Abstract
Vehicular Ad Hoc Networks (VANETs) are characterized by high node mobility and rapid topology changes, which pose significant challenges for reliable data routing. Traditional protocols like Ad hoc On-demand Distance Vector (AODV) prioritize the shortest path, often leading to frequent link breakages in dynamic environments. This paper introduces AODV-R, an enhanced protocol that utilizes a hybrid mobility model to predict link reliability. By selecting routes based on stability rather than hop count, AODV-R improves packet delivery ratios and reduces network overhead caused by route repairs.
Introduction
Intelligent Transportation Systems (ITS) rely on seamless communication between vehicles (V2V) and between vehicles and roadside infrastructure (V2I). However, the high-speed nature of vehicular movement makes maintaining stable communication links difficult. Standard routing protocols designed for mobile ad hoc networks (MANETs) often fail to account for the specific constraints of vehicular traffic, such as road geometry and driver behavior.
Problem Statement
In VANETs, the "shortest path" metric used by protocols like AODV is often suboptimal. For instance, two vehicles moving in opposite directions might provide the shortest hop but will quickly move out of range, causing a link failure. These frequent breakages lead to dropped packets and increased control overhead as the network repeatedly initiates route discovery.
Methodology
The proposed AODV-R protocol addresses these issues through a two-pronged approach: a mathematical reliability model and specific protocol modifications.
Hybrid Reliability Model
The authors developed a model that combines macroscopic traffic flow data with microscopic driver behavior parameters (e.g., acceleration/deceleration). By utilizing real-time GPS data—including position, speed, and direction—the model calculates the probability that a link between two nodes will remain active for a predefined prediction interval ($T_p$). The model employs the Gauss error function to translate these physical variables into a reliability score between 0 and 1.
AODV-R Protocol Modifications
To implement this model, the standard AODV packet structures were extended. Route Request (RREQ) and Route Reply (RREP) packets now include fields for coordinates, speed, and direction. As an RREQ propagates, it accumulates a "Route Reliability" score, representing the product of all individual link reliabilities along the path. Unlike standard AODV, the destination node in AODV-R selects the path with the highest reliability score rather than the first one received.
Experimental Setup
The performance of AODV-R was evaluated using the OMNet++ simulator. The simulation environment consisted of a 5,000-meter three-lane highway with bidirectional traffic. Traffic density was maintained at 10 vehicles per lane, with average speeds ranging from 40 to 80 km/h. The network was tested using UDP packets of varying sizes (500 to 3,000 bytes) to assess the protocol's robustness.
Results and Discussion
Simulation results indicate that AODV-R significantly outperforms standard AODV in high-mobility scenarios:
- Packet Delivery Ratio (PDR): AODV-R maintained a much higher delivery rate. At high speeds (140 km/h in the fast lane), AODV-R achieved ~56% PDR compared to AODV’s ~37%.
- Link Stability: The protocol saw a dramatic reduction in link breakages—often nearly half the failures of standard AODV—because it avoided fragile links from the start.
- Trade-offs: The focus on reliability introduces a slight increase in end-to-end delay and routing overhead due to the additional data processing and larger packet headers required for GPS and velocity information.
Conclusion
AODV-R demonstrates that incorporating mobility awareness into routing decisions can substantially enhance the performance of vehicular networks. While the protocol requires accurate GPS data and introduces some latency, the gains in reliability are critical for safety-oriented ITS applications. Future work will focus on integrating environmental factors like signal-to-noise ratios and testing the protocol in complex urban grid environments.
Summary updated