A Trajectory Analysis-based Critical Intersection Discovery Approach for Roadside Unit Deployment in Vehicular Networks

The deployment of roadside units plays an important role in the performance of vehicular networks. The main challenge of the roadside unit deployment algorithm is the trade-off between the service performance and the deployment cost. To address this challenge, this paper proposes a road network critical intersection discovery scheme based on real vehicle trajectory mining for the deployment of roadside units. In the proposed scheme, we first convert the trajectory of each vehicle to a sequence of intersections it visited. Then we transform the roadside unit deployment problem into a weighted set cover problem by selecting some critical intersections from the intersection sequence as the deployment locations of roadside units and introducing a weighted metric to trade off the network performance and deployment cost. We prove this problem is NP-hard and propose a greedy algorithm to solve it and discover the critical intersections to deploy roadside units. The simulation results show the proposed approach outperforms some existing schemes.