LSTM-Based Proactive Congestion Management for Internet of Vehicle Networks

TL;DR

IoV networks suffer from channel congestion that degrades delivery of safety messages. The authors propose a proactive open-loop congestion management framework that combines a stacked LSTM sequence-to-sequence predictor trained on NS3-SUMO generated data with a K-means-based packet prioritization scheme using TTL and priority headers. Key contributions include a realistic IoV congestion dataset, a two-layer stacked LSTM achieving about 99% test accuracy with RMSE around 2147, and a two-class packet clustering validated by silhouette scores. The approach enables real-time congestion forecasting and prioritized safety-message delivery, with potential integration into future RAN architectures such as O-RAN.

Abstract

Vehicle-to-everything (V2X) networks support a variety of safety, entertainment, and commercial applications. This is realized by applying the principles of the Internet of Vehicles (IoV) to facilitate connectivity among vehicles and between vehicles and roadside units (RSUs). Network congestion management is essential for IoVs and it represents a significant concern due to its impact on improving the efficiency of transportation systems and providing reliable communication among vehicles for the timely delivery of safety-critical packets. This paper introduces a framework for proactive congestion management for IoV networks. We generate congestion scenarios and a data set to predict the congestion using LSTM. We present the framework and the packet congestion dataset. Simulation results using SUMO with NS3 demonstrate the effectiveness of the framework for forecasting IoV network congestion and clustering/prioritizing packets employing recurrent neural networks.

Figures (6)

• Figure 1: IoV-based vehicular networks congestion scenario with the proposed solution components.
• Figure 2: Overall process of proposed strategy.
• Figure 3: LSTM Memory Block structure.
• Figure 4: Predicted test verus test data.
• Figure 5: The RMSPE performance of the proposed LSTM solution versus RNN benchmark technique.