Collaborative vehicle sensing in bus networks: A Stackelberg game approach

Collecting urban data has become a crucial issue for smart city. Unlike mobile devices and other vehicles, buses have unique characteristics: fixed mobility trajectories, strong periodicity, uniform sensor interfaces and low influence of exposure privacy. We leverage the characteristics and consider a collaborative vehicle sensing paradigm in bus networks. Data center and buses are two sides of trading sense service in the paradigm. The buses are coordinated to collect data while they are competitors for the reward from the data center. We design a primary-secondary coordination strategy using a Stackelberg game approach, where the data center is the leader while the buses are the followers. With the proposed heuristic algorithm, the Stackelberg game can reach a unique Stackelberg equilibrium (SE), where the data center and each bus respectively obtain maximum utility. The simulations validate the effectiveness of our coordination strategy.

I. Introduction

Smart city is envisioned to sense, analyze and integrate key information of city core operating system to make global intelligent decision and response [1]. However, collecting urban information is a crucial challenge due to the wide sensing coverage and the time sensitiveness of sensed data. Traditional wireless sensor networks deploy static sensors over fixed areas and assign mobile vehicles to carry the sensed data back to a data center [2]. However, the high cost of deployment and the limited battery energy of widespread sensors significantly constrain their large-scale application.