Spatio-Temporal Motifs for Optimized Vehicle-to-Vehicle (V2V) Communications

TL;DR

This paper introduces a novel spatio-temporal caching strategy for V2V networks using temporal graph motifs, significantly improving data dissemination efficiency in vehicular communication systems.

Contribution

It proposes a new spatio-temporal caching method based on temporal graph motifs, addressing large-scale variations in V2V networks for better content dissemination.

Findings

Increases average data rate by 45% in simulations.

Effective content placement using V2V motifs.

Improves network performance in real city traces.

Abstract

Caching popular contents in vehicle-to-vehicle (V2V) communication networks is expected to play an important role in road traffic management, the realization of intelligent transportation systems (ITSs), and the delivery of multimedia content across vehicles. However, for effective caching, the network must dynamically choose the optimal set of cars that will cache popular content and disseminate it in the entire network. However, most of the existing prior art on V2V caching is restricted to cache placement that is solely based on location and user demands and does not account for the large-scale spatio-temporal variations in V2V communication networks. In contrast, in this paper, a novel spatio-temporal caching strategy is proposed based on the notion of temporal graph motifs that can capture spatio-temporal communication patterns in V2V networks. It is shown that, by identifying such V2V motifs, the network can find sub-optimal content placement strategies for effective content dissemination across a vehicular network. Simulation results using real traces from the city of Cologne show that the proposed approach can increase the average data rate by $45\%$ for different network scenarios.