GraphEdge: Dynamic Graph Partition and Task Scheduling for GNNs Computing in Edge Network

2025-04-22

AI Summary

To address the high communication overhead, latency, and energy consumption in edge-computing-based Graph Neural Network (GNN) inference—caused by strong inter-dependencies among user data—this paper proposes a co-optimization framework tailored for dynamic graph topologies. We introduce HiCut, a hierarchical graph partitioning algorithm that performs GNN-aggregation-aware weakly-coupled subgraph division. Building upon this, we design DRLGO, a subgraph-granularity deep reinforcement learning–based offloading and scheduling algorithm that jointly optimizes inference latency, energy consumption, and inter-server communication. Our approach tightly integrates GNN inference with task orchestration in distributed edge architectures. Experimental results demonstrate a 47.3% reduction in cross-edge-server communication overhead, a 39.1% decrease in end-to-end latency, and a 32.8% reduction in energy consumption, while maintaining robustness under dynamic graph topologies.

Technology Category

• Machine Learning: Graph-based Machine Learning
• Search and Optimization: Distributed Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Federated Web and WoT systems, including distributed, federated and edge-based data processing
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

With the exponential growth of Internet of Things (IoT) devices, edge computing (EC) is gradually playing an important role in providing cost-effective services. However, existing approaches struggle to perform well in graph-structured scenarios where user data is correlated, such as traffic flow prediction and social relationship recommender systems. In particular, graph neural network (GNN)-based approaches lead to expensive server communication cost. To address this problem, we propose GraphEdge, an efficient GNN-based EC architecture. It considers the EC system of GNN tasks, where there are associations between users and it needs to take into account the task data of its neighbors when processing the tasks of a user. Specifically, the architecture first perceives the user topology and represents their data associations as a graph layout at each time step. Then the graph layout is optimized by calling our proposed hierarchical traversal graph cut algorithm (HiCut), which cuts the graph layout into multiple weakly associated subgraphs based on the aggregation characteristics of GNN, and the communication cost between different subgraphs during GNN inference is minimized. Finally, based on the optimized graph layout, our proposed deep reinforcement learning (DRL) based graph offloading algorithm (DRLGO) is executed to obtain the optimal offloading strategy for the tasks of users, the offloading strategy is subgraph-based, it tries to offload user tasks in a subgraph to the same edge server as possible while minimizing the task processing time and energy consumption of the EC system. Experimental results show the good effectiveness and dynamic adaptation of our proposed architecture and it also performs well even in dynamic scenarios.

Problem

• Minimize server communication cost in GNN-based edge computing
• Optimize graph layout for weakly associated subgraphs in GNN tasks
• Reduce task processing time and energy consumption in edge networks

Innovation

• Dynamic graph partition using hierarchical traversal algorithm
• DRL-based subgraph offloading for edge servers
• Minimizes GNN communication and processing costs