Abstract:Because traffic flow data has complex spatial dependence and temporal correlation, it is a challenging problem for researchers in the field of Intelligent Transportation to accurately predict traffic flow by analyzing spatio-temporal traffic data. Based on the idea of spatio-temporal data fusion, fully considering the correlation of traffic flow data in the time dimension and the dependence of spatial structure, this paper proposes a new spatio-temporal traffic flow prediction model based on Graph Neural Network (GNN), which is called Bidirectional-Graph Recurrent Convolutional Network (Bi-GRCN). First, aiming at the spatial dependence between traffic flow data and traffic roads, Graph Convolution Network (GCN) which can directly analyze complex non-Euclidean space data is selected for spatial dependence modeling, to extract the spatial dependence characteristics. Second, considering the temporal dependence of traffic flow data on historical data and future data in its time-series period, Bidirectional-Gate Recurrent Unit (Bi-GRU) is used to process historical data and future data at the same time, to learn the temporal correlation characteristics of data in the bidirectional time dimension from the input data. Finally, the full connection layer is used to fuse the extracted spatial features and the learned temporal features to optimize the prediction results so that the Bi-GRCN model can better extract the spatial dependence and temporal correlation of traffic flow data. The experimental results show that the model can not only effectively predict the short-term traffic flow but also get a good prediction effect in the medium- and long-term traffic flow prediction.

RouteNet-Fermi: Network Modeling With GNN (Analysis And Re-implementation)

RouteNet-Fermi: Network Modeling with Graph Neural Networks

Traffic Analysis of Partially Observable Bearer Network under RouteNet Based GNN Model

Xnet: Modeling Network Performance with Graph Neural Networks

Xnet: Improving Expressiveness and Granularity for Network Modeling with Graph Neural Networks

Deeper and Wider Networks for Performance Metrics Prediction in Communication Networks

Design of Retransmission Mechanism for Decentralized Inference with Graph Neural Networks

Efficient Metropolitan Traffic Prediction Based on Graph Recurrent Neural Network.

G-Routing: Graph Neural Networks-Based Flexible Online Routing

DeepComNet: Performance evaluation of network topologies using graph-based deep learning

Graph Neural Networks for Routing Optimization: Challenges and Opportunities

Routing in Small Satellite Networks: A GNN-based Learning Approach

Graph Neural Networks for Intelligent Modelling in Network Management and Orchestration: A Survey on Communications

Graph Neural Modeling of Network Flows

SST-GNN: Simplified Spatio-temporal Traffic forecasting model using Graph Neural Network

Graph-neural-network-based delay estimation for communication networks with heterogeneous scheduling policies

Bi-GRCN: A Spatio-Temporal Traffic Flow Prediction Model Based on Graph Neural Network

Graph Neural Networks Inspired by Classical Iterative Algorithms

Investigation of a GNN approach to mitigate congestion in a realistic MANET scenario

HighwayGraph: Modelling Long-distance Node Relations for Improving General Graph Neural Network.

Building a Graph-based Deep Learning network model from captured traffic traces