Efficient Data Collection Scheme based on Information Entropy for Vehicular Crowdsensing

Abstract

Vehicular crowdsensing system plays an important role in monitoring the dynamic characteristics of real environment with the assistance of mobile edge computing (MEC). However, connected vehicles (CVs) continuously generate a huge amount of sensed data, which causes severe data redundancy and considerable communication overhead. To tackle the above challenge, we propose an efficient data collection scheme for vehicular crowdsensing to mitigate data redundancy and communication overhead while improving data quality. Particularly, we design a sensed data preprocessing mechanism to reduce data volume and guarantee data quality. Besides, a grid selection algorithm is proposed to select the regions with abundant information by taking advantage of information entropy theory. Then we design an online collecting parameter adjustment algorithm, which adjusts collecting parameters according to the result of grid selection algorithm and current data redundancy, thus communication overhead can be reduced. Finally, extensive simulations are conducted to evaluate the performance of proposed scheme. The simulation results demonstrate that our proposed scheme can significantly reduce communication overhead and data redundancy while meeting the data quality requirement of vehicular crowdsensing.