[Paper Review] Evaluation of Synchronous and Asynchronous Reactive Distributed Congestion Control Algorithms for the ITS G5 Vehicular Systems

Oyunchimeg Shagdar|arXiv (Cornell University)|Jun 25, 2015

Vehicular Ad Hoc Networks (VANETs)1 references3 citations

TL;DR

This paper evaluates synchronous and asynchronous reactive distributed congestion control (DCC) algorithms for ITS G5 vehicular networks, showing that synchronous DCC causes harmful channel load oscillations due to coordinated reactions. By introducing randomized rate setting in asynchronous DCC, the authors reduce oscillations and significantly improve network performance, demonstrating a practical enhancement to IEEE 802.11p-based V2X systems under high density.

ABSTRACT

The IEEE 802.11p is the technology dedicated to vehicular communications to support road safety, efficiency, and comfort applications. A large number of research activities have been carried out to study the characteristics of the IEEE 802.11p. The key weakness of the IEEE 802.11p is the channel congestion issue, where the wireless channel gets saturated when the road density increases. The European Telecommunications Standardization Institute (ETSI) is in the progress of studying the channel congestion problem and proposed so-called Reactive Distributed Congestion Control (DCC) algorithm as a solution to the congestion issue. In this report we investigate the impacts of the Reactive DCC mechanism in comparison to the conventional IEEE 802.11p with no congestion control. Our study shows that the Reactive DCC scheme creates oscillation on channel load that consequently degrades communication performance. The results reveal that the channel load oscillation is due to the fact that in the Reactive DCC, the individual CAM (Cooperative Awareness Message) controllers react to the channel congestion in a synchronized manner. To reduce the oscillation, in this report we propose a simple extension to Reactive DCC, Asynchronous Reactive DCC, in which the individual CAM controllers adopt randomized rate setting, which can significantly reduce the oscillation and improve the network performance.

Motivation & Objective

• To analyze the performance impact of reactive distributed congestion control (DCC) in IEEE 802.11p-based ITS G5 vehicular networks.
• To identify the root cause of performance degradation in synchronous DCC, particularly channel load oscillations.
• To propose and evaluate an asynchronous variant of DCC that mitigates oscillation through randomized rate setting.
• To compare the performance of conventional IEEE 802.11p with no congestion control against both synchronous and asynchronous DCC schemes.
• To improve communication reliability and efficiency in high-density vehicular environments through enhanced congestion control.

Proposed method

• Simulates a vehicular network using the IEEE 802.11p MAC layer with CAM (Cooperative Awareness Message) transmission under varying road densities.
• Implements a reactive DCC mechanism where each vehicle adjusts its CAM transmission rate based on local channel load sensing.
• Introduces an asynchronous DCC variant by adding randomization to the rate-setting function of individual CAM controllers.
• Uses a distributed, decentralized control approach where each node independently reacts to congestion without global coordination.
• Employs simulation-based evaluation to compare channel load stability, packet delivery ratio, and delay across configurations.
• Analyzes the dynamics of channel load over time to detect oscillation patterns in synchronous vs. asynchronous operation.

Experimental results

Research questions

• RQ1 What causes channel load oscillations in synchronous reactive DCC for ITS G5 systems?
• RQ2 How does the synchronization of individual CAM controllers affect network performance under high vehicle density?
• RQ3 To what extent can randomization in rate setting reduce oscillation and improve stability in distributed congestion control?
• RQ4 How does the performance of asynchronous DCC compare to both conventional IEEE 802.11p and synchronous DCC in terms of throughput and delay?
• RQ5 Can a simple extension to the reactive DCC algorithm significantly enhance reliability in dense vehicular networks?

Key findings

• Synchronous reactive DCC causes significant channel load oscillations due to coordinated reactions among vehicles, degrading network performance.
• The oscillation leads to unstable throughput and increased delay, particularly under high road density conditions.
• Introducing randomized rate setting in the asynchronous DCC variant effectively reduces oscillation amplitudes and stabilizes channel load.
• Asynchronous DCC achieves better overall network performance, including improved delivery ratios and reduced jitter, compared to synchronous DCC.
• The proposed asynchronous extension requires no central coordination and is easily deployable in existing IEEE 802.11p-based V2X systems.
• Simulation results confirm that the asynchronous approach maintains stable performance even at high vehicle densities, outperforming both baseline IEEE 802.11p and synchronous DCC.