Miao Ye

Zhenqian Chen, Baoquan Zhong, Xiang Li et al.

RL post-training for LLMs has been widely scaled to enhance reasoning and tool-using capabilities. However, RL post-training interleaves training and inference workloads, exposing the system to faults from both sides. Existing fault tolerance frameworks for LLMs target either training or inference, leaving the optimization potential in the asynchronous execution unexplored for RL. Our key insight is role-based fault isolation so the failure in one machine does not affect the others. We treat trainer, rollout, and other management roles in RL training as distinct distributed sub-tasks. Instead of restarting the entire RL task in ByteRobust, we recover only the failed role and reconnect it to living ones, thereby eliminating the full-restart overhead including rollout replay and initialization delay. We present RobustRL, the first comprehensive robust system to handle GPU machine errors for RL post-training Effective Training Time Ratio improvement. (1) \textit{Detect}. We implement role-aware monitoring to distinguish actual failures from role-specific behaviors to avoid the false positive and delayed detection. (2) \textit{Restart}. For trainers, we implement a non-disruptive recovery where rollouts persist state and continue trajectory generation, while the trainer is rapidly restored via rollout warm standbys. For rollout, we perform isolated machine replacement without interrupting the RL task. (3) \textit{Reconnect}. We replace static collective communication with dynamic, UCX-based (Unified Communication X) point-to-point communication, enabling immediate weight synchronization between recovered roles. In an RL training task on a 256-GPU cluster with Qwen3-8B-Math workload under 10\% failure injection frequency, RobustRL can achieve an ETTR of over 80\% compared with the 60\% in ByteRobust and achieves 8.4\%-17.4\% faster in end-to-end training time.

Miao Ye, Hongwen Hu, Xiaoli Wang et al.

The cross-domain multicast routing problem in a software-defined wireless network with multiple controllers is a classic NP-hard optimization problem. As the network size increases, designing and implementing cross-domain multicast routing paths in the network requires not only designing efficient solution algorithms to obtain the optimal cross-domain multicast tree but also ensuring the timely and flexible acquisition and maintenance of global network state information. However, existing solutions have a limited ability to sense the network traffic state, affecting the quality of service of multicast services. In addition, these methods have difficulty adapting to the highly dynamically changing network states and have slow convergence speeds. To this end, this paper aims to design and implement a multiagent deep reinforcement learning based cross-domain multicast routing method for SDWN with multicontroller domains. First, a multicontroller communication mechanism and a multicast group management module are designed to transfer and synchronize network information between different control domains of the SDWN, thus effectively managing the joining and classification of members in the cross-domain multicast group. Second, a theoretical analysis and proof show that the optimal cross-domain multicast tree includes an interdomain multicast tree and an intradomain multicast tree. An agent is established for each controller, and a cooperation mechanism between multiple agents is designed to effectively optimize cross-domain multicast routing and ensure consistency and validity in the representation of network state information for cross-domain multicast routing decisions. Third, a multiagent reinforcement learning-based method that combines online and offline training is designed to reduce the dependence on the real-time environment and increase the convergence speed of multiple agents.

Miao Ye, Zhibang Jiang, Xingsi Xue et al.

Attention-based transformers have played an important role in wireless sensor network (WSN) timing anomaly detection due to their ability to capture long-term dependencies. However, there are several issues that must be addressed, such as the fact that their ability to capture long-term dependencies is not completely reliable, their computational complexity levels are high, and the spatiotemporal features of WSN timing data are not sufficiently extracted for detecting the correlation anomalies of multinode WSN timing data. To address these limitations, this paper proposes a WSN anomaly detection method that integrates frequency-domain features with dynamic graph neural networks (GNN) under a designed self-encoder reconstruction framework. First, the discrete wavelet transform effectively decomposes trend and seasonal components of time series to solve the poor long-term reliability of transformers. Second, a frequency-domain attention mechanism is designed to make full use of the difference between the amplitude distributions of normal data and anomalous data in this domain. Finally, a multimodal fusion-based dynamic graph convolutional network (MFDGCN) is designed by combining an attention mechanism and a graph convolutional network (GCN) to adaptively extract spatial correlation features. A series of experiments conducted on public datasets and their results demonstrate that the anomaly detection method designed in this paper exhibits superior precision and recall than the existing methods do, with an F1 score of 93.5%, representing an improvement of 2.9% over that of the existing models.

Miao Ye, Jihao Zheng, Qiuxiang Jiang et al.

The existing segment routing (SR) methods need to determine the routing first and then use path segmentation approaches to select swap nodes to form a segment routing path (SRP). They require re-segmentation of the path when the routing changes. Furthermore, they do not consider the flow table issuance time, which cannot maximize the speed of issuance flow table. To address these issues, this paper establishes an optimization model that can simultaneously form routing strategies and path segmentation strategies for selecting the appropriate swap nodes to reduce flow table issuance time. It also designs an intelligent segment routing algorithm based on deep reinforcement learning (DRL-SR) to solve the proposed model. First, a traffic matrix is designed as the state space for the deep reinforcement learning agent; this matrix includes multiple QoS performance indicators, flow table issuance time overhead and SR label stack depth. Second, the action selection strategy and corresponding reward function are designed, where the agent selects the next node considering the routing; in addition, the action selection strategy whether the newly added node is selected as the swap node and the corresponding reward function are designed considering the time cost factor for the controller to issue the flow table to the swap node. Finally, a series of experiments and their results show that, compared with the existing methods, the designed segmented route optimization model and the intelligent solution algorithm (DRL-SR) can reduce the time overhead required to complete the segmented route establishment task while optimizing performance metrics such as throughput, delays and packet losses.