Increasing Intelligence in Inter-Vehicle Communications to Reduce Traffic Congestions: Experiments in Urban and Highway Environments

Meneguette, Rodolfo I.; Filho, Geraldo P. R.; Guidoni, Daniel L.; Pessin, Gustavo; Villas, Leandro A.; Ueyama, Jó · 2016 · OpenAlex-citations

DOI: 10.1371/journal.pone.0159110

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of traffic congestion in Intelligent Transportation Systems (ITS) by proposing INCIDEnT, an intelligent protocol for congestion detection and mitigation using Inter-Vehicle Communication (IVC). Traffic congestion causes significant economic losses and driver stress, while existing data dissemination protocols in Vehicular Ad Hoc Networks (VANETs) often struggle with broadcast storms in dense networks or intermittent connectivity in sparse networks. The authors aim to reduce average trip times, fuel consumption, and CO emissions by enabling vehicles to collaboratively detect congestion levels and suggest alternative routes. The INCIDEnT protocol utilizes a Multi-Layer Perceptron Artificial Neural Network (ANN) to detect and classify road congestion into three levels: Free, Moderate, and Congested. The ANN takes vehicle speed (derived from GPS) and neighbor density (calculated from periodic beacon messages) as inputs. The network topology consists of two input neurons, a hidden layer with four neurons, and a single output neuron using a hyperbolic tangent activation function. The system employs a back-propagation algorithm for training and performs local classification every two seconds. Upon detecting congestion, the protocol disseminates this information to nearby vehicles and suggests new routes to avoid the affected areas, aiming to balance traffic flow without requiring centralized infrastructure. The study evaluates INCIDEnT through simulations in both urban and highway environments, comparing its performance against metrics such as network density, dissemination coverage, communication delay, transmitted packets, CO emissions, fuel consumption, and trip time. The results demonstrate that INCIDEnT achieves a high hit rate in classifying congestion levels while maintaining low overhead and short delays. Specifically, the protocol effectively propagates congestion information over significant distances, ensuring broad coverage. The findings indicate that the use of INCIDEnT leads to a measurable reduction in average trip times, decreased fuel consumption, and lower CO emissions compared to scenarios without such intelligent routing assistance. The significance of this work lies in its ability to provide a decentralized, efficient solution for traffic management that adapts to dynamic network conditions. By leveraging neural networks, INCIDEnT offers superior generalization and learning capabilities compared to traditional static protocols, allowing it to handle both dense urban traffic and sparse highway conditions. The protocol successfully mitigates the broadcast storm problem while maintaining connectivity in sparse networks. These results suggest that integrating intelligent, machine-learning-based detection mechanisms into VANETs can significantly enhance traffic efficiency and environmental sustainability, providing a scalable approach for future ITS implementations.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• v2x connected vehicle