy, our custom simulation environment models key mobility-related 5G dynamics at a high level, including user mobility, pathloss-based signal degradation, and interference. We evaluate DRL agents-Deep Q-Network (DQN) and Proximal Policy Optimization (PPO)-against heuristic and hysteresis-based baselines. Results show that DRL-based hysteresis optimization provides strong and robust performance under the considered ultra-dense mobility conditions in handover success rate, average SINR, throughput, and fairness, with PPO demonstrating the most consistent behavior across configurations. This work offers a reproducible simulation framework for further research into adaptive mobility management.

Scalable Traffic Allocation in Dynamic Networks via End-to-End Imitation Learning

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.

A novel deep reinforcement learning-based cell selection algorithm for efficient mobility management in 6G networks

Simulation results position D3QN-PER as a strong candidate for deployment as a near-RT RIC xApp within the O-RAN architecture, advancing the vision of AI-native mobility management for 6G.