Context-Aware Rate Adaptation for Vehicular Adhoc Networks
Pravin Shankar

Vehicular Adhoc Networks (or VANETs) enable novel safety and
entertainment applications which present high performance requirements
under varying channel conditions. Rate adaptation is a critical
component to ensure optimal system performance. Current rate adaptation
mechanisms in 802.11 wireless networks try to estimate the channel
quality by means of physical and link layer metrics and select the
optimal transmission rate accordingly. Outdoor mobile environments such
as vehicular adhoc networks present channel conditions that vary rapidly
because of fading and mobility. Current rate adaptation mechanisms incur
a time delay in channel estimation owing to the estimation window. We
propose a rate adaptation mechanism for VANETs which makes use of
context information from the environment to proactively select the
optimal transmission rate. Examples of such context information are
vehicle speed and distance from neighboring vehicle. By means of outdoor
experiments, we model the effect of mobility and fading on bit error
rate for different transmission rates, and using this model we design
Context Aware Rate Adaptation (CARA), a novel rate adaptation algorithm
for vehicular adhoc networks. Our evaluation results in outdoor high
mobility scenarios show that CARA achieves better throughput than
existing well-known rate adaptation solutions (AARF, Onoe and
SampleRate) in the tested scenarios.