Broadcast reception rates and effects of priority access in 802.11-based vehicular ad-hoc networks

Marc Torrent‐Moreno, Daniel Jiang, Hannes Hartenstein2004 · DOI

TL;DR

This paper investigates broadcast reception in vehicular ad-hoc networks (VANETs), crucial for safety applications. Using simulations, it demonstrates that IEEE 802.11e's priority access mechanisms can significantly improve broadcast reception rates and reduce channel access times, especially in saturated network conditions. However, the benefits are less pronounced and overall reception rates are lower when realistic, non-deterministic radio propagation models (like Nakagami) are used, highlighting the need for advanced retransmission strategies.

What Problem Does This Paper Address?

Vehicular ad-hoc networks (VANETs) hold immense promise for enhancing road safety by enabling vehicles to communicate directly with each other. A key application is broadcasting critical information, such as emergency braking or hazard warnings, to nearby vehicles. However, in a dynamic and potentially congested environment like a highway, ensuring that these broadcast messages are reliably received by all intended recipients is a significant challenge. This paper tackles three core questions:

  1. How does the distance between a sender and receiver affect the probability of a broadcast message being received?
  2. Can priority access mechanisms, like those in IEEE 802.11e, improve the reception rate of important warnings?
  3. How do real-world radio signal fluctuations (fading) impact these reception rates and the effectiveness of priority access?

The authors aim to provide quantitative insights into these issues, particularly in saturated network conditions where bandwidth is limited and collisions are frequent.

Methods

The researchers employed a simulation-based approach using the ns-2 network simulator. They implemented a modified version of the IEEE 802.11a standard, tailored for the 5.9 GHz band with 10 MHz channels, as proposed for DSRC (Dedicated Short Range Communications).

Key methodological aspects include:

  • Priority Access Mechanism: They implemented a priority access mechanism based on the IEEE 802.11e Enhanced Distributed Channel Access (EDCA) concept. This involves adjusting interframe spaces (AIFS) and contention window sizes (CW) to give higher priority traffic preferential access to the channel.
  • Radio Propagation Models: Two models were used:
  • Two-Ray Ground Model: A deterministic model that assumes a simplified, predictable signal propagation.
  • Nakagami Distribution: A non-deterministic, statistical model that better captures the effects of radio channel fading and signal strength fluctuations, considered more realistic for VANET environments.
  • Simulation Scenarios:
    • Static Scenario: Used to deeply analyze the impact of protocol parameters (AIFS, CW) on reception probability under saturated conditions.
  • Dynamic Scenario: Modeled a realistic highway environment with moving vehicles across multiple lanes to assess performance under mobility and varying traffic loads.
  • Metrics: The study focused on two primary metrics:
  • Channel Access Time: The time from packet creation to its transmission on the MAC layer.
  • Probability of Reception: The percentage of vehicles successfully receiving a broadcast message within a specified distance from the sender.

Datasets

Not discussed in the paper. The authors generated their own simulation data based on the defined scenarios and propagation models.

Key Insights and Ideas

  • Priority Pays Off (Under Ideal Conditions): The IEEE 802.11e EDCA-inspired priority mechanism demonstrably improves broadcast reception rates and significantly reduces channel access times, especially when the network is saturated. This is achieved by allowing prioritized nodes to access the channel more frequently.
  • The Reality of Fading: Realistic radio propagation models, like Nakagami's distribution, paint a less optimistic picture. While priority still offers benefits, the overall probability of reception drops considerably compared to the idealized two-ray ground model. This highlights that signal fluctuations are a major hurdle for reliable VANET communication.
  • Hidden Terminal Problem is Amplified: The non-deterministic nature of fading exacerbates the hidden terminal problem. Nodes may fail to sense transmissions from distant hidden terminals, leading to collisions and reduced reception, even when the priority mechanism is in place.
  • Saturation is the Norm: The paper emphasizes that VANETs, especially for safety applications, are likely to operate under saturation conditions. This makes understanding and mitigating congestion effects paramount.
  • Beyond Basic Priority: The findings strongly suggest that simple priority mechanisms alone are insufficient for robust safety-critical applications in realistic VANET environments.

Key Results

  • Channel Access Time: In the most saturated scenario (200m range, 500B packets, deterministic model), priority access reduced channel access time by approximately 4 times (1.6ms vs. 4.8ms).
  • Probability of Reception (Deterministic Model): Averaged over distance, priority nodes achieved 9.6% to 16.3% higher reception probabilities than non-priority nodes in moderate congestion. In the most congested scenario, this difference reached up to 150%.
  • Probability of Reception (Nakagami Model): When using the Nakagami model, the overall probability of reception for both priority and non-priority nodes was significantly lower than with the two-ray ground model. The benefit of priority access was still present but less pronounced.
  • Channel Idle Time: The Nakagami model resulted in a significantly lower channel idle time (4.4%) compared to the two-ray ground model (10.8%), indicating a more persistently busy channel due to signal fluctuations.

Limitations and Future Work

  • Limitations:
    • The study primarily focuses on broadcast traffic and does not deeply explore unicast or mixed traffic scenarios.
  • The Nakagami distribution implementation did not incorporate spatial or temporal correlations, meaning each node's signal strength was determined independently. This is a simplification of real-world fading.
  • The "capture effect" (where a stronger signal can overcome a weaker interfering one) was modeled with a fixed SNR, which might not fully represent all scenarios.
  • Future Work:
    • Repetition and Multi-hop Retransmission: The authors strongly advocate for designing robust repetition or multi-hop retransmission strategies to enhance the reliability of high-priority broadcasts, especially under fading conditions.
  • Traffic Modeling: More sophisticated traffic models are needed to better represent real-world VANET applications and packet generation patterns.
  • Mobility and Correlation: Further investigation into the impact of mobility, spatial, and temporal correlations in realistic radio models is crucial.
  • Topology Control: Exploring how to adjust transmission power for better reception while managing interference is an open area.
  • Comparison of Factors: Understanding the relative contributions of mobility, dimensionality, and radio fluctuations to performance degradation. The authors suggest that radio fluctuations can have a greater impact than node mobility alone.

Summary updated

Abstract

One key usage of VANET is to support vehicle safety applications. This use case is characterized by the prominence of broadcasts in scaled settings. In this context, we try to answer the following questions: i) what is the probability of reception of a broadcast message by another car depending on its distance to the sender, ii) how to give priority access and an improved reception rate for important warnings, e.g., sent out in an emergency situation, and iii) how are the above two results affected by signal strength fluctuations caused by radio channel fading? We quantify via simulation the probability of reception for the two-ray-ground propagation model as well as for the Nakagami distribution in saturated environments. By making use of some IEEE 802.11e EDCA mechanisms for priority access, we do not only quantify how channel access times can be reduced but also demonstrate how improved reception rates can be achieved. Our results show that the mechanisms for priority access are successful under the two-way-ground model. However, with a non-deterministic radio propagation model like Nakagami's distribution the benefit is still obvious but the general level of probability of reception is much smaller compared to two-ray-ground model. The results indicate that -- particularly for safety-critical and sensor network type of applications -- the proper design of repetition or multi-hop retransmission strategies represents an important aspect of future work for robustness and network stability of vehicular ad hoc networks.
BibTeX
@misc{torrentmoreno2004broadcast,
  author = {Marc Torrent‐Moreno and Daniel Jiang and Hannes Hartenstein},
  title = {{Broadcast reception rates and effects of priority access in 802.11-based vehicular ad-hoc networks}},
  year = {2004},
  doi = {10.1145/1023875.1023878},
}
Figure