Joint Network Selection and Traffic Allocation in Multi-Access Edge Computing-Based Vehicular Crowdsensing
Luning Liu, Luhan Wang, Xiangming Wen · 2020
The emergence of various applications for intelligent vehicles poses technical challenges on both communication and computing in vehicular crowdsensing systems due to large data volume and high performance requirements. In this paper, an edge-assisted data uploading architecture in vehicular crowdsensing is presented, where vehicles can access to different types of networks simultaneously to satisfy various requirements of vehicular crowdsensing applications. In order to enhance the quality of service (QoS) of vehicles and network utilization, an effective network selection and traffic allocation scheme is developed. In particular, we formulate a two-objective optimization problem to maximize user satisfaction, which is designed to characterize the performance of selected networks. To tackle the problem efficiently, we first decouple it into network selection and traffic allocation sub-problems, and then solve these sub-problems using low-complexity algorithms based on particle swarm optimization (PSO) and convex optimization theories, respectively. Furthermore, an edge-based alternate search algorithm is designed to jointly solve the two intertwined sub-problems. Finally, by conducting extensive simulations, the superiority of the proposed scheme is validated.