Dynamic Compressive Data Acquisition in Vehicular Sensor Network

Abstract

Compressive Sensing (CS) is a promising approach to compress data with spatial-temporal correlation to reduce the communication cost in Vehicular Sensor Network (VSN). However, the current research on the application of CS in VSN does not consider dynamic changes in data sparsity and vehicle distribution, which may lead to unacceptable reconstruction accuracy. At the same time, the loss of compressed data in transmission process also seriously affects data reconstruction. A dynamic compressive data acquisition in VSN is proposed in this paper. This paper first proposes a VSN data acquisition framework based on CS. Then, for the problem of data sparsity and vehicle distribution changes, a measurement rate adjustment method is proposed. Finally, aiming at the problem of degradation of reconstruction accuracy caused by the loss of measurements, this paper presents a method to determine the priority of measurement, and a data transmission strategy based on priority is put forward. The experiment shows that the proposed dynamic compressive data acquisition in VSN improves the reconstruction accuracy by 18.4% compared with the fixed CS approach in the VSN.