Temporal Correlation of Interference in Vehicular Networks with Shifted-Exponential Time Headways

TL;DR

This paper studies how the shifted-exponential distribution of vehicle headways affects the temporal correlation of interference in vehicular networks, revealing that ignoring the shift leads to overestimation, especially in high traffic and short time-lags.

Contribution

It introduces a model considering shifted-exponential headways and quantifies the impact on interference correlation, improving accuracy over Poisson-based models.

Findings

Ignoring the shift overestimates interference correlation.

Overestimation is significant at high traffic conditions.

The model provides more accurate correlation estimates.

Abstract

We consider a one-dimensional vehicular network where the time headway (time difference between successive vehicles as they pass a point on the roadway) follows the shifted-exponential distribution. We show that neglecting the impact of shift in the deployment model, which degenerates the distribution of vehicles to a Poisson Point Process, overestimates the temporal correlation of interference at the origin. The estimation error becomes large at high traffic conditions and small time-lags.