Throughput Capacity of Hybrid Multi-channel Vehicular Ad Hoc Networks

Lin ChenAdvances 2015 · DOI

Abstract

The throughput capacity of hybrid multi-channel Vehicular Ad hoc Networks is addressed in the paper. Both the theoretic upper and lower bounds are obtained. Moreover, both single road and grid road are analyzed. The results show that the movement has little influence on the throughput capacity of hybrid multi-channel vehicular ad hoc networks.

Introduction

In wireless networking, the landmark study by Gupta and Kumar (2000) established a "disappointing result": as an ad hoc network scales, the per-node throughput drops toward zero due to interference and relaying overhead. Vehicular Ad Hoc Networks (VANETs) exacerbate this issue with high-speed mobility and constrained topologies. To mitigate these bottlenecks, researchers have proposed multi-channel communication and hybrid architectures incorporating fixed roadside base stations. This paper rigorously evaluates the maximum possible throughput of a VANET utilizing both solutions simultaneously.

System Model and Methods

The study employs asymptotic analysis to determine network behavior as the number of vehicles ($n$) grows. Two primary road topologies are evaluated: a linear Single Road model representing a highway and a Grid Road model representing a city's intersecting streets.

Grid road topology model.

The mathematical framework is built upon the $(m, c)$ network model, where each node possesses $m$ interfaces and can access $c$ channels. A key component is the Hybrid Routing strategy: data from a source vehicle is transmitted to the nearest base station, forwarded through a high-speed wired backbone, and finally delivered via the base station closest to the destination vehicle.

Hybrid routing architecture in a single road scenario.

Vehicle mobility is modeled as a Poisson process. The paper derives tight throughput capacity by aligning theoretical upper bounds (physical limits) with lower bounds (achievable via specific scheduling and routing schemes).

Key Insights and Scaling Laws

The analysis reveals several critical findings regarding the scaling of hybrid VANETs:

  • Movement Neutrality: Contrary to intuition, vehicle movement has negligible impact on throughput capacity in hybrid models. The base station backbone effectively decouples long-distance communication from the dynamic local topology.
  • The Critical Ratio ($c/m$): Capacity is governed by the relationship between available channels ($c$) and hardware interfaces ($m$).
  • Interference-Limited Regime: When the ratio $c/m$ is small ($O(\ln n)$), the network capacity is: $$\Theta\left( \frac{nW}{\ln n} \right) \text{ bits/s}$$ In this scenario, the bottleneck is the radio interference between nodes.

  • Interface-Limited Regime: When the ratio $c/m$ is large ($\Omega(\ln n)$), the capacity is: $$\Theta\left( \frac{nWm}{c} \right) \text{ bits/s}$$ Here, the bottleneck is the physical number of transmissions a node can handle simultaneously.

  • Topology Independence: Interestingly, the scaling laws for a single highway and a complex city grid are identical. In a hybrid system, the base stations effectively "flatten" the geographic constraints of the road.

These results demonstrate that the integration of base stations allows network capacity to scale linearly with the number of base stations, effectively bypassing the capacity collapse characteristic of pure ad hoc networks.

Limitations and Future Work

While providing a robust theoretical foundation, the model assumes an idealized backbone with infinite capacity and a simplified Poisson mobility model. Real-world factors such as backhaul bottlenecks, signal fading, and complex traffic patterns (e.g., congestion) are not fully captured. Future research may explore the impact of non-uniform base station placement and more realistic physical layer characteristics in urban environments.

Summary updated

Abstract

The throughput capacity of hybrid multi-channel Vehicular Ad hoc Networks is addressed in the paper. Both the theoretic upper and lower bounds are obtained. Moreover, both single road and grid road are analyzed. The results show that the movement has little influence on the throughput capacity of hybrid multi-channel vehicular ad hoc networks.
BibTeX
@inproceedings{chen2015throughput,
  author = {Lin Chen},
  title = {{Throughput Capacity of Hybrid Multi-channel Vehicular Ad Hoc Networks}},
  booktitle = {Advances},
  year = {2015},
  doi = {10.2991/meici-15.2015.201},
}
Figure